Beruflich Dokumente
Kultur Dokumente
of Tanzania
Edited by
Adam Szirmai and Paul Lapperre
The Industrial Experience of Tanzania
Also by Adam Szirmai
Adam Szirmai
Professor of Technology and Development Studies
Eindhoven Centre for Innovation Studies (ECIS)
Eindhoven University of Technology
The Netherlands
and
Paul Lapperre
Associate Professor of Technology and Development Studies
Director of Education MSc Programme Technology and Society
Eindhoven University of Technology
The Netherlands
Editorial matter and selection © Adam Szirmai and Paul Lapperre 2001
Chapter 1 © Donné van Engelen, Adam Szirmai and Paul Lapperre 2001
Chapter 3 © Adam Szirmai, Menno Prins and Wessel Schulte 2001
Chapter 7 © Bartelt Bongenaar and Adam Szirmai 2001
Chapter 9 © Raymond Duijsens and Paul Lapperre 2001
Chapter 12 © Paul Lapperre 2001
Chapters 2, 4–6, 8, 10, 11, 13–17 © Palgrave Publishers Ltd 2001
Chapter 5 was previously published in Public Choice 89: 3/4, pp. 375–92,
1996. Reprinted with kind permission from Kluwer Academic Publishers.
All rights reserved. No reproduction, copy or transmission of
this publication may be made without written permission.
No paragraph of this publication may be reproduced, copied or
transmitted save with written permission or in accordance with
the provisions of the Copyright, Designs and Patents Act 1988,
or under the terms of any licence permitting limited copying
issued by the Copyright Licensing Agency, 90 Tottenham Court
Road, London W1P 0LP.
Any person who does any unauthorised act in relation to this
publication may be liable to criminal prosecution and civil
claims for damages.
The authors have asserted their rights to be identified
as the authors of this work in accordance with the
Copyright, Designs and Patents Act 1988.
First published 2001 by
PALGRAVE
Houndmills, Basingstoke, Hampshire RG21 6XS and
175 Fifth Avenue, New York, N. Y. 10010
Companies and representatives throughout the world
PALGRAVE is the new global academic imprint of
St. Martin’s Press LLC Scholarly and Reference Division and
Palgrave Publishers Ltd (formerly Macmillan Press Ltd).
ISBN 0–333–80019–2
This book is printed on paper suitable for recycling and
made from fully managed and sustained forest sources.
A catalogue record for this book is available
from the British Library.
Library of Congress Cataloging-in-Publication Data
The industrial experience of Tanzania / edited by Adam Szirmai
and Paul Lapperre.
p. cm.
Includes bibliographical references and index.
ISBN 0–333–80019–2
1. Industrialization—Tanzania. 2. Tanzania—Economic
conditions—1964– 3. Technology—Tanzania. I. Szirmai,
Adam, 1946– II. Lapperre, Paul, 1942–
HC885 .I533 2001
338.09678—dc21
2001021199
10 9 8 7 6 5 4 3 2 1
10 09 08 07 06 05 04 03 02 01
Printed in Great Britain by Antony Rowe Ltd, Chippenham, Wiltshire
Contents
Preface vii
Adam Szirmai
Paul Lapperre
October 1999
vii
Conference Participants
‘The Industrial Performance of
Tanzania’
viii
Conference Participants ix
x
Notes on the Contributors xi
1
2 The Industrial Experience of Tanzania
been essential for early industrial development in Europe. Like Pack and
Paxson, he doubts that simply increasing formal educational levels will
provide solutions to Tanzania’s economic problems. On the other hand,
many of the articles in this volume (van Egmond, Duijsens and Lapperre,
Semboja and Kweka) emphasize the importance of improving education
and in particular technical education.
An article on the informal urban manufacturing sector by Gaillard and
Beernink has a dual purpose. On the one hand, the authors criticize the
neglect of the informal sector in past policy thinking and practice. On the
other hand, they present a state of the art of informal-sector studies, both
in Tanzania and worldwide. They are very critical of the scientific quality
of informal-sector studies, concluding that this branch of research is still
in a pre-scientific stage. Therefore, it is of less use for the formulation of
well-specified informal-sector policies than is to be desired. Nevertheless,
the authors do conclude on the basis of the available information that the
relevance of this sector for economic development is very substantial in
terms of employment, output, income generation and learning potential.
In 1991 no less than 56 per cent of the urban active population in
Tanzania was working in the informal sector, generating 32 per cent of
total urban income. Gaillard and Beernink try to identify the main con-
straints for the future growth of this sector, which include: limited access
to credit, lack of equipment and spare parts, and limited access to more
permanent sites. Training is mentioned least frequently by respondents, as
the informal sector is largely self-supporting with regard to training.
However, the authors conclude that from a long-run perspective, training
for upgrading labour in the informal sector should be part of the policy
package.
The articles by Mbelle, Moshi and Wangwe all focus on the process of
reform and structural adjustment. It is important to note that there is con-
sensus concerning the shortcomings of the model of state-led import-
substitution industrialization dominant in the past. None of the authors
advocates a return to this model. However, their assessments of ongoing
reform policies and their outcomes are mixed.
In Chapter 15, Moshi provides two interesting case studies of major firms
that have experienced a marked turnaround since their privatization, with
increases in sales, productivity, competitiveness and profitability. In an
econometric article on the impact of liberalization in chapter 13, Mbelle
starts by noting that important macroeconomic indicators, such as manu-
facturing growth and employment, have started improving since 1994. For
1998 he even records a GDP growth in manufacturing of 8.1 per cent. GDP
growth in the total economy has also picked up in recent years (Table 1.9).
Like Wangwe, Semboja and Kweka, and Moshi, he also points to increased
technological dynamism and ‘offensive strategies’ in some of the manufac-
turing firms.
8 The Industrial Experience of Tanzania
1 Introduction
11
12 The Industrial Experience of Tanzania
2 Initial conditions
During the colonial interlude lasting from 1885 to 1961, Tanzania is sub-
jected to German rule (until 1918) and later to British rule. During most of
the colonial period, industrialization receives little attention. There is some
domestic manufacturing in the nineteenth century (handicraft production
of iron products, cotton cultivation, spinning and weaving), but under
German rule (1890–1918) this is largely replaced by imports. Between the
two world wars British colonial policy discourages investment in manufac-
ture for the local market (Coulson, 1982b). In the twilight of colonialism
after World War II, the previously almost non-existent Tanzanian industrial
sector begins to take shape. Initially all industrial efforts are directed at the
processing of primary agricultural commodities such as sisal, cotton, coffee
and tobacco. These commodities are processed for the export market.
In the 1950s production of consumer and intermediate goods for the
domestic market gains in importance. But the share of these products (for
example, woodwork, paint, soap, steel) in total manufacturing remains
small (Silver, 1984). The demand for intermediate and capital goods (for
example, iron, vehicles and machinery) is largely satisfied by imports
(World Bank, 1987a). To encourage the establishment of industrial enter-
prises, the colonial government provides fiscal advantages (tax relief on
capital expenditures, refunds of custom duties) and some measure of pro-
tection (import duties) to investors. Furthermore, the government is
willing to participate in investment projects, provide specialist research
and advisory services, create industrial sites and improve infrastructural
facilities. Since Tanzania lacks a well-developed capitalist entrepreneurial
class (Rweyemamu, 1973), the incentive package in this period aims at
stimulating private foreign investment. Unfortunately for Tanzania, the
other member states of the East African Community, Kenya and Uganda,
are providing similar incentive packages. Moreover, Kenya possesses the
largest industrial base, the best financial and physical infrastructure, the
most concentrated domestic market and the largest settler community.
Thus it is a more attractive country for foreign investors than Tanzania.
Silver (1984) concludes that the policy incentive package of the 1950s
cannot have been very effective in stimulating foreign investment in
Tanzanian manufacturing.
As a result of the unstimulating policy environment and the continued
focus on agricultural commodity processing, Tanzania has not experienced
a great deal of industrial development at the eve of independence. In 1961
the manufacturing sector contributes 3.6 per cent to GDP at factor costs
(Rweyemamu, 1973). Of the total number of 891 industrial establishments,
including cottage establishments (Silver, 1984), 245 are engaged in primary
products processing (sisal, tea, tobacco, cotton), 248 in basic food process-
ing and manufacturing (sugar, salt, dairy products, grain milling, bakeries,
Public Policy and Industrial Development 13
3 Policy periods
I. Industrial trade strategy a. Import substitution Tariffs, import quotas, import licences, (real) exchange rate
appreciation, export taxes, export licensing, export duties
b. Monetary sector Low/moderate/high Interest rate control, credit allocation, money supply
e. Prices and internal trade Low/moderate/high Producer and consumer price controls, confinement
15
16
Table 1.1 (continued)
III. Relative roles attributed to the a. Private sector Privatisation, investment incentives (tax reductions,
private and the public sector monopoly privileges, subsidies), openness to DFI, land
allocation and tenure
IV. Inflow of foreign finance: relative a. direct foreign investment Subsidies and tax incentives for foreign investors,
roles attributed to (DFI) monopoly privileges
Source: van Engelen (1996). Notes: policy instruments which have been omitted are the provision of infrastructure and vocational training.
Nevertheless, these unclassifiable policy instruments are taken into account in Section 5 of the paper.
Public Policy and Industrial Development 17
Year Nominal exchange rate Parallel market rate Ratio parallel/ Real exchange rate
(TSh/US$) (TSh/US$)a nominal (%) (1966 = 100)b
Sources: Lofchie (1988); Maliyamkono and Bagachwa (1990); Musonda (1992); Tanzania
Economic Trends (1994); Bank of Tanzania (1995); Economic Research Bureau of the University
of Dar es Salaam, calculations made by A.V.Y. Mbelle and J.J. Semboja (1996).
Notes: a Estimates of parallel rates have been made by van Engelen (1996) using sources
mentioned above. b The real exchange rate index is defined as the Tanzanian inflation rate
relative to that of the USA, times 100. n.a. = not available.
one third of total capital expenditures. When the volume of aid inflows
starts declining in the early 1980s, manufacturing can still count on a sub-
stantial share of foreign aid. Most of the aid provided between 1973 and
1984 is in the form of donor-tied capital intensive projects (World Bank,
1987a). There is an increased reliance on foreign know-how.
Apart from the shift towards a higher degree of dependency on foreign
aid, this policy period is characterized by the emergence of a high degree of
direct regulatory control. From 1973 onwards a new import licensing
system is introduced. Along with a system for the administrative allocation
of foreign exchange (Finance Act of 1973), this system allows for full
control of (legally) imported goods (Musonda, 1992).
The allocation of the import licences to manufacturers is considered on a
firm-by-firm basis, and is driven by ‘the objective of supporting fiscal
revenue earners … and keeping most existing enterprises alive, on the one
hand, and by particular pressures, representations and ad-hoc decisions on
the other’ (World Bank, 1987a). The resulting allocation pattern is pre-
sented in Table 1.3, which shows that from 1979 onwards the manufactur-
ing sector can on average count on being allocated less than one fifth of
the requested amounts of foreign exchange for the purchase of raw materi-
als, machinery and spares from abroad. In 1981 no foreign exchange at all
is made available for the importation of machinery and spares for manufac-
turing purposes. The same is true for raw material imports for manufactur-
ing during the first six months of 1984.
Another way in which the government augments the degree of direct
regulatory control of the manufacturing sector is by introducing a full-scale
Sources: manufacturing data from Mbelle (1988); all other data from World Bank (1984).
Notes: a The 1984 data concern the January–June period. n.a. = not available.
22 The Industrial Experience of Tanzania
price control system (Price Control Act of 1973). Price controls are intro-
duced with the dual purpose of limiting the monopoly pricing power of
domestic producers, and at the same time ensuring that these producers are
guaranteed satisfactory financial profitability (Mongi, 1980). For the domes-
tic producers the financial profitability is taken care of by the use of a cost-
plus pricing method, allowing for a 30 per cent pre-tax rate of return on
assets. The price of imported goods is set using a fixed percentage mark-up
(Maliyamkono and Bagachwa, 1990). Trends in the numbers of
price-controlled products can be derived from Table 1.4. As a result of a
government directive, the number of price-controlled items is reduced from
1980 onwards (United Republic of Tanzania, 1983). Nevertheless, until
1985 almost all consumer goods, a significant number of agricultural inter-
mediate inputs and various construction materials, such as iron sheets,
rolled steel and cement, remain price controlled.
Apart from import licensing and price control, the government uses
confinement as yet another policy instrument to ensure a high degree of
direct regulatory control over the manufacturing sector. As described in the
discussion of the previous policy period, trade is confined to the State
Trading Corporation STC. In order to perform more effectively the STC is
decentralized and reorganized in 1973. The STC’s tasks are taken over by
6 parastatal importing companies and 18 regional trading companies
(Musonda, 1992). Up to 1984 wholesale (and a part of retail) trade is
confined to these trading companies. According to the World Bank (1987a)
this ‘has been associated with poor service, lack of payments and high mar-
keting costs’.
The three policy instruments mentioned above not only allow for
increased control of the manufacturing sector. They also allow the govern-
ment to provide protection to the locally based industries. Import licensing
restricts competitive imports from entering the country, whilst the price
controls and the confinement system raise the prices of imported products
Table 1.4 Price controlled items, 1973–91 (product groups and separate prod-
ucts)
Sources: 1973 data from World Bank (1984); 1979–1988 data from Kiondo (1991); 1991 data
from World Bank (1991).
Note: Items are groups of separate products; n.a. = not available.
Public Policy and Industrial Development 23
Sources: 1966 data from Rweyemamu (1973); 1984 data from World Bank (1987a); 1986 data from Ndulu et al. (1987).
Notes: inf = infinite; n.a. = not available.
Public Policy and Industrial Development 25
1973 and 1984. Nominal interest rate rigidity, preferential credit allocation
and interest rate discrimination once again emphasize the high degree of
direct regulatory control which prevails throughout this policy period.
3.4 Decontrol and the shift away from import substitution (1984–90)
The introduction of the own funds import scheme in June 1984 marks a
radical change in government policy. It marks the end of 11 years of full
scale direct regulatory control of the manufacturing sector. The scheme
encourages Tanzanian citizens to obtain foreign exchange on the parallel
foreign exchange market, from friends, family, accounts abroad and from
foreign investors, so that imports can be financed without making use of
official foreign exchange (Bank of Tanzania, 1988). In Table 1.6 the various
sources of foreign exchange are presented. The table shows that within two
years the own funds import scheme has become the largest foreign
exchange window. Administration of an own funds import licence is not
subject to bureaucratic procedures, and the imports under the scheme are
not subjected to any price controls or confinement rules. Due to the magni-
tude of the own funds window, these conditions imply that considerable
trade liberalization takes place. This trade liberalization is only partial, since
not all goods are allowed to be imported. Nevertheless, the decontrol of
trade marks a jump in policy.
Partial liberalization of imports funded with official foreign exchange
follows in February 1988, when the open general licence (OGL) is intro-
duced. The foreign exchange available through this window is provided by
the World Bank and other donors, and serves to finance high priority
imports. It is allocated on a non-discretionary and automatic basis,
considerably since 1984, ranging from 2 to 84 per cent at the branch level
in 1986. Furthermore, a shift in relative protection has taken place in
favour of the manufacturers of consumer goods.
These conclusions should be drawn with some caution, though. In 1986
the exchange rate is still overvalued, and price controls, import licensing
and confinement are still used as policy instruments. It is more likely that
the 1986 ERPs in Table 1.5 are good approximations for ERPs in 1988. By
then the 1973–84 system of protection is almost fully dismantled. The
remaining protection from this system is cancelled out by tariff reductions
(Musonda, 1992), tariff evasions and exemptions (see also the 1990–95
policy period on this issue).
A final incentive for production for export is given by the introduction of
several export promotion schemes. In 1986 a new retention scheme
replaces the general retention scheme described in the previous section.
Under the new scheme, producers of non-traditional exports products are
granted a 50 per cent retention rate. In 1985 the seed capital revolving
scheme (SCRS) is introduced. Under this scheme a producer is provided
with foreign exchange when starting production for the export market.
Between 1985 and 1989 the number of manufacturers which benefits from
the scheme rises from 18 to 51 (Ndulu and Semboja, 1996). Yet another
incentive for export-oriented production is provided by commodity
exchange programmes (CEPs), which allow manufacturers of export prod-
ucts to exchange their goods for raw materials and spare parts from abroad.
Finally, the export duty drawback scheme (DDS) is established for manufac-
turers of exports, compensating them for the import duties they have to
pay when acquiring foreign inputs (Mbatia, 1993).
Apart from the decrease in direct regulatory control and the shift in the
industrial trade strategy, no other significant changes occur in the other
dimensions of policy during this policy period. With regard to foreign
financial inflows, there is actually an increase in foreign aid flows to
Tanzania and to Tanzanian industry. From 1987 to 1989 the share of
official development assistance (ODA) in GDP rises from 25.5 to no less
than 32.3 per cent (Szirmai, 1997a, table 12.6). UNDP data show that the
share of total industry in external assistance rises from 7.8 per cent in 1986
to 24.8 per cent in 1989 (Table 1.7). From table 1.7 it becomes evident that
external assistance is increasingly provided in the form of financial trans-
fers, rather than in the form of technical assistance (transfer of skills and
technology). This change in emphasis should be understood in the context
of attempts to ameliorate the financial position of the often heavily
indebted public enterprises (Agrawal et al., 1993).
Apart from supporting public enterprises through aid inflows, the
predominance of the public sector is also enhanced by continued lending
(at negative real interest rates) by the government owned banking sector:
‘the financial sector has continued to lend to ailing firms, particularly
28 The Industrial Experience of Tanzania
Year Total external Assistance to Ratio (%) Technical assistance Ratio (%)
assistance industrya to industryb
(1) (2) (2/1) (3) (3/2)
31
32 The Industrial Experience of Tanzania
Year GDP growth rate Inflation Curr. acc./GDP Terms of Trade Debt/GDP GDP growth
(%)a (%)b (%)c (1987 = 100)c (%)c 9 African Economiesd
33
Notes: a Source: Bureau of Statistics (1995a, 1995b, 1995e, 1999). b Source: Bank of Tanzania (various issues).
c
Source: World Bank (various issues). d Source: Maddison (1995). n.a. = not available.
34 The Industrial Experience of Tanzania
Industrial strategy
IS/EP IS IS IS IS IS → EP EP
Control low moderate high high moderate low
Priv./public priv priv. → publ. publ. publ. publ. publ. → priv.
DFI/aid DFI DFI → aid aid aid aid aid → DFI
Macroeconomic
situation relatively relatively first signs crisis internal and internal and
stable stable of imbalance external imbalance external imbalance
35
36 The Industrial Experience of Tanzania
14
12
10
8
MVA/GDP (%)
0
1964 1966 1968 1970 1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994
Year
Figure 1.2 Index of manufacturing gross value added, 1961–95 (establishments with
10 or more persons engaged, 1976 = 100)
120
100
80
Index
60
40
20
0
1961 1963 1965 1967 1969 1971 1973 1975 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995
Year
Source: Chapter 3.
Public Policy and Industrial Development 37
120
100
80
Index
60
40
20
0
1965 1967 1969 1971 1973 1975 1977 1979 1982 1983 1985 1987 1989
Year
Source: Chapter 3.
from 1961 to 1978, a period of collapse from 1978 to 1985 and a period of
uncertain stabilization from 1985 to 1995. Figure 1.3 presents an index of
labour productivity in constant Tanzanian shillings using the output series
of Figure 1.2 and newly revised employment series (see Chapter 3). This
figure illustrates the secular decline in productivity after 1973. Figure 1.4
puts Tanzanian manufacturing productivity in comparative perspective.
This figure derives from a level comparison between Tanzania and the
world productivity leader, the USA, for 1989 (see Chapter 3). The binary
comparison is based on the calculation of industry of origin unit value
ratios, which are used to convert Tanzanian gross value added for purposes
of international comparison. In this paper, the benchmark comparison is
extrapolated using national time series of GDP per person. For Tanzania,
we used the series described in the source notes to Figure 1.2; for the USA
we used national accounts series (see Timmer and Szirmai, 1999).
One of the interesting results documented in Figure 1.4 is that relative
labour productivity in Tanzania shows secular decline from a rather high
initial level. Having started at around 9 per cent of US levels in 1965, com-
parative labour productivity increases until 1973, starts to decline after that
year and falls to 3.7 per cent in 1987.
In the following paragraphs, each period of analysis is discussed in terms
of the interrelatedness of policy, external influences and manufacturing
38 The Industrial Experience of Tanzania
12
10
Tanzania/USA (%)
0
1965 1967 1969 1971 1973 1975 1977 1979 1981 1983 1985 1987 1989
Year
Sources: 1989 benchmark from Szirmai et al. (Chapter 3); index of GDP per person in Tanzania
from Figure 1.2; GDP per person in USA from national accounts sources (see Chapter 3).
5.1 1961–67
Between 1961 and 1967 the growth performance of the manufacturing
sector is impressive. Double-digit growth rates of real value added are
recorded for the whole sector, causing the share of manufacturing value
added in GDP to rise to 10.2 per cent in 1967 (Figure 1.1). During this
period the index of real value added for enterprises with ten or more
persons engaged rises 68 per cent (from 22 to 37; see Figure 1.2), and there
are also increases in labour productivity (see Figure 1.3). Between 1965 and
1967 one can also discern some catch-up in labour productivity levels vis-à-
vis the world productivity leader, the USA (Figure 1.4).
The investment thrust necessary to raise output and labour productivity
levels is mainly provided by foreign owned firms. As Perkins (1983)
emphazises, many of the new firms established after 1961 are owned by
transnational corporations (TNCs), or their Kenyan subsidiaries. In 1965 no
more than 32 per cent of all industrial firms are exclusively owned by
Tanzanians. The predominance of private foreign owned firms can be
explained by the negotiations within the East African Community
Public Policy and Industrial Development 39
(see Section 4), and even more by the policy climate prevailing at the time.
As described in Section 3.1, between 1961 and 1967 policy instruments are
used to encourage private foreign investment. The most notable incentives
are the granting of monopoly or near monopoly power and tariff protec-
tion.
As pointed out in Section 3.1, tariff protection results in high effective
rates of protection, especially for the manufacturers of consumer and inter-
mediate goods. Changes in degrees of import substitution and changes in
industrial structure can be explained in this context. Table 1.11 presents
data on trends in the total supply of manufactured goods in Tanzania. The
total manufactured supply (TMS) is defined as the sum of locally produced
goods (for the domestic market, DM and exports, EXP) and imported
goods, IMP. A declining IMP/TMS ratio, combined with a rising DM/TMS
ratio, implies that imports are being substituted for locally manufactured
goods.2 Table 1.11 confirms that import substitution is taking place
between 1961 and 1965.
From the last three columns of the table it can be seen that import sub-
stitution is primarily taking place in the consumer goods sector, and to a
lesser extent in the intermediate goods sector. This coincides with the
structural changes taking place in the share of branches in gross value
added (10+). As Table 1.12 shows, the shares of consumer goods (ISIC 31
and 32) and intermediate goods (ISIC 33, 34, 35 and 36) in manufacturing
value added rise between 1965 and 1967, whilst the share of capital goods
declines. These changes can be traced back to the pattern of effective pro-
tection (Table 1.5), which stimulates (foreign) entrepreneurs to invest in
consumer and intermediate good import-substitution activities.
5.2 1967–73
Between 1967 and 1973 the role private foreign investors play in the devel-
opment of the manufacturing sector is considerably reduced. From 1967
onwards the public sector provides the main development thrust, as can be
deduced from the rising shares of the public sector in total manufacturing
value added and total manufacturing employment. In 1966 the public
sector accounts for only 5 per cent of total value added. This percentage
has increased to 32 per cent by 1973. The share of public sector employ-
ment in total manufacturing employment rises from 15.5 per cent in 1967
to 46.7 per cent in 1973 (Skarstein and Wangwe, 1986).
These major changes are realized through the nationalization of private
enterprises and by public investment in additional capacity. Public invest-
ment in manufacturing peaks in 1970, when the share of the public sector
in total manufacturing investment reaches 63.7 per cent. 3 Thereafter, the
rate of public investment in manufacturing slows down, since, as the
World Bank (1988) puts it, ‘little is left to nationalize [and] fewer joint ven-
tures are established’.
40
Table 1.11 Import substitution of manufactured goods, 1961–90
Domestic market (%) Exports (%) Imports (%) Consumer (%) Intermediate (%) Capital (%)
Sources: 1961–73 data from World Bank (1987a); 1980–90 data calculated from United Republic of Tanzania (1995c).
Public Policy and Industrial Development 41
5.3 1973–80
In terms of industrial performance, 1973 marks an important trend break.
From 1973 onwards, labour productivity in 10+ manufacturing enterprises
starts declining steadily (see Figure 1.3). The labour productivity index
drops from 110 in 1973 to 75 in 1980, a decline of 5.3 per cent per year.
42 The Industrial Experience of Tanzania
Comparative labour productivity declines from its peak value of 11 per cent
to 7.2 per cent of the US level.
There have been declines in value added between 1973–75 and 1978–80,
but the index of manufacturing value added in 1980 is higher than that in
1973, indicating a net expansion of output over the whole period (see
Figure 1.2). Given the fact that between 1973 and 1980 investments are
largely taking place in capital intensive projects (Perkins, 1983), the combi-
nation of declines in labour productivity and expansion of output suggests
that throughout this period manufacturing enterprises are producing less
and less efficiently.
The decline in efficiency, which also manifests itself in the drop in capac-
ity utilization (World Bank, 1987a; Maliyamkono and Bagachwa 1990), is
closely related to all four characteristics of the policy period prevailing at
the time. First, it should be noted that the increased reliance on foreign aid
results in investments in inefficient, large-scale and capital-intensive pub-
licly owned enterprises. On this matter Perkins (1983) writes that ‘in many
cases it appears that the government approves the establishment of new
parastatals … merely because a foreign equipment supplier or TNC is
willing to implement the project, and to provide equity capital, commercial
credit or a soft loan through the aid programme of its home country’.
These newly established publicly owned enterprises are not in accordance
with resources available within the country, since they prove to be
extremely import intensive, and require highly developed managerial skills
and highly trained workers (see Chapter 5).
On the latter issue, Gulhati and Sekhar (1982) point out that in sub-
Saharan Africa inefficient production in public manufacturing results to a
considerable extent from an ‘acute scarcity of skilled managers, administra-
tors and technicians’. By expanding the Tanzanian public sector through
aid programmes, the strain on the scarce availability of managerial skills
and skilled labour is increased. On top of this, by allowing for the establish-
ment of import-intensive enterprises, the already deteriorating balance of
payments (see Section 4) is put under even more strain. Because of the
balance of payments crisis and the continued establishment of new import-
intensive firms, available foreign exchange has to be spread out more
thinly over claimants. This results in shortages of machinery, spare parts
and raw materials (see Table 1.3). It is not surprising that the paradox of
capital intensive industrial expansion in a situation of scarcity of material,
financial and human resources results in declines in labour productivity.
Another cause of the decline in efficiency should be sought in the combi-
nation of the emphasis on the public sector and the high degree of direct
regulatory control prevailing from 1973 onwards. This combination, which
is not uncommon in developing countries (White, 1984; Szirmai, 1997a), is
known to result in inefficiency, since the imposition of direct regulatory
controls (for example, foreign exchange allocation, price control, confine-
Public Policy and Industrial Development 43
5.4 1980–84
A second turning point in Tanzanian manufacturing performance occurs in
1979. In this year the trend of manufacturing (10+) growth is reversed. As
Figure 1.2 reveals, the index of medium- and large-scale manufacturing
GDP declines rapidly after 1979. In 1984 it is already beneath its 1973 level.
The four years of negative growth cause the share of manufacturing value
added in GDP to shrink from 12.4 per cent in 1979 to 9.1 per cent in 1984
(Figure 1.1).
Between 1979 and 1983 the decline in labour productivity accelerates.
Comparative labour productivity declines dramatically from 8.1 per cent in
1979 to 4.8 per cent in 1983. The divergence is the result of the combination
44 The Industrial Experience of Tanzania
5.5 1984–90
In 1984 the decline in output and labour productivity levels comes to an
end. As Figures 1.2 and 1.3 show, output and labour productivity levels sta-
bilize between 1984 and 1989. So does the trend in the share of manufac-
turing value added in GDP: 1985 and 1989 shares both equal 8.4 per cent.
The declining trend in comparative labour productivity levels, however, is
not reversed. In 1984 the ratio of Tanzanian and US manufacturing labour
productivity levels equals 5.8 per cent. This ratio has dropped to 3.9 per
cent in 1990. Comparative productivity in medium- and large-scale manu-
facturing is well below that of developing economies in Asia, such as
China, India and Indonesia (between 5.7 and 10 per cent of the US level in
1987; see Timmer and Szirmai, 1999). The further decline in comparative
46 The Industrial Experience of Tanzania
5.6 1990–95
With the 1990–95 decontrol and shift away from the predominance of the
public sector in the industrial strategy, it would seem that the many policy
barriers to efficient production are removed. But this is not (or not yet)
reflected in the indicators of manufacturing performance. The share of
manufacturing value added hardly changes between 1990 (7.8 per cent)
and 1994 (7.6 per cent) (Figure 1.1). Output levels in medium- and large-
scale manufacturing do not rise substantially either: the 1995 output level
is only just above the 1989 output level (Figure 1.2). Furthermore, a sample
of 20 manufacturing activities shows only modest improvements in the use
of capacity throughout the period. In 1990, 32 per cent of the installed
capacity in the sample industries is used. In 1994 capacity utilization in
these industries has increased to 37 per cent (United Republic of Tanzania,
1995b).
The absence of significant improvements in manufacturing performance
can be traced back to the macroeconomic situation, the unfavourable
initial conditions Tanzania has had to face and the policy changes in the
period studied. Regarding parastatal reform, for instance, it cannot be
expected that privatization of 34 of the 170 manufacturing parastatals
immediately results in productivity gains. In the words of Adam (1994),
public sector reform is accompanied by ‘changes … in terms of managerial
attitudes, and a greater perception of operating autonomy, which will grad-
ually be translated into efficiency gains and increased profitability’. It is
unlikely that these desirable effects take place at short notice though, a
point which can also be made regarding the inflow of direct foreign invest-
ment after the opening up of the sector to private entrepreneurs.
Another reason why policy changes do not result in improvements in
manufacturing performance can be found in the combination of the liber-
alisation of the banking system and the high inflation rates prevailing
between 1990 and 1994 (Table 1.9). The high inflation rates force the banks
to charge extremely high nominal interest rates in order to maintain posi-
tive real interest rates (see Section 3.5). These high nominal interest rates
are a threat to the liquidity of manufacturing enterprises. Furthermore, the
high inflation rates erode the purchasing power of the bank overdrafts.
Since banks are reluctant to raise bank overdrafts, manufacturers have less
and less access to short-term credit to finance procurement of raw materials
and intermediate inputs. On top of that, non-creditworthy enterprises are
48 The Industrial Experience of Tanzania
no longer allocated medium- and long-term credit, since this would contra-
dict the commercial principles banks operate on from 1991 onwards. As
such, credit has replaced foreign exchange as the major production con-
straint the manufacturing sector faces.
Other production constraints are related to the conditions Tanzania has
had to face since independence. As discussed in Section 2 and in this
section (1980–84 period), the manufacturing sector has been operating in
an environment characterized by poor infrastructure and a shortage of
supervisory and technical manpower. In a more general sense, Helleiner
(1994) points out that ‘industrial development hinges, in any case, upon
local capabilities and appropriate supporting institutions no less than on
incentives, and, in sub-Saharan Africa these have been typically lacking’. In
Tanzania, this lack of local capabilities and supporting institutions has con-
tinued to hamper manufacturing development between 1990 and 1995
(United Republic of Tanzania, 1996b).
Along with the unfair competition by tax-evading importers (see Section
3.5), the production constraints mentioned above result in an environment
which is not very conducive to rapid industrial growth. Neither is the envi-
ronment conducive to structural change. As Table 1.12 shows, in 1995
more than 50 per cent of manufacturing value added (10+) is produced in
the food, beverages and tobacco branch. In 1995 the total consumer goods
sector accounts for 64 per cent of manufacturing value added, higher than
the share recorded 30 years earlier. Thus, in terms of structural change little
has changed throughout these 30 years, a conclusion which can also be
drawn regarding the importance of the manufacturing sector for the
economy as a whole (see Figure 1.1).
6 Concluding remarks
Notes
* Section of Technology and Development Studies and Eindhoven Centre for
Innovation Studies, Faculty of Technology Management, Eindhoven University
of Technology
1 Some classifications focus on the level of dependence on foreign finance, but in
the case of Tanzania this dimension does not help distinguish policy periods.
The dependence on foreign finance continues to be high throughout the whole
period studied.
2 For total manufacturing, one can also take production for the domestic market
(DM) as a proportion of production for the domestic market plus imported
goods (IM). However, these ratios are not available for subcategories of
manufacturing.
3 Calculated from United Republic of Tanzania, Analysis of Parastatal Accounts, and
Bureau of Statistics, National Accounts (various issues).
2
Is African Manufacturing Skill
Constrained?
Howard Pack and Christina Paxson*
1 Introduction
50
Is African Manufacturing Skill Constrained? 51
T (t ) − A(t )
A(t ) / A(t ) = ( h) (2.1)
A(t )
Is African Manufacturing Skill Constrained? 53
Given (2.1) and (2.2), the underlying differential equation implies that the
potential equilibrium path of technology of an LDC firm is
Equation 2.4 underlines the fact that, at a given point in time, the level of
productivity A(t) for a given firm may be weakly associated with h unless
firms’ ability to obtain new technology varies systematically with h. Unlike
formulations which treat education or purchases of knowledge as conven-
tional inputs in the production function, Equation 2.4 implies that human
capital will have no effect on the level of output obtained with conven-
tional inputs unless F > 0, which can only occur if new productive inputs
are introduced.
Schultz (1975) argued that in addition to the ability to absorb new tech-
nology, education would be important if there were a changing sectoral
structure or internal product mix of firms, more highly educated workers
54 The Industrial Experience of Tanzania
Country Growth rate of Share of machinery and transport Share of other manufactured
imports, 1980–93 equipment in total imports imports in total imports
Kenya –0.8 23 25 28 29
Ghana 2.7 40 26 28 31
Zimbabwe 0.2 65 36 26 31
Korea 11.4 34 34 23 29
Malaysia 9.7 46 54 28 30
55
56 The Industrial Experience of Tanzania
Light industry 57.8 59.9 52.7 55.2 49.3 53.3 41.5 29.6 47.8 30.8
Heavy industry 42.2 40.1 47.2 45.1 50.7 46.7 58.5 70.4 52.2 69.2
SD of the % change
in sectoral value
shares in 28 ISIC
sectors 1.58 1.25 2.05 2.18 2.98
The Kenyan data are drawn from a survey of 223 manufacturing firms.
The original sample consisted of 162 formal sector firms drawn from the
Central Bureau of Statistics register of firms, and 61 informal (that is non-
registered) firms. Four sectors are represented: food processing, textiles and
garments, woodworking, and metalworking. The sample is stratified by
sector and (for formal firms) by firm size. Detailed information on the
survey design and the procedures used to compute weights for the strata
are in World Bank (1993b). Although the descriptive statistics presented
below make use of data on all 223 firms, we had to exclude 51 firms when
estimating cost functions, because of missing or incomplete data on costs,
output, capital values and other variables used in the analysis.
In Zimbabwe 201 firms were surveyed in the summer of 1993. The firms
were selected from two sources. Large firms, with 50 or more employees,
were drawn from a firm registration list provided by the Central Statistical
Office. Smaller firms were drawn from a list compiled by the Gemini Survey
of small-scale firms, conducted in 1991 (see World Bank, 1994b, for further
details on the sample design and the sample weighting procedure). We use
data on 200 firms (one firm had no sector information and was excluded),
and lose an additional 30 firms for the estimates of cost functions because
of missing data.
The sample from Ghana consists of 186 firms surveyed in the summer of
1992. The sample design is described in World Bank (1993c). The design of
the survey appears to have been complicated by the fact that two censuses,
an industrial census and a population census, yield inconsistent informa-
tion on the numbers of small-scale enterprises (World Bank, 1993c, p. 28.)
Because of this ambiguity, we do not use population weights, the number
of firms in the population in each stratum, in calculating the summary sta-
tistics that follow, but the means of all firms in the sample. It is thus possi-
ble that our sample statistics may not accurately reflect the characteristics
of the population of firms in Ghana.
The firms in the Ghanaian sample tend to be smaller than the firms in
Zimbabwe and Kenya. Furthermore, there is no simple definition of
“formal” versus “informal” firms, as there is for the other two countries. In
what follows we sometimes split the Ghanaian sample between ‘large’
firms, defined as those with 10 or more employees (as opposed to more
than 50 employees for Zimbabwean formal-sector firms), and small firms.
The Ghanaian data also have a larger fraction of missing values than the
other two surveys. Of the original 186 firms, only 132 can be used to esti-
mate cost functions. Valuation of the capital stock of firms appears to have
been a particular problem in Ghana, and so we do not include capital stock
in the cost functions: doing so would have resulted in losing 19 more firms.
The results for Ghana should be treated with extra caution, because of the
small sample and the selection problems that may result from missing
dependent variables.
58 The Industrial Experience of Tanzania
Note: Numbers indicate numbers of sampled firms. Percentages indicate estimated percentages
of firms in the population, and are calculated using population weights provided by the survey
study.
Note: Numbers indicate numbers of sampled firms. Percentages indicate estimated percentages
of firms (with owners) in the population, and are calculated using population weights provided
by the survey study.
Is African Manufacturing Skill Constrained? 59
Note: Numbers indicate numbers of sampled firms, and percentages indicate percentages of
sampled firms. Population weights are not used.
Note: Numbers indicate numbers of sampled firms. Percentages indicate estimated percentages
of firms in the population, and are calculated using population weights provided by the survey
study.
60 The Industrial Experience of Tanzania
Note: Numbers indicate numbers of sampled firms. Percentages indicate estimated percentages
of firms in the population, and are calculated using population weights provided by the survey
study.
and cannot be used), and these indicate that the majority of managerial
employees have attended secondary school, and roughly a quarter of the
sampled firms with managerial employees list the average education level
of their managers at the university level. In the cost functions presented
below we use a variety of skill measures that reflect the education levels of
both owners and managers.
We have no direct measures of the rate at which different firms acquire
new technology (that is, the term F in Equation 2.4). However, it is likely
that the acquisition of new technology from abroad is positively correlated
with activities in foreign markets, and Table 2.6 presents information on
these activities.
The predominant picture is of firms that have little involvement in inter-
national markets. The vast majority of all firms import none of their raw
materials, export none of their output, and do not have any foreign owner-
ship. The firms were also asked about the use of foreign technology licences
and foreign technical assistance. In Kenya only 9 out of 220 firms that
answered these questions had foreign licences, and 17 out of 220 received
Is African Manufacturing Skill Constrained? 61
Note: Numbers indicate numbers of sampled firms. For Kenya, the percentages are calculated
using population weights provided by the survey. The data from Ghana are unweighted.
where c is costs defined as the cost of labour, raw materials and indirect
costs such as utilities and rent; p is a vector of input prices, q is the value of
output, k is the replacement value of the firm’s capital equipment, and A is
a measure of the technology level of the firm. This form of the cost func-
tion follows from the assumption that the short-run production function
has the form qi = AiQ(Mi, ki), where Mi is a vector of variable inputs and Ai
depends on the level of ‘technological’ inputs such as education. The
Nelson-Phelps view discussed in Section 2 implies that Ai is a function of
both the skill level of the firm (measured by hi) and also by factors that
62
Table 2.6 Involvement in foreign markets
Kenya
Percentage of raw materials 121 (76.1%) 32 (11.4%) 15 (3.3%) 25 (3.3%) 25 (5.9%)
imported
Percentage of output exported 170 (78.0%) 29 (6.6%) 7 (0.1%) 4 (0.2%) 8 (0.3%)
Percentage of foreign 184 (96.5%) 4 (0.2%) 6 (1.7%) 10 (0.3%) 14 (1.3%)
ownership of the firm
Zimbabwe
Percentage of raw materials 92 (93.0%) 57 (4.4%) 20 (0.9%) 13 (0.4%) 12 (1.3%)
imported
Percentage of output exported 104 (92.2%) 58 (4.4%) 23 (1.1%) 9 (1.1%) 6 (1.3%)
Percentage of foreign 155 (96.7%) 12 (0.5%) 7 (0.6%) 8 (0.9%) 15 (1.3%)
ownership of the firm
Ghana
Percentage of raw materials 127 (72.6%) 12 (6.9%) 6 (3.4%) 11 (6.3%) 19 (10.9%)
imported
Percentage of output exported 163 (90.1%) 11 (6.1%) 2 (1.1%) 1 (0.6%) 4 (2.2%)
Percentage of foreign 154 (82.8%) 2 (1.1%) 12 (6.5%) 13 (7.0%) 5 (2.7%)
ownership of the firm
Note: The numbers in parentheses indicate percentage of firms. The percentages for Kenya and Zimbabwe are calculated using population weights. The
percentages for Ghana are unweighted.
Is African Manufacturing Skill Constrained? 63
affect the firm-specific rate of flow of new technologies from abroad (mea-
sured by F), the two entering the cost function interactively.
We start by estimating basic cost functions that do not include any vari-
ables that may affect Ai. Because we have small samples of cross-sectional
data, and no data on input prices p, we employ simple specifications for
the short-run cost functions. We work with the following:
Table 2.7 Basic cost functions for formal sector firms or (for Ghana) large
firms (absolute values of t-statistics in parentheses)
66
Kenya: formal firms (obs = 125) Sample means
Owner/manager’s education:
Primary school –0.038 (0.32) –0.017 (0.15) –0.053 (0.45) 0.32
Secondary school –0.062 (0.47) –0.065 (0.49) –0.009 (0.07) 0.18
University –0.056 (0.47) –0.041 (0.33) 0.032 (0.23) 0.31
% raw materials imported 0.0006 (0.46) 0.0000 (0.04) 26.8
% output exported –0.0034 (1.78) 0.0014 (0.59) 8.9
% foreign ownership –0.0019 (1.16) 0.0001 (0.04) 10.2
Owner/manager education is
university, times:
% raw material imported 0.0010 (0.43) 16.5
% output exported –0.0115 (3.20) 4.3
% foreign ownership –0.0023 (0.72) 6.5
R2 .95 .95 .96
Kenya: all firms (obs = 172)
Owner/manager’s education:
Primary school –0.054 (0.56) –0.034 (0.35) –0.042 (0.42) 0.31
Secondary school –0.005 (0.05) 0.004 (0.03) 0.033 (0.26) 0.15
University 0.016 (0.15) 0.039 (0.34) 0.076 (0.60) 0.24
% raw materials imported 0.0000 (0.02) –0.0008 (0.47) 22.3
% output exported –0.0026 (1.26) 0.0027 (1.09) 6.5
% foreign ownership –0.0020 (1.27) –0.0021 (0.84) 8.2
Owner/manager education is
university, times:
% raw material imported 0.0017 (0.74) 12.4
% output exported –0.0123 (3.33) 3.1
% foreign ownership 0.0011 (0.32) 5.3
2
R .96 .96 .97
Notes: Absolute values of t-statistics in parentheses. Also included in cost functions were the logarithms of output and capital, each of these terms
squared, and an interaction term between output and capital (as in Table 2.7).
Table 2.8b Cost functions with skill measures included, Zimbabwe
Owner/manager’s education:
Secondary school 0.0216 (0.15) .0107 (0.08) –0.041 (0.29) 0.48
University –0.0373 (0.25) –0.0321 (0.21) 0.102 (0.63) 0.42
% raw materials imported 0.0004 (0.24) 0.0024 (1.19) 19.2
% output exported 0.0030 (1.46) 0.0033 (1.33) 13.0
% foreign ownership 0.0011 (0.80) 0.0032 (1.62) 14.0
Owner/manager education is
university, times:
% raw material imported –0.0044 (1.50) 8.7
% output exported –0.0025 (0.61) 5.7
% foreign ownership –0.0041 (1.58) 8.0
68
Ghana: large firms (10 or more employees) (obs = 80) Sample means
Owner/manager’s education:
Primary school 0.043 (0.22) 0.038 (0.19) 0.044 (0.22) 0.39
Secondary school –0.087 (0.42) –0.091 (0.44) –0.135 (0.61) 0.33
University 0.427 (1.71) 0.426 (1.70) 0.340 (1.22) 0.16
% raw materials imported –0.0006 (0.29) –0.0041 (1.29) 20.0
% output exported –0.0023 (0.78) –0.0034 (0.77) 6.7
% foreign ownership 0.0044 (1.36) 0.0099 (1.71) 10.6
Owner/manager education is
university, times:
% raw material imported 0.0059 (1.40) 11.5
% output exported 0.0022 (0.38) 3.7
% foreign ownership –0.0070 (1.14) 6.6
R2 .93 .93 .94
Ghana: all firms (obs = 132)
Owner/manager’s education:
Primary school 0.389 (2.29) 0.386 (2.24) 0.389 (2.23) 0.45
Secondary school 0.216 (1.16) 0.215 (1.15) 0.190 (0.96) 0.27
University 0.526 (2.12) 0.519 (2.07) 0.465 (1.72) 0.11
% raw materials imported –0.0002 (0.09) –0.0019 (0.57) 12.8
% output exported –0.0027 (0.75) –0.0038 (0.72) 4.0
% foreign ownership 0.0027 (0.80) 0.0045 (0.92) 7.2
Owner/manager education is
university, times:
% raw material imported 0.0035 (0.75) 7.1
% output exported 0.0024 (0.35) 2.3
% foreign ownership –0.0037 (0.54) 4.0
R2 .92 .92 .92
Notes: Absolute values of t-statistics in parentheses. Also included in cost functions were the logarithm of output and the logarithm of output squared
(as in Table 2.7).
Is African Manufacturing Skill Constrained? 69
The firms in our sample have not been challenged by newer technologies
where the payoffs to domestic skills are likely to be largest. Simultaneously
the failure of agricultural productivity to grow has limited the economy-
wide growth in per capita income. Thus the change in sectoral structure
engendered by different income elasticities among industrial goods has not
had an effect. The absence of a need to alter production methods or sec-
toral mix has reduced the payoff to education.
Even if there were an inflow of technology, and high education made its
absorption potentially productive, the absence of a competitive environ-
ment limits these gains. Much of the empirical testing of the role of educa-
tion in adjusting to technical change has been in the agricultural sector,
which is usually much more competitive than the industrial sector, particu-
larly in developing countries (see the references in footnote 3). In contrast,
in Ghana, Kenya and Zimbabwe the survival environment is lax. Tariffs
and quotas protect industrial firms from external competition, while inter-
nal rivalry is reduced by the relatively small number of firms producing a
specified product. The shift in short-run supply curves induced by abnor-
mal profits in competitive markets is not forthcoming, as there are a
limited number of firms capable of entering most industries. Firms possess-
ing technical abilities may choose not to utilise them to increase productiv-
ity or reduce costs. The abilities of skilled personnel may be devoted to
insuring the existence and exploitation of profit possibilities made possible
by tariffs, quotas and cheap credit.
Firm-level efforts to improve productivity are the result of an income and
a substitution effect. High levels of protection, ceteris paribus, increase the
firm’s profits per unit of effort. If the foregone ‘leisure’ or easy life is a
normal good, protection will tend to reduce cost saving efforts. However,
higher levels of effective protection increase the opportunity cost of forgo-
ing additional output from productivity-augmenting activities. Thus, the
impact of firm-level abilities on productivity in the presence of protection
is ambiguous. An absence of an association between firm skills, however
measured, and productivity levels among firms is not an indication that
such skills do not provide the potential to raise productivity. Rather, such a
capacity may not be deployed in pursuit of greater cost reduction but in
the search for greater rents or an easier life.
6. Conclusions
measured. Our results are not necessarily in conflict with this view. They
should be interpreted as indicating that in the non-competitive industrial
sector with very little inflow of new technology the contribution of techno-
logical abilities, however measured, is limited. If there were a liberalisation
of the economy that generated greater competition or if there were an
acceleration of export growth, permitting the import of new technology
embodying inputs, the contribution of local skills would become more
significant in raising output. The experience of other countries also sug-
gests that as the economy becomes open to flows of international knowl-
edge, whether through informal transfers from purchasers of exports or
through technology licences, the technological capacity of local industry
becomes important.10
The policy implications of the empirical analysis are clear. Absent greater
prospective competition, continued efforts to develop high level industrial
skills may be wasteful. The conundrum is that the absence of these skills
may limit the benefits to the industrial sector from future liberalisation,
and lead to a weak supply response to improved incentives.
Notes
* University of Pennsylvania and Princeton University. We are grateful to Tyler
Biggs of the Africa Technical Department of the World Bank for making avail-
able the data upon which this paper is based. We received helpful comments
from participants at a symposium at the Eindhoven Technical University. A.
Szirmai provided many helpful suggestions. Part of the research was financed by
the World Bank’s Regional Program on Enterprise Development. Pack gratefully
acknowledges partial support from the University of Pennsylvania Research
Foundation and the World Bank Development Research Group.
1 Meier and Steel (1989) contains a good selection of representative views.
2 For a general discussion of the technological requirements for industrial devel-
opment see Lall (1990). Pack (1993) provides an overview of industrial develop-
ment in Africa and extensive references to the literature.
3 For studies demonstrating the interaction of education and technological
change in agriculture see Welch (1970), Foster et al. (1995a, 1995b), Rosenzweig
(1995).
4 The measures described in the text and shown in Table 2.1 could be improved.
For example, standard international trade data contained in the UN Yearbook of
International Trade Statistics allow a disaggregated breakdown of imports that
would allow more precision with respect to both the equipment and intermedi-
ate imports. However, the latest date for which these are available for the African
economies is 1990, two years before the firm-level surveys were carried out. Thus
we have cited the aggregate data generated by the World Bank.
5 In almost all cases the sectoral shares of the 28 sectors, not reproduced in Table
2.2, show little change, with the exception of the Zimbabwean food and bever-
age sectors. The greater summary measure in Zimbabwe compared to Ghana and
Kenya is largely due to the expansion in the share of the beverage sector, a sector
in which the technology employed has not changed.
72 The Industrial Experience of Tanzania
6 It should be noted that the use of sample weights together with small sample
sizes can yield a misleading picture of the activities of these firms. For example,
the Kenyan numbers indicate that 58.7 per cent of wood-sector workers are in
the ‘miscellaneous’ category that includes ‘carvings, camping equipment, and
automobile trim.’ This is due to one 70-person firm that makes camping equip-
ment, and which was given a sample weight four times the sector average.
Likewise, the result that 41 per cent of food and beverage workers make animal
feed and hay is due to one poultry-feed firm with 420 workers and a sample
weight three times the sector average.
7 The survey does not ask whether the owner and manager are the same person,
and this might account for the similarity between the education of owners and
managers.
8 We also experimented with including dummy variables for the city in which
firms are located, and these were also insignificant. The results are consistent
with there being neither regional nor sectoral variation in input prices.
9 One of the benefits of panel data would be the ability to address the question of
whether efficient firms are more successful exporters or whether exporting gen-
erates learning.
10 See the East Asian Economic Miracle (World Bank, 1993d), chapter 6.
3
Measuring Manufacturing
Performance in Tanzania: GDP,
Employment and Comparative
Labour Productivity, 1961–95
Adam Szirmai, Menno Prins and Wessel Schulte*
1 Introduction
73
74 The Industrial Experience of Tanzania
1946 101 b
1957 293 b 17
Independence (1961) 380 b 22 22 3.5
1965 569 38 28 9.2
1978 1 282 51 110 12.3
1989 886 –33 124 8.4
Source: Number of establishments 1946–65 from Survey of Industries 1965 (Central Statistical
Bureau, 1966), table 5; persons engaged and value added 1961 from Skarstein and Wangwe
(1986), p. 79; persons engaged 1978 and 1989 from Census of Industrial Production 1978 (Bureau
of Statistics, 1995b) and Census of Industrial Production 1979 (Bureau of Statistics, 1994a); value
added 1978 and 1989 from Bureau of Statistics (1995c), table 7.2.
Notes:
a
All data refer to establishments with 10 or more persons engaged, except for value added data
which refer to the entire manufacturing sector. Data until 1965 include activities in sisal
decortication and motor vehicle repair, which are excluded in 1978 and 1989.
b
Number of establishments which were still in production in 1965.
c
Average annual increase in the number of establishments.
The main data sources for the manufacturing sector in Tanzania are the
annual survey of industrial production (Bureau of Statistics, ASIP, various
issues), the 1978 and 1989 censuses of industrial production (Bureau of
Statistics, 1985b, 1993a, 1994a) and the quarterly surveys of industrial pro-
duction covering establishments with 50 or more persons engaged (Bureau
of Statistics, QSIP, various issues). Additional sources are input-output
tables (Bureau of Statistics, 1986), informal sector surveys (Planning com-
mission, 1991, 1995), and price indices and economic surveys (Bureau of
Statistics, 1972, 1977, 1982, 1985a). Our main source for nominal value
added in 10+ manufacturing are the surveys and censuses. Figure 3.1 sum-
marizes these sources and compares their coverage.
IS- IS-
Coverage a b
Census I/O table Census surveyd surveyd
1+
Census
5+
Notes: ASIP = Annual Survey of Industrial Production. IS-survey = Informal Sector Survey.
I/O table = Input–Output table. Ec. Surv. = Economic Survey.
a
1961 Census results are not comparable to later industrial surveys/censuses.
z
The coverage of the I/O table is not based on a canvassing procedure of 1+ establishments, but
on estimates based on data such as the ‘final demand’ for certain manufacturing products.
Estimates thus attempt to reflect 1+ coverage.
c
Although no ASIP was published in 1975–77, data were collected through the ASIP question-
naires and the aggregates were published in the Economic Survey. For 1978, 10+ estimates have
been published in the Economic Survey, since census results did not became available till 1985.
d
The Informal Sector Surveys covered establishments with 5 or fewer paid employees not using
high technology. This does not correspond to our categorization of 1–4 persons engaged.
76 The Industrial Experience of Tanzania
Adjustments for these sources of error are made for two periods: 1978–90
and 1965–78.
Additional sources of error, which will not be discussed in this article, are
errors in the raw data of the 1989 census, the changing categorization of
activities in economic sectors and inadequate treatment of the small-scale
and informal sector in the national accounts, using rules of thumb (see
Prins and Szirmai, 1998, sections 2.2.2, 2.4 and 2.5). The issue of changes
in categorization has been dealt with by aggregating manufacturing sectors
into six major branches. The informal sector merits treatment in a separate
article (see also Chapter 14). This paper limits itself to medium- and large-
scale enterprises with ten or more persons employed.
3.1 Coverage
The framework for data collection is the directory of establishments
maintained at the industrial section of the Bureau of Statistics. The cover-
age of medium- and large-scale establishments in the directory varies over
time and does not adequately reflect changes in the real volume of activi-
ties. There are particularly large differences in coverage between survey
years and census years (see Prins and Szirmai, 1998, table 2.2). For
instance, the directory for the 1974 survey lists 499 enterprises. This
jumps to 1282 in the census year 1978. In the survey year 1988 the direc-
tory includes 711 establishments; this jumps to 866 in the census year
1989. In some cases establishments, such as furniture making and tailor-
ing, have deliberately been omitted from the annual surveys (omitted
establishments). In most cases, the differences are unintentional (under-
coverage).
St
tnrt nrt (t − i ) CPI t
UEt = ∑ ue
i= 1
t ,i and uet ,=
i
Nt tnrt
CPI t − i
− 1
(3.1)
The estimate for the overall underestimation of value added was calcu-
lated as a weighted average of the degrees of underestimation per size class,
weighted with value added weights per size class from the census.6
enormous amounts of costs allocated to the residual all other costs cate-
gory. In particular they have allocated large amounts of interest payments
to this category, where they do not belong.7
This conclusion is based on the following reasoning. The questionnaire
distinguishes 16 categories for production costs. One of these categories is
bank charges and insurance paid, which explicitly excludes interest costs.
However, the instructions for the category all other costs do not indicate
that interest costs should be excluded. They explicitly exclude labour costs,
sales taxes, corporate taxes, excise duties and depreciation, but there is no
mention of interest at all. In Figure 3.2 we can see how important all other
costs are compared to most other intermediate input cost categories. From
an analysis of firm level-data for 175 large firms accounting for 96 per cent
of ‘all other costs’, we conclude that the all other costs category indeed
includes interest payments which should have been allocated to value
added. Some other components of value added have also been erroneously
included in all other costs, because they were not identified elsewhere in the
questionnaire. These cost categories are bad debts, directors’ fees and dona-
tions.
A less important conceptual error is that profits from sale of fixed assets
should not have been included in the calculation of gross output. As gross
output is too high, gross value added at factor cost is overstated by
0.3 billion TSh.
60
Intermediate inputs (billion TSh)
50
40
30
20
10
0
Electr. & Water
Transport
Chem. & Pack.
Miscellaneous
Bank charges &
Raw materials
Professional &
Cost of resales
services
Fuel, Lubr.,
office exp.
materials
insurance
80 The Industrial Experience of Tanzania
1961 243
1962 272
1963 276
1964 350
1965 267 282 307 405 6 9 32 52
1966 295 297 327 441 1 10 35 49
1967 319 323 344 460 1 6 34 45
1968 378 390 402 522 3 3 30 38
1969 475 497 659 5 33 39
1970 561 571 766 2 34 37
1971 643 657 874 2 33 36
1972 806 1 087 35 35
1973 914 1 244 36 36
1974 1 157 1 537 33 33
1975 1 246 1 863 50 50
1976 1 480 2 143 45 45
1977 2 075 2 678 29 29
1978 2 186 2 926 34 34
1978* 2 842 2 842 2 842 2 926 0 0 3 3
1979 2 927 2 927 2 985 3 238 0 2 8 11
1980 2 891 2 900 3 279 3 622 0 13 10 25
1981 3 108 3 148 3 983 4 555 1 27 14 47
1982 3 204 3 333 4 862 5 873 4 46 21 83
1983 3 620 3 994 5 706 6 812 10 43 19 88
1984 4 417 5 269 7 241 8 094 19 37 12 83
81
82
Table 3.2 (continued)
Sources: Unadjusted data: 1965–74: ASIP; 1975–78: Economic Survey; 1978: Census; 1979–1988, 1990: ASIP; 1989: 1989 Census; adjusted data: Prins and
Szirmai (1998) (appendixes A–G).
a
Change calculated as the percentage change of adjusted value compared to the pre-adjustment value.
b
Other costs based adjustment.
c
Non-response based adjustment.
d
Omitted establishments based adjustment.
e
Undercoverage based adjustment.
Table 3.3 Adjusted nominal value added, 1961–1995 (thousands TSh)
Adjusted Unadjusted
83
84
Table 3.3 (continued)
Adjusted Unadjusted
1982 2 026 862 1 198 534 499 515 982 171 246 378 919 253 5 872 714 3 203 832
1983 2 067 669 1 812 035 549 270 889 370 326 831 1 167 309 6 812 485 3 619 760
1984 2 582 369 2 022 511 760 200 1 228 673 170 774 1 329 517 8 094 045 4 417 219
1985 3 148 163 2 238 412 870 414 1 398 033 384 734 1 772 483 9 812 238 5 111 606
1986 3 099 117 2 577 458 1 107 314 2 189 489 917 068 2 469 293 12 359 739 6 412 236
1987 5 157 851 5 331 612 1 546 226 4 903 191 1 237 203 3 503 219 21 679 302 11 062 008
1988 7 666 631 6 865 440 2 044 644 4 960 480 1 359 874 3 744 791 26 641 860 11 357 863
1989 12 441 829 5 826 626 3 017 332 4 901 342 1 422 835 5 042 905 32 652 870 21 474 018
1990 17 149 464 5 597 233 3 755 861 5 432 448 1 176 377 5 570 034 38 681 417 23 955 853
1991 21 232 538 6 459 244 4 328 090 6 944 192 2 059 175 7 105 139 48 128 378
1992 25 522 856 7 033 218 4 138 842 7 650 103 1 809 982 7 317 953 53 472 955
1993 30 743 364 7 567 179 7 639 361 8 002 358 2 427 974 7 154 866 63 535 101
1994 43 202 958 8 666 806 6 716 699 10 917 273 2 781 754 9 873 702 82 159 192
1995 53 931 263 12 257 957 8 543 564 12 284 511 3 770 135 8 030 217 98 817 647
For most years official real value added series have only been constructed
for total manufacturing. An exception is the index of industrial production
(IIP), published in the quarterly survey of industrial production (QSIP) from
1985 onwards. Real value added series have been constructed at 1966 base
prices (1960–80), 1976 prices (1964–94) and 1985 prices (1986–94). For
1976–84 an indirect approach has been followed, deflating current value
added (1+) from the national accounts with a cost of living index of cloth-
ing and footwear (Bureau of Statistics, 1985a). For all other years an direct
approach has been followed, in which quantity relatives have been
weighted with base year value added shares. For the 1966 series no sources
and methods are available. The 1976 series are based on quantity relatives
for 14 commodities. The 1985 series are based on 120 commodities derived
from the quarterly survey among 50+ enterprises.
We constructed a new index 1965–95 based on the following principles:
use of as much commodity information as possible, consistent application
of the direct approach of weighting quantity relatives to calculate a
Laspeyres fixed base weighted index, and construction of indexes for six
branches of manufacturing as well as for total manufacturing.
data from the 1985 survey. Moreover, the analysis of the surveys
made clear that the reliability of the value added data in the mid-eighties
is questionable. In our opinion the most reliable source for reweighting
the IIP (1985-95) is the data from the 1989 census of industrial pro-
duction, as adjusted in this report (see Section 3.4). We have reweighted
the index of industrial production utilizing (adjusted) 1989 census value
added as base year weights. The weighting procedures are the same as
explained above for the 1965–85 index (see Prins and Szirmai, 1998,
appendix I and Table I-3).
(1976=100)
1982 64 83 91 71 92 132 83
1983 67 62 68 70 68 167 77
1984 62 60 58 120 102 194 88
1985 57 57 62 77 103 137 72
1986 54 58 83 71 121 152 73
1987 48 75 114 70 126 115 71
1988 57 82 111 64 133 102 75
1989 55 76 109 76 136 110 75
1990 62 77 101 84 157 141 84
1991 64 73 90 84 214 133 84
1992 62 65 71 78 154 100 74
1993 65 68 101 72 164 88 75
1994 66 66 72 79 139 125 79
1995 69 58 66 75 153 69 72
89
90
Table 3.5 Persons engaged in 10+ manufacturing, by branch (1965–90)
91
Source: Tables 3.4 and 3.5.
92 The Industrial Experience of Tanzania
1993a, 1994a) and the data files underlying the Tanzanian published
census. These sources provide information on product quantities and corre-
sponding gross output values, making it possible to derive unit values for
products or groups of products for both economies.
The basic approach is to make matches of comparable products or
product groups from the two censuses and to calculate unit value ratios
(UVRs) for each of the matches. Subsequently these are aggregated into
average unit value ratios for industries, branches and total manufacturing.
These unit value ratios are used as conversion factors.
Matches were made in 16 sample industries. On the US side these consist
of one or more four-digit industries. For Tanzania, the commodity and
output information collected in the 1989 industrial census has not been
published. The information from the basic census data files on products
and their output value was rearranged into four-digit ISIC industries (1968
version of ISIC; see United Nations, 1968b). These were combined into
sample industries comparable to those on the US side.
The conversion factors are calculated in a number of steps.
1. For each product match, UVRs are calculated (TSh/$). The initial unit
value ratios derive from 1989 Tanzanian unit values and 1987 US unit
values.
2. UVRs are put on a 1989–89 basis, using US 1987–89 price movements for
each product group from the Bureau of Labour Statistics (1998). The
1989/1989 UVRs are used in all the subsequent calculations.
3. All the UVRs are aggregated into average UVRs at sample industry level
using output quantities of either countries as weights:
S S
∑Q X
ij ∗ PijX ∑ (Q U
ij ∗ PijX )
UVR XU ( X )
j = i= 1
S
UVR XU (U )
j = i= 1
S
(3.2)
∑
i= 1
Q ijX ∗ PijU ∑
i= 1
(Q Uij ∗ PijU )
where
UVRjXU(X) is the unit value ratio of the Tanzanian shilling against the
US dollar in sample industry j, at quantity weights of
Tanzania
UVRjXU(U) is the unit value ratio of the Tanzanian shilling against the
US dollar in sample industry j, at quantity weights of the USA
i = 1…s is the sample of matched items.
sectors. Where there is more than one sample industry in a branch (as in
food products), the 1989 sample industry UVRs are aggregated at manu-
facturing branch level by taking the weighted average of sample industry
UVRs using 1989 sample industry gross output as weights:
∑ [GO
j= 1
U (U )
j ∗ UVR XU
j ]
(U )
UVR XU (U )
k = O
(3.3)
∑ GO
j= 1
U (U )
j
∑ GO
j= 1
X (X )
j
UVR XU ( X )
=
∑ [GO ]
k O
X (X )
j / UVR XU
j
(X )
j= 1
where
GOjU(U)is gross output in US sample industry j in dollars
GOjX(X)is gross output in Tanzanian sample industry j in Tanzanian
shillings
k is the branch of industry
j=1…o is sample industries belonging to a branch k
5. The branch UVRs are aggregated into UVRs for total manufacturing,
using branch gross output weights according to Equation 3.2. However,
if the reliability of branch UVRs is too low, we use the sample industry
gross output weights of step 4 to weight sample industry UVRs. 11 The
general rationale behind these stepwise weighting procedures is to
ensure that unit value ratios in large sample industries and large
branches receive heavier weights than in small ones (see van Ark, 1993).
However, where the unit value ratios are insufficiently reliable, we use
gross output of matched items in a sample industry as weights at branch
level and sample industry gross output weights at manufacturing level
(see Timmer, 1996, for more detail). Thus, the more reliable a UVR the
heavier the weight it receives in the aggregation procedure.
6. At each level of aggregation – sample industry, branch and total manu-
facturing – the UVRs can be used to convert 1989 value added into the
currency of the other country for purposes of real value added compar-
isons. In binary comparisons one gets two UVRs at every level of aggre-
gation, one at quantity weights of country X, the other at quantity
Measuring Manufacturing Performance 95
gross output, see Prins and Szirmai, 1998, table C-2). These estimates, espe-
cially for value added, are substantially higher than the official ones.
Gross value added in the US census is measured without deducting the
cost of services purchased from outside the manufacturing sector. Thus the
US census concept of value added involves a degree of double counting.
To ensure comparability with the USA, the Tanzanian value added concept
has been adjusted to the US census concept of value added, which is gross
of service inputs from outside the manufacturing sector. This was done by
adding bank charges, insurance costs, transport costs, communication ser-
vices, accountancy and professional services (see Bureau of Statistics,
1993a, codes 408–411). The basic data for 1989 are reproduced in
Appendix Table A.1.
Table 3.7 Unit value ratios and price levels by major manufacturing branch
Tanzania/USA (TSh/$), 1989
Notes: a No sample industries for this branch. We used the UVR for the total of branches. b The
UVR for total manufacturing is the gross output weighted average of branch or sample industry
UVRs (see Timmer, 1996). c Implicit UVRs calculated from the summed branch value added
totals in Table 3.8. These are the preferred UVRs.
the USA will tend to be expensive and rare in a low-income country such as
Tanzania. Products which are cheap and common in Tanzania will tend to
be rare in the USA. Therefore matches with high unit values will have high
quantity weights in the USA and low quantity weights in Tanzania.
Matches with low unit values will have high quantity weights in Tanzania
and lower weights in the USA.
6 New insights
At Tanzanian At US prices
prices (in TSh) (in US$) Geometric
average,
Tanzania USA Tanzania/ Tanzania USA Tanzania/ Tanzania/
USA (%) USA (%) USA (%)
Food manufacturing 203 370 8 387 897 2.4 4 511 82 572 5.5 3.6
Beverages 816 907 10 413 211 7.8 10 027 150 548 6.7 7.2
Tobacco products 712 352 15 691 221 4.5 13 871 305 552 4.5 4.5
Textile mill products 205 282 4 229 796 4.9 1 871 40 281 4.6 4.7
Wearing apparel 48 669 1 306 552 3.7 1 734 30 798 5.6 4.6
Leather products & footwear 120 675 1 305 577 9.2 6 111 36 648 16.7 12.4
Wood products, furniture & 150 251 3 828 677 3.9 1 754 42 331 4.1 4.0
fixtures
Paper products, printing 321 831 11 517 612 2.8 2 318 72 141 3.2 3.0
& publishing
Chemical products 845 705 66 647 074 1.3 7 711 141 811 5.4 2.6
Rubber & plastic products 726 943 17 537 943 4.1 2 234 52 669 4.2 4.2
Non-metallic mineral products 390 347 5 703 193 6.8 2 737 63 071 4.3 5.4
Basic & fabricated metal products 470 994 10 680 022 4.4 6 072 60 344 10.1 6.7
Machinery & transport equipment 419 871 12 819 273 3.3 5 413 72 431 7.5 4.9
Electrical machinery & equipment 778 159 11 234 739 6.9 10 032 63 478 15.8 10.5
Other manufacturing industries 267 095 12 307 000 2.2 3 443 69 537 5.0 3.3
Total manufacturing 312 562 13 844 486 2.3 4 278 69 787 6.1 3.7
Source: Gross value added and employment from Appendix Tables A.1 and A.2; UVRs from Table 3.7.
99
100 The Industrial Experience of Tanzania
Figure 3.3 Unadjusted and adjusted nominal value added for 10+ manufacturing,
1965–90
3 000 000
Unadjusted
Adjusted
2 500 000
2 000 000
1 500 000
1 000 000
500 000
0
1972
1973
1974
1975
1976
1977
1978
1965
1966
1967
1968
1969
1970
1971
40 000 000
Unadjusted
35 000 000 Adjusted
30 000 000
25 000 000
20 000 000
15 000 000
10 000 000
5 000 000
0
1981
1984
1986
1978
1983
1985
1988
1989
1990
1979
1989
1982
1987
eighties. The light industries food, beverages, tobacco, textiles and leather
together account for 56 per cent of value added, as against 54 per cent in
1978 and 63 per cent in 1966.
Figure 3.4 National accounts, QSIP and the new index of real value added, 1965–94
140
120
100
Index 1985 = 100
80
RMVA series BoS 1995b
RMVA series QSIP
60 New IIP
40
20
1987
1989
1991
1993
1995
1961
1963
1965
1967
1969
1971
1973
1975
1977
1979
1981
1983
1985
45 000
NA&QSIP
Million constant 1989 TSh
New IIP
40 000
35 000
30 000
25 000
20 000
15 000
10 000
5 000
0
1961
1963
1965
1967
1969
1971
1973
1975
1977
1979
1981
1983
1985
1987
1989
1991
1993
1995
Sources: Adjusted and unadjusted nominal value added for 1989 from Table 3.2. Adjusted value
added extrapolated with new index of industrial production from Figure 3.4, unadjusted value
added extrapolated from 1985 to 1995 with QSIP index from Figure 3.4 and from 1961 till 1985
with the national accounts index (Bureau of Statistics, 1995b) from Figure 3.4.
with the new index of production. The bottom line represents unadjusted
1989 value added extrapolated with an unadjusted index of industrial pro-
duction. This unadjusted index consists of the national accounts index up
to 1985 (Bureau of Statistics, 1995; see Prins and Szirmai, 1998, table L-3),
linked with the QSIP index available since 1985. Figure 3.5 has a double
message. On the one hand, the rise and collapse of manufacturing are more
dramatic in the new estimates than in the old ones. Real value added in
1995 is about the same as in 1972. But, on the other hand, the adjusted
level of manufacturing value added in 1995 turns out to be substantially
higher than in the old estimates. An important outcome of this research
consists of indices of real output for different manufacturing branches for
the period 1965-85. No such series were previously available. In Table 3.4
above we have presented the new indices for six manufacturing branches
for the period 1965–95.
The growth trends for the branches food, beverages and tobacco, and tex-
tiles and leather more or less correspond to the pattern for total manufac-
turing, represented in Figure 3.4. The trends for the branch wood are
irregular and show long-term decline. Chemicals, petroleum, rubber and
plastics show modest growth until 1977 and stability thereafter.
104 The Industrial Experience of Tanzania
Figure 3.6 Labour productivity index for total manufacturing, 1965–90 (1976 = 100)
120
110
100
Index, 1976 = 100
90
80
70
60
50
40
1965 1967 1969 1971 1973 1975 1977 1979 1981 1983 1985 1987 1989
12
11
10
9
8
Tanzania/USA (%)
7
6
5
4
3
2
1
0
1965 1967 1969 1971 1973 1975 1977 1979 1981 1983 1985 1987 1989
Year
Source: 1989 Tanzania/USA benchmark from Table 3.8; index of GDP per person in Tanzania
from Figure 3.6; GDP per person in USA from national accounts sources. 1965–82 GDP at con-
stant prices from US Department of Commerce (1986); 1977–90 from Survey of Current Business,
various issues; persons employed, 1965–90, from NIPA; 1959–88, US Department of Commerce
(1992), and Survey of Current Business, various issues.
Number of Gross value Gross value Gross value Gross value Employment Gross value
establishments of output at added at added at factor added in (person) added per
factor cost factor cost cost, US census branch as % person
(mill. TSh) (mill. TSh) concepta of total employed
(mill. TSh) TSh
1 Food manufacturing (311/12) 146 24 531.6 5 630.9 7 177 18.24 35291 203 370
2 Beverages (313) 19 8 959.6 3 003.1 4 116 10.46 5039 816 907
3 Tobacco products (314) 3 6 237.2 3 359.7 3 528 8.97 4952 712 352
4 Textile mill products (321) 83 17 701.0 5 256.4 6 367 16.18 31014 205 282
5 Wearing apparel (322) 54 519.5 105.8 136 0.35 2803 48 669
6 Leather products & 21 1 883.2 401.1 520 1.32 4311 120 675
footwear (323/324)
7 Wood products, furniture & 214 4 399.7 1 115.4 1 454 3.69 9674 150 251
fixtures (331/2)
8 Paper products, printing & 62 7 411.8 1 865.7 2 335 5.94 7256 321 831
publishing (341/2)
9 Chemicals products (351–3) 51 11 680.3 3 515.8 4 201 10.68 4967 845 705
10 Rubber and plastic 15 4 900.7 954.0 1 243 3.16 1710 726 943
products (355/6)
11 Non-metallic mineral 22 5 608.1 1 413.7 1 979 5.03 5069 390 347
products (361–9)
12 Basic & fabricated metal 94 13 583.2 2 231.6 2 882 7.32 6118 470 994
products (371–81)
13 Machinery & transport 72 8 062.1 1 768.3 2 234 5.68 5320 419 871
equipment (382/4)
Table A.1 (continued)
Number of Gross value Gross value Gross value Gross value Employment Gross value
establishments of output at added at added at factor added in (person) added per
factor cost factor cost cost, US census branch as % person employed
(mill. TSh) (mill. TSh) concepta of total TSh
(mill. TSh)
14 Electrical machinery & 6 2 303.4 710.0 830 2.11 1067 778 159
equipment (383)
Sources: Prins and Szirmai (1998): gross output and value added table A-4, employment table A-5. Original source: data files of 1989 census of
production.
Note: a US census value added defined as: gross value of output at factor cost minus intermediate inputs, except intermediate service inputs from outside
the industrial sector.
108 The Industrial Experience of Tanzania
Table A.2 Basic data on output and employment in manufacturing, USA, 1989 (establishments with 10 or more persons
engaged)a
Note: Coverage refers to matched output as percentage of total gross value of output. The measure for reliability is calculated as the variation of unit
value ratios divided by the UVR for the sample industry or branch. The 90 per cent confidence interval equals the UVR plus or minus a percentage
equal to twice the reliability measure.
111
112 The Industrial Experience of Tanzania
Notes
* Eindhoven Centre for Innovation Studies and Section of Technology and
Development Studies, Faculty of Technology Management, Eindhoven
University of Technology. Menno Prins is currently working with
Ericsson/Sweden, Wessel Schulte with UNIDO/Uganda. Thanks are due to John
Komba and Russell Freeman for supervision and support in Tanzania, and to
Marcel Timmer and Cees Withagen for valuable comments and advice.
1 Unless indicated otherwise, GDP always refers to the factor cost concept.
2 The work on Tanzanian time series is based on fieldwork carried out by M. Prins
at the Tanzanian Bureau of Statistics, from April to October 1996 (M. Prins,
‘Manufacturing Statistics: Reconstructing Tanzanian Manufacturing Value Added
1965-1995’, MSc thesis, Technology and Development Studies, August 1997).
Lex Lemmens collected the original data files for the benchmark study. We
thank him for making these data available to us. We thank Marcel Timmer for
advice and assistance with the calculation of unit value ratios and their
reliability.
3 This means that we can correct for price changes, but not for real output
changes of the non-responding establishments. Inspection of the real index of
industrial production discussed in Section 4 of this article shows that aggregate
real output was more or less stable between 1983 and 1989. Therefore, our
adjustment for price changes does not seriously underestimate aggregate value
added.
4 Note that this information is not available for the period 1978–90 for the full
population of establishments in the census.
5 The assumption here is that value added is roughly proportionate to the number
of establishments. We have to make this assumption as no complete value added
data are available for the sample for all the intervening years between 1978 and
1990.
6 The important assumption here is that the degree of underestimation calculated
from our sample of 50+ establishments is considered to be representative for 10+
manufacturing.
7 Intermediate inputs or intermediate consumption consists of the value of the
goods and services consumed as inputs by a process of production. (United
Nations, 1993, 143). Cost categories such as interests costs, directors’ fees and
donations are not intermediate inputs, but part of value added.
8 The construction of a price index in this way derives from Mr R. Freeman, who
has prepared a similar estimate in draft data files which he has kindly made
available to us. The CPI is used for food manufacturing (food index), beverages
and tobacco (beverages and tobacco index), textiles and apparel (clothing and
footwear index), and furniture (furniture and utensils index) and for all other
branches the total CPI is used. From 1992 onwards the PPI is available at indus-
try level. The CPI and the PPI are spliced in 1992 to construct a series for
1990–95.
9 Commodity gross output values were not available for the period 1965–85. We
have used 1985 unit value data to cross check and incidentally adjust our 1989-
based intra-industry weights.
10 The reliability of a UVR depends primarily on the variation in unit values within
the given category. The smaller the variation of UVRS around the weighted
average, the more reliable the calculated UVR is as a representation of the under-
lying UVR for the category. Reliability is measured as variation of unit value
Measuring Manufacturing Performance 113
ratios divided by the average unit value ratio; see Timmer (1996). We used .10 as
the cut-off value for reliability.
11 If matched output value was taken as the sample industry weight in step 4 of the
aggregation procedure, it is used as the sample industry weight in step 5 as well.
12 This file contained quantity information for approximately 720 enterprises.
Though this is not explicitly indicated, we may safely conclude that this file
refers to 10+ establishments.
13 There is a typical index-number type of discrepancy between the directly calcu-
lated UVRs for total manufacturing and the implicit UVRs calculated from
summed branch values added at US and Tanzanian prices. We choose for the
lowest degree of aggregation and therefore use the implicit UVR for total manu-
facturing.
4
The Role of Technological Factors in
the Early Stages of Industrial Exports:
A Note
Charles Cooper*
1 Introduction
Over the past decade or so, the economic performance of the Southeast
Asian NICs, particularly their performance in the export of manufactured
goods, has exercised a considerable influence on policy thinking in devel-
oping countries. In many countries there is the hope that the remarkable
demand growth generated in export markets might be emulated and lead
to similar achievements to those of the NICs in terms of full employment
accompanied by rising real wages and labour productivity. Admittedly, the
desirability of an NIC pattern of development may be more questionable
now, in the wake of the Southeast Asian financial and economic crisis, but
the attraction of some, if not all, elements of what is seen as NIC export-
promoting policy still has strong influence elsewhere in the world.
This paper is not about the economic and financial crisis. It proceeds
from the assumption that, by and large, the crisis may be seen as a latter-
day phenomenon, afflicting the NICs in the past tumultuous two years, but
not necessarily discrediting all past NIC industrialisation policies. After all,
as many writers have pointed out, different countries have followed differ-
ent economic policies, so that it is misleading to think that some generally
defined group of common NIC policies have been put in question.
Moreover, whatever may have been put in question, the reality is that there
were historically unique successes in economic and especially industrial
growth over a period of 30 years, and it remains relevant to ask how they
were achieved, and especially (for present purposes), what role technologi-
cal change and policies to induce it, played in the three decades of income
growth which have so recently and abruptly come to an end. There are
many new questions to be asked about the NICs at the moment, but the
crisis does not gainsay the continuing relevance of the question which is
asked in this paper.
Particularly not, perhaps, because of the way it is intended to approach
the question here. The first contention of the paper is that the technological
114
Technological Factors in Industrial Exports 115
The Lewis dual economy model and its development in an open economy
form by Fei and Ranis had a profound influence on thinking about econ-
omic development. This part of the paper examines what these models of
economic development had to say about technology. In particular it will
deal with the implications which can be drawn from the Fei and Ranis
form of the model for technology in the early stages of entry into world
markets for manufactures.
The Lewis ‘unlimited supplies’ model deals with the processes of capital
accumulation in a labour surplus economy, leading to the emergence of a
modern sector in the context of a large subsistence-oriented rural sector.
There is a labour surplus in the rural sector in the sense that the migration
of workers to the modern sector will not cause a fall in output. It is
assumed that arrangements in the subsistence sector are such that all
persons working there enjoy access to the average product of labour in the
sector – and this average product of labour is what determines the
minimum real wage in the modern industrial sector. This is one of
the more debatable and debated assumptions of the model, but we will not
enter into that. The level of output in industry is determined by the pre-
vailing modern sector technology and the minimum real wage. Production
is expanded to the point where the marginal product of labour is equal to
the real wage. At this point the surplus value added in production above
the wage bill accrues as profit to the owners of capital. It is this surplus,
properly reinvested, which provides for reinvestment and expansion, and
which therefore drives the economy. Reinvestment of surplus and the accu-
mulation of capital stock will expand the modern sector so that eventually
surplus labour will be fully absorbed.
Lewis addresses the problem of international trade in his original paper,
but it is probably fair to say that his main concerns are with the internal
operation of the dual economy, and the process of accumulation of capital
in the modern sector. Thus, a central matter for Lewis is the effect of move-
ments in the internal terms of trade – between the urban and rural sectors –
on profits and the capital accumulation process in the modern sector. In
this particular sense, and despite the discussion of the international
economy in the second half of the paper, it is probably fair to say that the
Lewis formulation deals essentially with a closed economy. Twenty years
Technological Factors in Industrial Exports 117
later Fei and Ranis consider the implications of the Lewis type of accumula-
tion in an open economy and apply their framework to the (early) develop-
ment of Korea and Taiwan (Fei and Ranis, 1976). More recently Ranis
(1988) has given a useful reformulation of the original ideas of the earlier
paper. The centrepiece in both analyses is the onset of a phase of ‘export
substitution’, starting at the point where traditional exports are replaced by
exports of labour-intensive manufactured goods. 3 This is a key turning
point, because thereafter the absorption of surplus labour is greatly acceler-
ated. So much so, claim Ranis and Fei, that debates on trade-off between
growth and employment, which were characteristic of the seventies,
became largely irrelevant. Once the economy had got into the export sub-
stitution phase it was expected to move rapidly to the next turning point,
called by Ranis and Fei the ‘commercialization point’. At the commercial-
ization point, surplus labour is fully absorbed, the real wage is no longer
‘institutionally’ determined, but becomes equated to the marginal product
of labour in the rural sector.
What implications does this model of entry into the international market
for manufactured exports have for technological change and factor produc-
tivities? The question is simply answered (in the model) and the answer
follows from labour market conditions.
After the process of export substitution has started, and up until the com-
mercialisation point, the idea is that the institutionally determined low real
wage will rule. Once labour is fully absorbed, that is, once the commercial-
ization point is past, the real wage will naturally rise. In the first, ‘pre-
commercialisation’ phase, ‘the existence of relatively constant (and low) …
real wages … should induce labour-intensive technology choices and, more
importantly, labour-using technology change … in the dual economy’
(Ranis, 1988, p. 82). Then, after the commercialization point and full
absorption of surplus labour, ‘increase in real wages … is expected to be
accompanied by a shift towards more capital and skill intensive technology
and output mix’ (Fei and Ranis, 1976). In short, labour productivity and
real wages will remain low and stagnant after the initial shift to what Fei
and Ranis call ‘export substitution’, whilst manufactured exports will rise
rapidly. Thereafter, when surplus labour is absorbed, wages and labour pro-
ductivity will rise more rapidly. As Fei and Ranis tell the story, the key role
of the labour-intensive export phase is to accelerate the rate of absorption
of excess labour in the economy above the closed economy rate of the
Lewis model in its early form. The implication for technology is simple:
labour intensity will prevail until the commercialization point is reached
and the real wage begins to rise.
In broad outlines Fei and Ranis give a fairly convincing framework for
understanding the actual pattern of development in Korea and Taiwan in
the early stages of their post-war development. There was, of course, a
shift away from primary product exports as these economies successfully
118 The Industrial Experience of Tanzania
countries, and there may therefore be more than one explanation for the
various differences. It is not surprising that such a complex set of phenom-
ena cannot easily be reduced to a single simple pattern.
Plainly, despite the basically sensible framework which the Fei–Ranis
model provides, there are some complexities of the real world that it does
not address in an adequate way, and some of the points made above point
in their direction. These complexities indicate that the framework of the
model, though convincing and in part empirically supported, leaves some-
thing out.
Essentially the Fei–Ranis framework says very little about the implications
of technological change for the process of entry into world markets in manu-
factures. In this respect it is in line with most development economics of its
time. In fact, for the sub-Saharan countries seeking to enter international
trade in manufactures there are conditions in the international market which
strongly influence the nature and requirements of the Fei–Ranis export
substitution phase. The next section touches on implications of generic
technological change7 and innovative competition, as well as on the question
of path dependencies resulting from development of labour-intensive export
development. In the modern world economy these considerations are an
essential complement to the preceding analysis.
still open, though it is narrower than before. In these products, there is still
a high degree of conventional price competition, which developing coun-
tries are in a better position to meet.
line of production, the learning processes this sets in train – whether ‘auto-
matic’ in the Arrow tradition (Arrow, 1962) or the result of conscious manage-
rial decision and resource allocation12 – reinforce the inter-industry pattern of
comparative advantages, and, since the same thing is happening in trading
partner countries, it becomes increasingly difficult to change the pattern. This
is an example of ‘path dependence’ – which might briefly be described as a
recognition that ‘history matters’. Learning processes13 will, obviously,
produce many situations of path dependence. From the present point of view,
path dependence is important because it will influence the possibilities of
shifting between the types of growth path (that is, high versus low productiv-
ity growth paths) which differentiate developing countries. There are two
levels at which relevant kinds of path dependency may be set up.
First, the technological learning processes within firms are path depen-
dent. David recognizes this: ‘Because technological learning depends on the
accumulation of actual production experience, short sighted choices about
what to produce, and especially about how to produce it using presently
known methods, also in effect govern what subsequently comes to be
learnt’ ( David, 1975, p. 4).
Dosi (1988) describes the cumulative learning processes which underlie
the accumulation of technological capability in enterprises. 14 There are
three distinctive features of these learning processes. First, they tend to
have important firm-specific features. Although there may be spillovers of
technological know-how between firms, a good deal of the learning process
in a firm differentiates it from its competitors. Secondly, learning processes
create a good deal of ‘tacit’ knowledge – that is knowledge specific to the
application of particular processes inside the firm, and which is neither
codified nor easy to codify. This is the type of technological capability that
can only be acquired by ‘doing’. Thirdly, whilst some knowledge may accu-
mulate ‘spontaneously’ through the experience of production, for the most
part the accumulation of technological capabilities depends on the alloca-
tion of time and effort by the personnel of the firm, and depends on
explicit management decisions.
But though accumulation of technological capabilities takes place in the
first instance within production units (and increasingly in service enter-
prises too), the broader institutional environment within which firms
operate is also important. In recent times this environment has become
called the ‘national system of innovation’, and important attempts have
been made to describe it systematically (Nelson, 1993). The national
system of innovation is the second level at which there are important path
dependencies. It has a number of components other than enterprises.
These differ in form from country to country, but are present in most. In
the first place there is the education system – especially those parts con-
cerned with scientific and technical education. The early creation of a
highly skilled and educated workforce is generally agreed to have been a
128 The Industrial Experience of Tanzania
key element in the success of the first generation of NICs. On its impor-
tance in Korea see Pack and Westphal (1986). Second, there are the various
institutions engaged in scientific and technological research (outside of
enterprises). These normally include the universities, as well as various
national laboratory organisations. Sometimes, especially in developing
countries, a large part of the scientific and technical capability of a
country is ‘tied up’ in these institutions, and a major policy problem is
how to relate this capability to national development objectives.
Sometimes also, as in the United States for example, these institutions
grew out of major national programmes – such as the space programme or
defence programmes. Third, there is a set of important ancillary institu-
tions – survey systems, technical information systems, standards systems,
technology transfer organisations and so on.
In most countries the institutions making up the national system
of innovation play an important part in technological development
within enterprises, whether through creating a supply of skilled persons,
or through facilitating the acquisition of technology from abroad, or
through provision of technological information, or through the support
of university or other research activities on which enterprises can draw.
It is important not only that the institutional structure of the national
system of innovation should be present, but that it should be functionally
related to the requirements of the enterprises which are at the sharp
end of the process of acquisition of technological capability. The long-
term development of these institutions and their organic relations to the
enterprise sector has played a large part in the process of technological
development in many of the high productivity growth developing
countries.
How then are high and low productivity growth economies distinguished
from one another as far as the technological capabilities are concerned? We
can give an impressionistic but probably reasonably accurate response
along the following lines. First, we expect that in the high productivity
growth economies we will find production and service enterprises, espe-
cially in the export sectors, in which there are considerable concentrations
of technically skilled persons, and where – more importantly – there is a
vigorous process of technical learning taking place within firms. Second, we
would expect that there will be close links between production and the rest
of the national system of innovation. In the low productivity growth
economies we would expect to find firms which are solely concerned with
repetitive production tasks, in which there is no concern with learning or
change. Very little research has been done on these differences, but there is
a good deal of impressionistic evidence to support the picture we have
drawn. This hypothetical description will also make clear that the shift
from low productivity to high productivity paths is not as easy as may
appear. It will depend on generating learning processes within firms on the
Technological Factors in Industrial Exports 129
one hand, and on linking the key elements of the national system of inno-
vation to services and production on the other.
4 Points in conclusion
Notes
* The author is Director of the United Nations University Institute for New
Technologies (UNU/INTECH). This paper draws heavily on two discussion
papers in the UNU/INTECH series, which are referred to in the text. The author
benefited from the comments of his colleagues at UNU/INTECH as well as from
others including Prof. Ed Steinmueller of MERIT, Prof. Jorge Katz of ECLAC in
Santiago, Prof. Larry Westphal at Swarthmore College, USA, as well as partici-
pants at the INTECH/ECLAC conference in Marbella, Chile, in 1995, and at the
UNIDO Global Forum on Industrialisation in Delhi. Prof. Eddy Szirmai, of TU
Eindhoven, as Editor of the conference volume in which this paper appears,
gave invaluable comments which are acknowledged at various points in the
text.
1. It seems to me likely that some people will see this observation as an example of
a kind of determinism. If so, they have an ahistorical case to make, and it is their
obligation to do so.
2. Sustained growth of exports is defined as follows. The growth rate of exports is
determined by fitting the logarithmic form of the standard expression for com-
Technological Factors in Industrial Exports 131
pound growth to the time series of manufactured exports (1970–90). The growth
rate is determined from the coefficient of the logarithmic term. If this coefficient
is significant for the whole time series, at the 1 per cent level, growth is defined
as sustained in our term. Some more recent research will result in slight
modifications to these statistical observations, but not sufficient to require
modifications to the points made in the ensuing text.
3. As Szirmai pointed out in a private note, the implicit symmetry between the
terms ‘import substitution’ and ‘export substitution’ is potentially misleading.
Import substitution refers to the competition between imports from abroad and
domestic production for the domestic market. Export substitution refers to the
increasing share of manufacturing in exports, but it is not clear whether there is
a process of substitution. Nevertheless, in this paper, we will continue to use the
accepted terminology.
4. The data on which this conclusion is based are discussed in Charles Cooper,
‘Technology, Manufactured Exports and Competitiveness’, UNU/INTECH
Discussion Paper no. 9513, December 1995.
5. Although for these countries the matter is somewhat different since – as their
export data indicate – they have not really entered the ‘export substitution’
phase.
6. The data available are too weak to support tests of causality.
7. The terminology ‘generic technological change’ is used very frequently in a
certain part of the ‘technology’ literature, often in an unclear way. The dictio-
nary definition of ‘generic’ is ‘pertaining to a class, not specific or special’. So
generic technologies are not sector specific or unique to particular lines of pro-
duction, but are widely relevant to many. They are technologically and not
industrially defined.
8. This is what Barros seems to have in mind when he writes of specialization
having a ‘negative effect on productivity increase’ (Barros, 1993, p. 545), but his
discussion is vague and unconvincing.
9. The analysis is based on the assumption that whilst there may be international
spillovers of technological capability within industries, there are no spillovers
between industries. The lock-in effect would be much less severe if there were
inter-industry spillovers. It might be argued that one of the implications of
generic technological change is that such inter-industry spillovers will be
important.
10. Krugman’s analysis has not been extended to the case of labour surplus
economies operating with a constant institutionally determined wage rate, but
that does not change the validity of the present line of argument.
11. The second point to emerge from Krugman’s analytics is that there is clearly
a way out of the ‘lock-in’ – along lines which he identifies with Japanese
industrial policy, and which is nowadays more commonly associated with the
policies of selective protection followed by Korea. The idea is that governments
may use ‘temporary protection to permanently shift comparative advantage’.
The protection will be directed to goods which are just outside the present
pattern of national comparative advantage, and applied for just so long as is
necessary to raise relative productivities in their production to the point where a
new area of comparative advantage is established. Krugman refers to this as the
policy of a ‘narrow moving band’ of protection (Krugman, 1987, pp. 48–9). It is
an interesting reflection of the notion of ‘technological upgrading’, and has
considerable empirical foundation in the history of industrial policies in some
countries.
132 The Industrial Experience of Tanzania
12. The recognition that learning processes involve important resources has a sub-
stantial history. As far as work on developing countries is concerned, Katz’s work
in Latin America provided the essential empirical basis (Katz, 1987) and was the
point of departure for a substantial literature. Much later the point became
embodied in theories of endogenous technological change.
13. On the kinds of technological learning which are important in firms in develop-
ing countries see Dahlmann et al. (1987).
14. This and other basic material on accumulation of technological capability is
surveyed in Cooper (1993).
15. There is an implication that the transition will be from import substitution
to ‘export substitution’, but it is not clear what mechanisms will make this
happen.
Part II
Innovation, Technological
Capabilities and Choice of
Techniques
This page intentionally left blank
5
Public Choice, Technology and
Industrialization in Tanzania: Some
Paradoxes Resolved
Jeffrey James*
1 Introduction
The public choice approach has already been used in the African context to
explore the political rationality of policies that seem difficult, if not entirely
impossible, to justify on purely economic grounds. A well-known study by
Robert Bates (1981), for example, sought to explain why governments in
Africa tend to adopt agricultural policies that are blatantly harmful to the
interests of most farmers in the region. More recently, rent-seeking behav-
iour was used by Gallagher (1991) to explain variations in growth rates
across a wide range of African countries. What has not been applied to any
of those countries, however, is the area of public choice theory that deals
specifically with the preferences and behaviour of government bureaucrats:
the so-called political economy of bureaucracy. Yet, as we shall argue
below, this important strand of the public choice literature helps to explain
some of the most paradoxical aspects of technology and industrialization
in the public sector of one particular African country, Tanzania.
More specifically, our argument will be that these paradoxes can be
explained by the following propositions. The first is that, contrary to what
is almost always assumed in the literature on development economics,
bureaucrats do not in fact have preferences defined over technologies (that
is, they do not choose technologies in the usual sense of the word). Rather,
their preferences are thought to be defined over projects and particularly
those projects that enable the institutions to which they belong to grow as
rapidly as possible.
To a much greater extent than elsewhere in the Third World, the industrial
sector in Tanzania (as in sub-Saharan Africa more generally) has been dom-
inated by enterprises owned by the state and to a correspondingly greater
135
136 The Industrial Experience of Tanzania
even when we are building factories which serve the whole nation,
we have to consider whether it is necessary for us to use the most
modern machinery which exists in the world. We have to consider
whether some older equipment which demands more labour, but
labour which is less highly skilled, is not better suited to our needs, as
well as being more within our capacity to build and use. (Nyerere,
1968b, pp. 98–9)
There were indeed many cases, he believed, where the needs of society
could better be met by labour-intensive, small-scale technologies than by
large-scale mass production. These cases, furthermore, were closely in
accord with his view that industry should be decentralized to the
maximum possible extent. ‘In so far as there is a choice, we in Tanzania
would infinitely prefer to see many small factories started in different
towns in our country rather than one big factory started in any one of
them’ (Nyerere, 1968a, p. 107).
These early ideas were consistent with and indeed embodied in the key
planning documents that were to form the basis of Tanzania’s industrializa-
tion strategy in the post-Arusha period (see Chapter 1). The second five-
year plan, covering the years 1969 to 1974, for example, not only provided
explicit encouragement of labour-intensive techniques, but also empha-
sized the importance of more decentralization and more effective linkages
between large- and small-scale industries (ILO, 1982).
For the period after 1974, the so-called ‘basic industry strategy’ strongly
reaffirmed the principle of self-reliance, which played so central a role in
the Arusha declaration and in subsequent policy pronouncements by the
top political leadership. This goal was to be achieved, among other ways,
by the encouragement of a local capital goods industry which would lessen
Tanzania’s dependence on imports of foreign technology. Moreover, the
major objectives laid down by the ‘basic industry strategy’, such as employ-
ment creation, equality of income distribution and dispersion of industry,
suggested implicitly or explicitly that where a choice existed, technologies
should generally be relatively low cost, labour intensive, simple and small-
scale (Williams, 1976).
Beginning in the 1970s, however, a long list of scholars in Tanzania and
elsewhere pointed out that the technologies actually being used in the
public sector typically had just the opposite features. That is to say,
they were usually large-scale and inefficient, as well as being capital and
import intensive.2 As such, therefore, these technologies also tended to
have the effect of centralizing, rather than decentralizing, the location of
industry in Tanzania. Far, therefore, from promoting the most important
goals of the state, public enterprises in the industrial sector seemed to be
doing just the reverse. It was not that there were no choices available to the
managers of state-owned enterprises. On the contrary, over a wide range of
138 The Industrial Experience of Tanzania
Rice huller
4–8 tons per day 93.0 2.4 7.94a
Rice roller
24 tons per day 1 752.5 14.3 2.52
60 tons per day 4 410.0 11.9 1.69
120 tons per day 7 927.0 16.7 1.52
of the most efficient in the entire industry, whereas the other performed
poorly on virtually all the usual indicators. In one enterprise, indigenous
technological capabilities were successfully acquired, while in the other
they were not.
Still another case that is difficult to reconcile with the general pattern of
technological behaviour in the public sector is to be found in the farm
implements subsector, for there is one state-owned enterprise in that sector
that has used relatively simple labour-intensive technology (in conjunction
with other factors) to attain a highly competitive position (as measured by
domestic resource costs) (World Bank, 1987a).
This firm is also unusual in the rapidity with which it was able to dispense
with foreign technical expertise – that is, in the rapidity with which it
was able to acquire indigenous technological capabilities of various kinds
(Barker et al., 1986).
The earliest and most detailed attempt to explain the foregoing paradoxes
was made by David Williams (1976) with particular reference to the textile
industry. He showed, among other things, that the pronounced disparity
between the technologies chosen in the industry could not be explained on
the basis of the existing literature on the choice of technology in develop-
ing countries. It is often argued in this literature, for example, that techno-
logical differences between firms reflect differences in the types of products
they manufacture (where differences refer to variations in the characteris-
tics that the various products embody). In particular, it is commonly
argued that labour-intensive technologies tend to produce goods with a
higher proportion of functional or ‘low-income’ characteristics than
capital-intensive techniques. Williams, however, found ‘no grounds for
assuming that any particular type of technology in the observed range was
dictated by product characteristics’ (Willams, 1975, p. 3). Nor was he able
to find much evidence in support of another well-known category of expla-
nations in the choice of technology, namely, those that impute various
types of non-economic preferences to decision makers in developing coun-
tries. Perhaps the best known of these explanations is the ‘engineering-
man’ hypothesis advanced by Louis Wells (1975). His contention, in brief,
is that under conditions of imperfect competition the preference of ‘engi-
neering-man’ for sophisticated automated technology and modern prod-
ucts dominates the concern of ‘economic-man’ to minimize costs of
production. Yet, while there were certainly enough departures from perfect
competition in Tanzania’s textile industry at the end of the 1960s to enable
‘engineering-man’ to hold sway, one would then have expected a uniform
bias in favour of capital-intensive techniques, rather than the coexistence
of techniques with markedly different factor intensities, as is most clearly
Public Choice, Technology and Industrialization 141
illustrated by the comparison between the two textile mills (see Table 5.2
below). As Williams puts it, ‘The engineering-man hypothesis would not
predict that a single investor … would set up both capital-intensive and
labour-intensive plants at the same time’. (Williams, 1975, p. 7). The same
problem, one should note, would apply just as much to the various other
technology-related objective functions that have been proposed in the lit-
erature, such as, for example, the sense of ‘national pride’ which is some-
times said to be evoked by the use of the most modern technologies in
developing countries (Winston, 1979).
the budget were required. But if circumstances were such that the more
acceptable package available consisted of a high-cost capital-intensive
mill, then that would be chosen. (Williams, 1976, p. 165)
Let us now consider how this observation can be applied to the concrete
case of the two textile mills that were established, as noted earlier, at more
or less the same time by the National Development Corporation.
Friendship Mwanza
It is easy enough to show that these goals find frequent expression in the
writings of political scientists who work on sub-Saharan Africa in general
Public Choice, Technology and Industrialization 145
Public sector fixed assets Public sector employment in Share of public sector in Share of public sector in
in manufacturing manufacturing manufacturing employment manufacturing value
(million shillings) (absolute numbers) (per cent) added (per cent)
Sources: (Col. 1) World Bank (1988, p. 4); (Col. 2) Perkins (1980), Clark (1978) and World Bank (1987a); (Cols 3 and 4) Skarstein and Wangwe (1986).
Note: a Refers to 1964.
Public Choice, Technology and Industrialization 147
1966 65 708 43
1967 80 239 73
1971 99 564
1980 295 342 380
From the perspective of a sector engaged in a struggle for its very sur-
vival, therefore, what seemed to matter much more than its efficiency or
profitability were its size and power. And it is from these political points of
view, we suggest, that the central role of new projects and foreign capital
needs to be understood.
To what extent then did the public sector actually grow in Tanzania?
Table 5.3 provides evidence on this question for the manufacturing sector,
while Table 5.4 refers to the public sector more generally (in both cases the
data cover the period from the 1960s to the early 1980s). As one would
expect, the figures in the tables show a marked growth in the public sector
after the Arusha declaration. For there was then a greatly enhanced need
for civil servants and other officials to run the institutions that had just
been nationalized (Mukandala, 1985). What is remarkable, though, is that
the public sector continued to grow very rapidly after the early post-Arusha
period, when no such obvious need for personnel was apparent and at a
time when in fact the ‘political stratum’ of the state was bent on reducing
the sector’s size and power. Between 1971 and 1981/2, for example, public
sector fixed assets in manufacturing grew more than threefold, while the
share of public enterprises in manufacturing value added almost doubled
(see Table 5.3).
These pronounced increases in the absolute and relative size of the public
sector were closely related to the external sector, for in the ‘frantic spate of
subsidiary creation’ referred to above ‘there was a renewed aggressiveness
toward recruiting foreign project partners; the call for self-reliance notwith-
standing. … This in turn led to a faster increase in the importation of
capital rather than intermediate goods and over-installation of new indus-
trial capacity’ (Mukandala, 1988, pp. 128–9). Tanzania’s increasing reliance
on imported capital goods over this period was to a large extent made pos-
sible by growing amounts of foreign aid, which, as Table 5.5 shows,
financed an ever-greater proportion of gross investment and government
expenditure.
The effect of all of this activity on the technological dimensions of
the industrialization process was hardly surprising. On the one hand, the
Public Choice, Technology and Industrialization 149
1960s 30–35
1970 29.9 11.4 40
1975 114.9 25.3 45
1980 493.6 46.6
1982 514.9 54.3 50
Sources: (col. 1 and 2) Skarstein and Wangwe (1986); (col. 3) Wangwe (1992a).
increased rather than decreased in the reform period, as compared with the
seven previous years. In particular, one can see from Table 5.6 that the
state-owned share of the GDP, non-agricultural economic activity and gross
domestic investment actually increased in the later period, while its share
in employment remained unchanged between the two periods.
Since the period covered by Table 5.6, however, the pace of liberalisation
and privatisation has quickened somewhat. Between 1990 and 1995, for
example, 34 of the 170 parastatals in manufacturing were privatised. This,
in turn, will tend to have increased the share of the private sector in invest-
ment and other macroeconomic variables.
6 Conclusions
Notes
* Department of Economics, Tilburg University, PO Box 90153, Tilburg, The
Netherlands. This article was originally published in Public Choice, 89, 3/4
(1996), pp. 375–392, and is reprinted with kind permission of Kluwer Academic
Publishers.
1. For details see James (1995). This book was the first attempt to apply public
choice theory to African industrialization in general.
2. This was already clear from Clark’s study (1978) of the period from 1964 to
1973. He pointed to a ‘failure of the post-Arusha companies to distinguish them-
selves significantly and favourably from the pre-Arusha companies’
152 The Industrial Experience of Tanzania
(Clark, 1978, p. 140), and concluded that ‘There has been as yet no developmen-
tal innovation on the part of parastatals to make themselves more consistent
with the Tanzanian ideology’ (ibid.). Some years later a very detailed study by
Perkins (1980, 1983) of more than 300 firms in ten industrial sectors arrived at a
similar conclusion, namely, that despite the existence of efficient labour-inten-
sive technologies and ‘Despite the rhetoric, Tanzania’s industrialization pro-
gramme has in general promoted the establishment of [publicly owned]
enterprises using large-scale capital-intensive, often technically and invariably
economically inefficient technologies’ (Perkins, 1983, p. 231).
3. This argument is most clearly elaborated in Stewart (1977).
4. These data are taken from de Valk (1992), who does not cite his primary source.
5. See World Bank (1994a). This source does not however reveal the amount over
5 per cent by which the civil service grew over this period.
6
The Form and Role of Innovativeness
in Enhancing Firms’ Productivity:
The Case of Selected Manufacturing
Firms in Tanzania
Haji H. Semboja and Josephat P. Kweka*
1 Introduction
Development experience has shown that the countries that have under-
gone fast economic growth and structural change in the second half of the
twentieth century are those that have based their productivity growth on
rapid absorption of foreign technology and on increasing efficiency in the
use of technology over time, and that have successfully developed and
applied innovation through learning. Generally, the scale of operation,
firm size, organisational structure, characteristics of the market, industrial
policies, infrastructure, resource availability and the supporting institu-
tions are crucial factors determining the nature and level of innovative
activities. At the broader level, human resource development and access to
finance and the legal and regulatory framework contribute to innovative
achievement.
At the firm level, technological development has occurred through inno-
vation which has become not only an important determinant of the level
of productivity but also of firms’ international competitiveness. However,
during the pre-reform era innovation was never a major determinant of the
level of productivity in Tanzanian manufacturing.
Technological development has been limited. What development there
was, has occurred through absorption and adaptation of foreign technol-
ogy rather than indigenous technological change. It is upon this under-
standing that the process of technological development, the nature of
innovations and the factors influencing innovative activities need to be
understood as a basis of formulating policies for stimulating productivity
and competitiveness of the manufacturing industry. This article intends to
make a contribution in this direction.1
We set out to describe the various forms of innovative activities at
firm level and to determine the factors influencing these activities,
154 The Industrial Experience of Tanzania
In this article we shall consider innovation in its broader sense, that is, as
all that is new to the organisation. At firm level, innovation refers to both
the technology of an enterprise and its products (Rothwell, 1977; Souder
and Chakrabarti, 1980).3 Technical and innovative capability is an asset for
adaptation to all sorts of societal changes. Innovation is caused through
interactions between individuals, business units and social-economic
sectors. In the developing world, public socioeconomic supporting institu-
tions can make important contributions to importing, modifying and dif-
fusing new technologies. Innovation contributes to the creation of wealth
and induces growth, structural change and development. The innovative
capacity of a country constitutes its ability to adapt to shocks, be they
internal or external (DeBresson, 1994). This capacity may break the con-
straints of the internal and external environment, and enable a country to
forge ahead, thus improving the market position of firms and challenging
competitors and the external environment, making them adapt.
According to DeBresson (1994), innovative activity is an aspect of com-
petition, growth and development. Competition by technology is an
important force for a firm and an industry in the context of both industri-
ally developed and developing economies. With the opening of markets,
for an organisation to survive, it is imperative to undertake innovative
activities. Import substitution strategies or industrial protection introduced
to enable an industry to learn and catch up often have the perverse effect
of inducing inertia. Some forms of learning must exist in an organisation if
it is to survive. Even the so-called traditional manufacturing industries are
no longer exempt from the impact of new techniques.
Innovation has also been identified as an important determinant of
success in small-scale enterprises. It is through product rather than techno-
logical innovation that small-scale enterprises are able to develop their
competitive edge. Romano (1990) found that, for high-growth firms, the
determinants of product innovation that influenced the success of small
business included technology, R&D, the product life cycle, market change,
product/market mix and customer base. For low-growth firms, customer
base was a major determinant of product innovation.
Innovativeness and Productivity 155
programmes for their employees while others do not. About 61 per cent of
the surveyed firms in the ESRF Multi-Country Comparative Study on
Private Enterprise Development survey (Semboja et al., 1997b) have indi-
cated that they have in-house training programmes, whereas 39 per cent do
not have them. The nature of such training is mainly on the job training,
and its main objectives are to enhance/upgrade workers’ skills, equip them
with new production techniques, and teach them better quality control
and other machine operations. This is especially important for new
recruits, and for improving productivity and the quality of goods and ser-
vices. Many firms send their employees to local institutions, especially
technical colleges or vocational training colleges, such as the Vocational
Educational Training Authority (ESRF-VETA, 1998).
Prior experience in business has been suggested as important in provid-
ing owners with knowledge of current operations. Training obtained in a
successful innovative firm will make management responsible for innova-
tion more progressive and open-minded (Rothwell, 1977). Without the
support of previous managerial experience for innovation, it will be
difficult to develop a successful innovation programme. In Tanzania,
however, in most firms long experience of the workers or managers did not
translate into significant innovation. This is because managers are preoccu-
pied with routine tasks and responsibilities, where their work experience
has made them capable of solving the day-to-day technical/managerial
problems. However, the experience and exposure of the managers of high-
growth firms (for example Tanzania Breweries Ltd) are shown to be a
dynamic source of innovation, high productivity and competitiveness
(ESRF-PSRC, 1996).
The Tanzanian manufacturing sector is characterized by a critical short-
age of technical skills necessary to enhance firms’ innovativeness. As a
result, some of the high-growth firms have resorted to hiring foreign
experts. Besides foreign technical experts, some firms engage foreign con-
sultants. The Semboja et al. (1997b) survey data showed that 22 per cent of
the firms engage foreign consultants while 78 per cent do not. Those that
engage foreign consultants seek expert advice in the technical and manage-
rial fields.
lack of finance to carry out scientific activities, and failures in the commer-
cialization of R&D initiatives. The basic infrastructure for R&D is underde-
veloped, and links between productivity/innovation centres and the users
are weak (see also Chapter 7). The government role in this aspect is not
very effective, given the modest expenditure allocations for such institu-
tions/centres. In addition, firms have a tendency to invest in short-term
profit-intensive plans, and shy away from long-term investment in innova-
tiveness and R&D activities.
Further analysis of the R&D efforts reveals a critical shortage of research
and technological personnel, not only in the technological centres but also
in the firms themselves. The impact of this shortage is aggravated by three
other factors. First, there is an unfavourable ratio between the number of
engineers and the number of technicians who are supposed to support and
complement the work of those engineers. Most of the engineers have been
offered administrative tasks in many technology institutions and have not
been preoccupied with innovation activities. Second, there is a shortage of
research inputs and facilities with which R&D activities can be done. Most
of the funds allocated to research are used to pay staff salaries (which are
themselves low), leaving little for basic inputs and facilities for research.
Third, there is a weak link between innovative activities and production.
One explanation of the weak link between R&D and production has
centred around the problems of relevance of the R&D activities for prob-
lems facing production activities in the country (Mlawa and Sheya, 1990).
The other side of the coin is the fragility of the productive sector and its
limited financial capacity to engage in new investments, reinforced by the
unfavourable economic environment in which enterprises have been
operating.
In general, however, there are but few industrial firms that engaged
in R&D activities. The RPED (1994) survey categorized the number of
firms engaged in R&D according to the size of firms: 3 per cent of small-
scale firms, 12 per cent of medium-scale firms and 16 per cent of large-
scale firms engaged in R&D. Most scientists, engineers and technicians are
found in the large firms (food industry). Of professionals employed by
the firms, engineers are the most frequently engaged in R&D activities
(RPED, 1994).
between them and the final consumers (that is, the firms) are weak and
rudimentary.
Notes
* Economic and Social Research Foundation (ESRF), Dar es Salaam.
1. There are no comprehensive studies on innovation in Tanzania. The few existing
studies are based on micro-firm-level data and/or are founded on a sound theo-
retical framework (for example, innovative activity matrix). Recent industrial
case- and macro-based studies show that both private firms and public innova-
tion supporting institutions operate within an unfavourable economic environ-
ment (Wangwe, 1986; Watanabe, 1987; RPED, 1994). However, there are
significant adaptive and incremental innovative activities conducted in some
medium- and large-sized manufacturing firms. Direct foreign investment, expa-
triates and imported technology remain the major sources of acquisition of tech-
nology and industrial performance.
2. The article draws some insights from theoretical and empirical works on this
subject and utilises data and information on various manufacturing sector
surveys done by the ESRF and other institutions. The surveys of the manufactur-
ing sector include the following:
• the ESRF-PSRC (1996) study of the impact of divestiture on industrial perfor-
mance (six case studies on Tanzania Breweries Ltd, ABB Tanelec, Morogoro
Textile Mills, Moshi Tanneries Ltd, Tanzanian Parkers Limited, G&T
Company)
• the UNU-INTECH QEH-ESRF (1996) study (also referred to as Semboja and
Kweka, forthcoming) on the impact of import liberalisation on the manufac-
turing sector, based on a survey of 47 light engineering firms and 17 garment
firms
• the ESRF Multi-Country Comparative Study on Private Enterprise
Development, also referred to as the Semboja et al. (1997b). This study pre-
sents results of a survey of 74 firms from three sub-sectors, food, construc-
tion, and metal and light engineering, held in 1997.
Innovativeness and Productivity 169
other institutions of higher learning undertake R&D activities, besides their prin-
cipal teaching function.
13. Innovative efforts may be rewarded with promotion, recognition and productiv-
ity bonuses. For example, rewards for innovative efforts are similar to and some-
times substitute those for productivity, that is, normally in the form of salary
increments, with promotion coming only occasionally. Two companies in the
garment sample report using an incentive bonus system to achieve higher levels
of productivity. The management sets an output target for every section/depart-
ment and for the entire firm. Any production above the target level is rewarded
by a bonus system on a percentage basis. The second bonus system rewards the
department that achieves the best results at the end of each year. A third is the
general bonus, which is awarded to all the workers depending on the general
performance of the firm.
7
Development and Diffusion of
Technology: The Case of TIRDO
Bartelt Bongenaar and Adam Szirmai*
1 Introduction
171
172 The Industrial Experience of Tanzania
first national science and technology policy for Tanzania was formulated in
1985. The STP policy was derived from the general long-term Basic
Industrial Strategy. The shift towards a more open and market-oriented
economy since 1986 and the passing of the national investment act in
1990 necessitated the formulation of a new science and technology policy.
Although a draft version of such a plan has been drawn up, a final version
still awaits publication at the time of writing this article.
R&D in Tanzania is mainly conducted within government organizations.
R&D in firms is almost non-existent. The government of Tanzania is also the
most important provider of funds to the R&D sector. Government spending
on R&D is modest, varying between 0.99 and 1.34 per cent of total govern-
ment expenditures (Mlawa and Sheya, 1990, table 3.4).
Some institutes succeed in getting some form of external support from
international donors or generate funds by offering consultancy, training
and technical services to enterprises. Usually, the income thus generated is
modest (Mlawa and Sheya, 1990, pp. 14–17).2
Eight institutes are primarily involved in industrial research, develop-
ment and design. The organizations have similar or partly similar goals.
Four of them have goals that overlap with those of TIRDO: the Tanzania
Automotive Technology Centre (TACT), the Tanzanian Engineering and
the Mechanical Design Organization (TEMDO), the Centre for Agricultural
Mechanization and Rural Technology (CAMERTEC) and the Institute for
Production Innovation (IPI).
3 Theoretical background
assumption, based on the charter of TIRDO, is that the target group of the
research and development organization consists of domestic industrial
enterprises.
Figure 7.1 presents the six main phases in a model of the technology
development process, seen from the perspective of an R&D institute:
identification and selection of technologies, acquisition of technology,
adaptation, selection of firms, technology transfer to firms, and implemen-
tation of innovations. These steps have to be executed for each technology
development project and will be analyzed separately. The phases are
analytical rather than sequential. There are relationships of circular causa-
tion, from ‘later’ phases to ‘earlier phases’ (feedback) and from earlier to
later phases, and activities in different phases can – and often should – be
undertaken simultaneously.
This article examines the innovation process from the perspective of the
research institution. Thus, we examine the transfer of technology from the
R&D institute to industrial organizations. It should be stressed that R&D is
only one of the aspects of a more complex model of technological change
and innovation developed by Kline and Rosenberg (Malecki, 1991,
pp. 114–17). In that model the interactions between R&D, technology and
markets are very important, as the market ultimately defines the technol-
ogy needs and specifications, and is the customer for R&D outputs.
Technological change only occurs if these different aspects are coordi-
nated. The present analysis concentrates on the functioning of an R&D
organization and tries to fill part of the black box of ‘research’ in the Kline
and Rosenberg model of technological change.
3.3 Adaptation
The purpose of adaptation is to make a technology more appropriate to
local conditions. The adaptation phase is an extremely important phase in
the transfer of technology to developing economies (Dar, 1990, pp. 137ff.;
van Straaten et al. 1992). Many innovations fail because of insufficient
attention to differences in the socioeconomic and physical conditions in
the originating and target environments.
The basic goal of the adaptation process is to compensate for the differ-
ences between environmental conditions before and after the technology
transfer. Therefore, the adaptation process depends on the characteristics of
the environment of origin embodied in the design of the technology and the
characteristics of the target environment for which the technology was
designed. The differences between the circumstances in which the technol-
ogy has functioned and those in which the technology has to function
define the setting and need for adaptations.
The adaptation of a technology not only requires the adaptation of
knowledge to local factor conditions, but also the capability to modify and
add products and processes to suit local preferences and requirements.
Adaptation also involves measures to speed up absorption of a technology,
by making it more appropriate for local use and specifying required
methods for introduction (for example, training) (UNIDO, 1991, p. 177).
In the adaptation phase the participating technical staff is the main
input. But technical staff members cannot operate without non-technical
capabilities within the team. Market knowledge, process control knowledge
and knowledge of the social and economic implications of the technology
also are essential for successful adaptation of technology (Mourik et al.,
1991, pp. 111–112). Therefore, the characteristics of the adaptation team and
its external partners are of considerable importance
phase and the persuasion phase. In the knowledge phase a decision making
unit (DMU) is exposed to the existence of the innovation and gains some
understanding of how it functions. In the persuasion phase the DMU
develops a favourable or unfavourable attitude towards the innovation,
based upon the information available to it.
To influence these processes communication instruments can be used
(Rogers 1983, p. 318; Wissema and Euser, 1988, pp. 75–80). These include
informal communication with opinion leaders in innovation networks, the
use of change agents, the use of financial and non-financial incentives, and
approaching gatekeepers.
The first author of this article spent eight months doing fieldwork at
TIRDO (from November 1995 until June 1996). Two projects – national
dyes and caustic soda – were investigated in great detail. Research methods
included repeated open interviews with project officers and document
research. Another ten projects were examined with help of a standardised
questionnaire completed by the project leaders, again supplemented by
documentary research. (For more detailed discussion of the technical
aspects of the projects, their output and the research findings, see
Bongenaar, 1997.) With the exception of three projects (castor oil, satellite
receivers and turkey oil), all projects were well documented. The 12 projects
are briefly described in Section 4.1. In the subsequent sections, we discuss
the different analytical phases of technology development, making use of
the theories and concepts introduced in Section 3.
• Satellite receiver project: project for the local design of a satellite receiver
for televisions. The project was started as a follow-up to a satellite dish
project. The project started in 1995, and was still going on in 1996.
The goal was to produce simple receivers using the organization’s
equipment.
• Turkey red oil project: a project on the local production of turkey red oil
using locally available castor seeds, used within the textile industry to
dissolve dyes. The project was shelved because of a lack of funding and
lack of interest from the target group. The goal for the project was to
identify and adapt a process for the production of the oil.
National shortages 2
Substitution of imports 9
Availability of raw materials 10
Contacts with industry 3
Follow-up research 3
Pressure to develop proposal 4
Note: n = 12.
Development and Diffusion of Technology 181
Notes: a n = 12.
b
Importance of each flow mechanism measured on scale 1–5.
c
Relative importance based on average scores per mechanism (average score/summed average
scores).
Development and Diffusion of Technology 183
4.4 Adaptation
The purpose of the adaptation phase is to make a technology appropriate
for its target environment. TIRDO’s main aim in this phase is the adaptation
of a technology to locally available raw materials. This has always been the
main goal for projects executed by the chemical department. Aspects of the
technological environment (available knowledge and equipment), the
social environment (special social benefit groups) and the economic en-
vironment (financial implications) are sometimes mentioned as secondary
goals of the adaptation process. The adaptive criteria tend to be related to
physical characteristics of the technology, and reflect the capabilities and
areas of interest of the project team. These interests are chiefly technical in
nature.
Most staff involved in the adaptation process come from within the orga-
nization. Besides the project team other departments are sometimes
involved, especially during the pilot plant stage. The first part of the
research, the laboratory investigation, is normally executed by the project
team itself: 80 per cent of the staff involved are chemical and mechanical
engineers. Within the organization non-technical involvement in the adap-
tive work is very low. In two-thirds of the cases, actors from the environ-
ment (companies and other institutes) are involved in the work. This
contribution is normally limited to the testing of materials. Only R&D
institutes cooperate in the execution of the work itself.
184 The Industrial Experience of Tanzania
Technical Non-technical
Note: a Project results meeting technical and non-technical expectations both scaled from 1–10,
n = 11. In the case of shelved projects, the question whether results met expectations referred to
the partial results. See Appendix A, variable IV, for details of scale construction.
Development and Diffusion of Technology 185
Note: n = 8.
186 The Industrial Experience of Tanzania
the school chalk and the wood adhesives projects, a market survey has
been performed to analyze a target group for the innovation.
Some activities are undertaken to persuade the enterprises to accept an
innovation by reducing entrepreneurs’ uncertainties. The tailor-made
design of production systems based upon the specific needs of the entrepre-
neurs is mentioned in four projects. One method, though not specifically
mentioned, is subsidizing the use of a technology. The R&D organization
does not try to recover the total development costs of a technology, only
charging direct costs and (usually) some fixed fee for development costs.
Besides the project team, the information department is involved in
60 per cent of the projects. In some cases, members of the information
department participate in the work of the project team; in some other
cases their involvement is on request and is limited to selectively approach-
ing some entrepreneurs.
The number of interested firms differs considerably across the projects.
The number of instruments TIRDO uses to interest and persuade also differs
from project to project, and a relation between the instruments used and
number of firms expressing interest seems evident. As indicated, the total
activities for technology diffusion are modest. The organization does not
put enough effort into interesting and motivating entrepreneurs. So far, no
firms have been persuaded to adopt an innovation. One may conclude that
TIRDO’s execution of the selection phase is not very effective.
required management expertise, TIRDO itself certainly did not possess the
knowledge of how to run a plant.
In the case of the caustic soda project there was also no transfer of the
innovation. The staff did indicate what the implications of technology
transfer would be. They described the equipment and knowledge prerequi-
sites for transfer in general terms. (for example, ‘are expected to have
“feeling for chemicals”’). There was interest on the part of firms, a pilot
plant was in operation and one entrepreneur seemed to be willing to adopt
the technology. However, the worsening economic climate and decreasing
prospects of protection of domestic industry prevented the entrepreneur
from following up his interest.
A few remarks are in order concerning two instruments of importance for
the diffusion of technology in the last three phases of technology develop-
ment projects: the use of change agents and use of contacts within industrial
networks. Change agents are normally not used in TIRDO projects. Only in
one case, the school chalk project, could a kind of change agent could be
identified: one member of the project team 10 focused on supporting all
entrepreneurs with the different aspects of the innovation decision.
Contacts with firms and individuals within a well-developed industrial
network can contribute positively to the selection of a technology for
development and the subsequent success in the diffusion of the innova-
tion.11 The involvement of networks in TIRDO’s projects turns out to be
low. Direct influence of domestic industry on project content and on activ-
ities of project teams is lacking. Though there are some contacts with
industrial firms in most of the projects investigated, the scope of these con-
tacts is limited.
Note: a Calculated with respectively n = 12, n = 12, n = 12, n = 11, n = 8, n = 12. The variables are roughly scaled from a negative pole of 1 to a positive
pole of 10.
Development and Diffusion of Technology 189
5 Conclusions
Appendix
the scores on the following three items: age of technology (1: young – 5 old),
familiarity of technology (1: hardly studied – 5: widely studied) and accessibility
(1: mainly private – 5 mainly public).
• unpacking capabilities: the capabilities of the project team to unpack a technol-
ogy reflecting technical capabilities and non-technical capabilities (scale 0–5, after
rescaling of scores to a 0–5 range).
Notes
* Section of Technology and Development Studies and Eindhoven Centre for
Innovation Studies, Faculty of Technology Management, Eindhoven University
of Technology. Bartelt Bongenaar is now working at IBM, The Netherlands. We
thank Emilia van Egmond, Leon Oerlemans and Henny Romijn for valuable
comments.
1. This article is based on the MSc thesis of B. Bongenaar, ‘Analyzing Technology
Development’, Eindhoven, Technology and Development Studies, March 1997.
This thesis was based on eight months of fieldwork at TIRDO. We thank TIRDO
and its former director Dr G. Njau for the opportunity to execute this research
project. We should like to note the impressive degree of openness on the part of
the staff of TIRDO to outside examination of their projects. The aim of this
paper is not to offer facile criticisms, but to contribute to better understanding of
the factors which may contribute to or hamper the success of technology
development.
Development and Diffusion of Technology 193
2. Income from services in the R&D sector is estimated at 15 per cent of total
income, external assistance at 30 per cent of total income.
3. For a more extended discussion of the relevant theoretical framework see
Bongenaar and Szirmai (1999).
4. Invention is the output and the process by which a new idea is discovered or
created. An innovation is an idea or object perceived as new by a group (Rogers,
1983).
5. There is much refined discussion of concepts and definitions in the literature,
which we will try to avoid in this article. Technique generally refers to the total
of means and procedures for the production and marketing of an existing good.
Technology is the wider frame of reference (knowledge, ways and means) within
which techniques can be applied to a set of objectives (van Egmond, 1993,
pp. 15–16). However, the difference between the two concepts is a matter of
degree. In this paper we will use the term technology in a rather rough sense,
referring both to specific techniques and the knowledge required to make these
specific techniques work. Where the wider concept of technology is referred to,
this will be clear from the context.
6. TIRDO projects do not involve the acquisition of domestic technologies.
7. No costs being charged for acquisition, or knowledge being provided without
profit.
8. Both pilot projects had problems with the designs and the implementation of
the designs.
9. For example, in the case of caustic soda, TIRDO expected the producers of soap
to use the technology in their production process; wood adhesives were
expected to be used by the producers of particle board.
10. A Dutch student temporarily involved in the project.
11. Better contacts with firms and clients will also have positive effects in the acqui-
sition and adaptation phases.
12. This problem is not limited to TIRDO. It is also mentioned in the literature on
the Tanzanian R&D sector in general (Sinda, 1991; Kisimbo, 1994; Mageni,
1994).
8
Technological Capabilities: A Core
Element for National Development
Opportunities?
A Study of Technological Capabilities in
the Dwelling Construction Sector for
Lower-Income Households in Tanzania
Emilia van Egmond-de Wilde de Ligny*
1 Introduction
Only two decades ago the search for factors determining national economic
competitiveness led to the introduction of the concept ‘technological capa-
bility’. Like technology, technological capability is a complex concept com-
prising both the utilisation and development of technologies, either
through indigenous efforts or through international technology transfers.
The search for useful definitions and operational indicators has yielded an
extensive body of literature.
First, the theoretical views underlying the definition of the concept of
technological capabilities are reviewed. Then the preliminary results of inves-
tigations regarding the technological capabilities in the dwelling construc-
tion sector for lower-income households in urban Tanzania are presented.
194
Technological Capabilities 195
process in which at least four elements play a role: (1) technology itself, or
‘technoware’, (2) people and their skills and knowledge, or ‘humanware’,
(3) information of all sorts, or ‘infoware’, and (4) the ways and means by
which production is organised, or ‘orgaware’ (ESCAP, 1989).
3.5 Conclusion
There is consensus with respect to the fact that improved production per-
formance and further social development can be attributed to the status of
technological capabilities and technology utilisation in production
processes. In spite of this, there is no consensus yet on an all-comprising
and workable definition for the various concepts. Neither is there consen-
sus with respect to a successful way to make them operational.
Technological Capabilities 199
All this implies that technological capability building means the increase
of the quantity and quality of these components. The creation and
200 The Industrial Experience of Tanzania
1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993
Average annual GDP growth (%) –2.4 3.4 2.6 3.3 5.1 4.2 4.0 4.8 3.9 3.5 3.7
Share of constr. in GDP (%) 2.4 2.8 2.5 2.8 4.0 4.3 3.3 5.4 4.4 4.6 4.0
Growth in constr. GDP (%) –41.0 20.2 –8.9 17.3 49.2 11.9 19.2 69.8 14.1 9.4 –10.3
Contribution of constr. to GDP growth (%) –0.1 0.6 –0.2 0.5 2.0 0.5 –0.6 0.4 –0.4
GFCF
Buildings 1 847 2 510 3 096 4 339 6 057 6 802 11 618 20 423 22 101
Works 1 687 2 203 2 762 4 665 13 020 28 089 20 492 48 039 56 871
Equipment 4 218 7 260 11 014 19 675 45 998 62 410 96 912 149 942 183 406
1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993
Growth in total GFCF (%) –33.2 45.7 22.6 3.0 29.4 2.3 5.1 6.5 32.0 11.0 5.4
Growth in GFCF buildings (%) –32.7 17.2 3.2 10.1 13.1 19.4 21.4 29.5 49.7 3.33 16.6
Growth in GFCF other works (%) –17.0 14.1 6.5 27.6 98.3 1.3 –25.3 136.2 32.1 9.1 12.3
Share of buildings in GFCF (%) 19.9 16.0 12.7 14.4 9.7 11.3 9.3 12.9 14.5 16.9 13.7
Share of other works in GFCF (%) 18.3 14.3 10.9 14.4 22.0 21.8 17.2 43.4. 22.2 27.2 30.0
Source: Bureau of Statistics. Selected Statistical Series 1951–1991. 1994e, p. 35; Statistical Abstract: 1993. 1995, p. 31. The data in the table are based on
constant 1976 prices.
Table 8.4 The formal and informal construction industry in Tanzania
No of peoples employed (× 1000) 160 000 (TIS 1991) 91 649 (LFS 1991) 251 649
22 113 (WB 1994) 182 113
% of total 64 36 100
Annual gross output 14 577 million TSh 1991 14 416 mill TSh 1991 28 993
88 554 mill TSh 1994
Annual value added 10 864 million TSh 1991 4 324 mill TSh 1991 15 188
26 420 mill TSh 1994
Sources: Statistical Abstract 1991, Labor Force Survey 1990/1991, National Informal Sector Survey 1991 (Planning Commission, MLYD, 1991), Survey of
Construction, Trade & Transport, Dar es Salaam1994.
205
206 The Industrial Experience of Tanzania
Equipment item Class I Class II Class III Class IV Class V Class VI Class VII
Tower crane 1 1 – – – – –
Concrete mixers 3 2 1 1 1 – –
Block making mixers 2 1 1 1 – – –
Steel bending machines (set) 1 set 1 set 1 set – – – –
Water pumps 2 1 1 – – – –
Concrete dumpers 2 2 – – – – –
Heavy duty motor vehicles 3 2 1 – – – –
Compactors 2 1 – – – – –
Compressors 1 1 – – – – –
Concrete vibrators 2 1 1 – – – –
Total number of contractors 45 20 56 70 157 146 650
registered (1990)
207
208 The Industrial Experience of Tanzania
7 Conclusions
1. quantity of output investment in the housing sector and its backward linkages for the alleviation
of bottlenecks in the construction process
2. supply of facilities in the house investment in physical infrastructure: access roads, water, electricity and sewerage
3. application of appr. materials and investment in establishment of local production of building materials with
constr.systems high local content
4. application of proper phys.-techn. investment in training and education, information and documentation
constr. details systems
5. simplification of production processes investment in r&d of construction technologies and systems; investment in
on site proper diffusion of available technologies
6. cost reduction investment in r&d for the determination of solutions for improvements and
lower costs of construction technologies and systems
1. access to proper equipment and tools investment in the establishment of local production of equipment and tools
with high local content
2. knowledge and skills of project managers, investment in the establishment of proper training and education systems
general foremen, craftsmen, labourers adjusted to the needs with regard to:
project management, supervision and technical control, crafts, training
possibilities on the job and literacy
3. avail. of and access to proper info. and investment in:
doc. regarding technical specific. and doc., (a) establishment of well functioning and accessible consultancy and advisory
planning and control systems, material organizations
and equip. data bases (b) establishment, accessibility and diffusion of planning and control systems
213
(c) establishment, accessibility and diffusion of materials & equip. databases
Table 8.6 (continued)
214
Process technologies Requirements for TC building
quantity of required labour force investment in education and training of an additional 4000–5000 peoples
professionals (10% of total additionally required)
supervisors & technicians (40% of total additionally required)
craftsmen and labourers (50% of total additionally/required)
quality of required labour force investment in:
1. improvement of existing educational system
2. (on-going) education and training of the existing manpower in
(a) enterprise management, (b) project management, (c) supervision and
technical control, (d) crafts, (e) specific training on the job and (f) literacy
land delivery system investment in surveying and delivery system of dwelling construction plots
energy supply investment in the establishment of an adequate energy supply system
mineral resources investment in adequate r&d and production of local building materials with
high local content of inputs
agro-based and forestry resources investment in adequate r&d and production of local building
materials with high local content of inputs
Table 8.6 (continued)
clients investment in establishment and accessibility of info. and doc. systems for
clients and house owners
consultants investment in:
(a) training and education
(b) establishment and accessibility of financing
materials & equipment suppliers investment in:
(a) mining, quarrying, transport and infrastructure
(a) r&d on product and process technologies
(c) training and education of manpower
(d) availability of equipment and tools
r&d institutions investment in:
(a) training and education of r&d manpower
(b) facilities and equipment
(c) info. and doc. system for a proper diffusion of the results
information and documentation centres investment in establishment and accessibility of info. and doc. facilities
for the construction industry including all actors of its network.
educational institutions investment in the establishment and accessibility of education and training
programmes adjusted to needs.
financing organisations investment in establishment accessibility of financing systems for the
construction industry
branch organisations investment in improvement and accessibility of branch organisations
labour organisations investment in establishment of incentive systems
215
216
Table 8.6 (continued)
(a) improved insight in the actual needs for an optimal performance of the
construction industry through further research
(b) establishment of an enabling environment for the improvement of the
performance of the construction industry by judicious application of policies
and strategies in terms of legal, financial or fiscal measurements. These
refer, for example, to the establishment of adapted regulations, standards
and norms, financing and saving systems, tax regulations on imports and
investments
Technological Capabilities 217
Notes
* Section of Technology and Development Studies, Faculty of Technology
Management, Eindhoven University of Technology.
1. The field studies were executed from 1993 to 1996 by Dankers, Rijkenberg,
Tegelaars and Treffers, MSc alumni of the Eindhoven University of Technology
and supervised by van Egmond and Gaillard, senior lecturers in Technology and
Development Studies, Eindhoven University of Technology.
2. The data for this component of the technological capabilities are derived from
Statistical Abstracts (National Bureau of Statistics, formerly Bureau of Statistics,
various issues), UNESCO (1995) and World Bank (1995c).
9
Technical Education, Knowledge and
Skills in the Metalworking Industry
in Tanzania
Raymond Duijsens and Paul Lapperre*
218
Technical Education, Knowledge and Skills 219
get some form of primary education. In Asia and Africa, however, enrol-
ment for girls is much lower than for boys, and is likely to hamper the
development process in the decades to come. Despite the growth in
primary education, illiteracy continues to rise in terms of absolute
numbers. A large gap still exists between industrialized and developing
countries in enrolment rates beyond primary level. With respect to sec-
ondary education, most developing countries were not even expected to
reach the 1960 industrial country levels by the year 2000. In the field of
tertiary education, the situation is even more a reason for concern. In par-
ticular, enrolment in technical education stays behind, both at the sec-
ondary and tertiary levels.
In most developing countries, education is provided publicly (by the gov-
ernment), and the determinants of demand (private costs versus expected
private benefits: the demand is a derived demand) are generally more
important than the determinants of supply (determined largely by political
processes unrelated to economic criteria other than limitations to public
expenditure) (Lapperre, 1993). UNESCO forecasts for the next century
suggest a continuation of these trends.
Tanzania
Netherlands
1 000 000
Tanzania
500 000
0
10–14
20–24
30–34
40–44
50–54
60–64
70–74
0–4
1 500 000
1 000 000
0
50–54
70–74
20–24
30–34
10–14
0–4
40–44
60–64
5 000 000
4 000 000
3 000 000
2 000 000 Tanzania
1 000 000
0
20–24
30–34
40–44
50–54
60–64
70–74
10–14
0–4
Technical Education, Knowledge and Skills 221
1 500 000
1 000 000
Netherlands
500 000
0
0–4
10–14
20–24
30–34
40–44
50–54
60–64
70–74
absolute number of potential pupils/students (age bracket 5–24) almost
tripled. In the Netherlands the percentage of the population under
25 years, between 1960 and 1990, decreased significantly and the
absolute number of potential pupils/students (age bracket 5–24) remained
virtually constant.
In view of Tanzania’s financial constraints, bringing the educational
infrastructure (buildings, teaching materials, teachers) in line with the
demographic developments will raise serious problems indeed.
120 140
100 120
100
80
80
60
60
40 40
20 20
0
0
1970
1975
1980
1985
1990
1995
Tanzania
Tanzania
Netherlands
Netherlands
Technical Education, Knowledge and Skills 223
Figure 9.3c Gross enrolment in tertiary Figure 9.4 Technical students in sec-
education (% age group) ondary education (% total secondary)
45
40
35
30
25
20
15
10
5
0
1980 1982 1984 1986 1988 1990 1992
Tanzania
Netherlands
industrialized ones by some 10 per cent. Figure 9.4 presents the technical
students in secondary education in Tanzania and the Netherlands as a
percentage of the total of students in secondary education (Centraal Bureau
voor de Statistiek, 1995; Duijsens, 1995; Mitchell, 1995).
Compared to the Netherlands, Tanzania has a low gross enrolment in
primary education. The net enrolment is even more revealing: between 1980
and 1995, in the Netherlands, net enrolment was approximately
95 per cent. In Tanzania, between 1980 and 1995, it dropped from 70 to
50 per cent. Compared to the Netherlands, gross enrolment in secondary
and tertiary education in Tanzania is extremely low. In the Netherlands,
from 1980 onwards, there is an upward trend. In Tanzania enrolment was
almost constant from 1980 onwards. Between 1960 and 1995, the percent-
age of technical students in secondary education (as a percentage of total
secondary education) in the Netherlands dropped from almost 22 per cent to
a mere 11 per cent. In Tanzania the percentage of technical students in
secondary education remained almost constant between 1 and 2 per cent.
2.1 Introduction
Metalworking enterprises play a small but significant role in the Tanzanian
industry (Bureau of Statistics, 1994d). A state-of-the-art survey of 168 enter-
prises in the Dar es Salaam region was conducted in 1990 through the
Development of Improved Production Systems project (DIPS) (Masuha et
al., 1992). The main goal was to make an inventory of available technolo-
gies in the metalworking industry, and little attention was paid to knowl-
edge and skills of employees. Not only did the researchers of the University
of Dar es Salaam want to rectify this omission, also in the literature increas-
ingly doubts were voiced with respect to the quality of (vocational) techni-
cal education (Psacharopoulos, 1991).
At the request of the University of Dar es Salaam, the Eindhoven
University of Technology agreed in 1994 to a research effort to investigate
the nature and causes of shortcomings in knowledge and skills of employ-
ees with formal technical (vocational) education in the metalworking
industry. Only employees who had recently completed their education
were to be included in the research. The preparation of the research took
place in the second half of 1994, and the fieldwork was executed from
January to September 1995.
2.2 Methodology
2.2.1 Identification of the nature of shortcomings in knowledge and skills
In order to identify the nature of shortcomings in knowledge and skills,
these were defined in such a way that they could be judged by means of a
Technical Education, Knowledge and Skills 225
2.3 Results
2.3.1 Nature of shortcomings in knowledge and skills
Table 9.1 presents the outcomes of the survey in 28 metalworking enter-
prises. From the table one may conclude that, for all educational levels,
reproductive skills are consistently less developed than productive skills,
Table 9.1 Nature of shortcomings in knowledge and skills
Educational Skills (scores, 1–4 range)a Knowledge and skills based on single
level of employee based on closed questions open question
VTC 2.5 3.1 2.8 3.4 2.6 3.3 3.0 Low understanding of English; low knowledge
of basic metal working processes; low knowledge of
modern metalworking processes; low attitude of
striving for excellence; low awareness of safety-related
aspects
TSS 2.4 3.0 2.6 3.2 2.6 3.2 3.0 Low understanding of English; low knowledge of
modern metalworking processes; low attitude of
striving for excellence; low awareness of safety related
aspects; low awareness of preventive maintenance
TC 2.6 3.2 2.8 3.2 2.5 3.3 3.0 Low understanding of English; low knowledge of
modern metalworking processes; low attitude of striving
for excellence; low awareness of safety-related aspects;
low management and supervision capabilities
Note: aAll skills range from 1 (not capable of executing the specific type of skill) to 4 (very capable); Cogn = cognitive skills, Inter = interactive skills,
Rep = reproductive skills, Psym = psychomotoric skills, R = reactive skills and P = productive skills.
227
228 The Industrial Experience of Tanzania
and that cognitive reproductive and productive skills tend to be less devel-
oped than psychomotoric, interactive and reactive reproductive and pro-
ductive skills (for reactive skills, the productive variant was not measured).
In addition, and again for all educational levels, the understanding of
English, the knowledge of modern metalworking technologies, the attitude
to strive for excellence and the awareness of safety-related aspects in the
working environment were found to be low. School leavers from VTCs,
TSSs and TCs appeared to have respectively a limited knowledge of basic
metalworking processes, a limited awareness of preventive maintenance,
and a lack of management and supervision capabilities.
Students. Many of the indicated shortcomings find their origin in the edu-
cational background of the students.
Tanzanian education, especially primary education, is known for its high
teacher/pupil ratios (1:40 and higher) (Ministry of Education and Culture,
1988; Dar es Salaam Technical College, 1995; Vocational Education and
Training Authority, 1995). This forces teachers to use the most basic expos-
itive methods and ‘tight rule’ over a class. This, in turn, makes students
passive and less receptive to other methods in subsequent education.
Furthermore, and although the selection of new students is, among others,
based on their mastery of English, their knowledge of English is
insufficient.
This situation is even worse at VTCs, where the student population in all
classes is a mixture of graduate O-level students and primary school leavers.
Teachers. At the TSS and particularly the VTC level, the mastery of English
of the teachers is, in general, poor. In many cases, their educational back-
ground is hardly higher than that of their students, and family and com-
munity characteristics apply to them in a same way. Furthermore, at
teacher training colleges emphasis is put on (technical) contents, rather
than on the mastery of the medium of instruction.
In addition, the teacher training colleges encounter the same instruc-
tional and institutional difficulties as the VTCs and TSSs. As far as TCs are
Table 9.2 Relation between shortcomings in knowledge and skills and institutional and instructional characteristicsa
Reproductive skills √ √ √ x x x x x
Cognitive skills √ √ x x x x x
Understanding of English √ √ √ x x x x x
Knowledge of modern √ √ √ x x x x
metalworking technologies
Striving for excellence √ √ √ x x x
Awareness of safety-related √ √ √ x x x x
aspects
Theoretical knowledge (basic √ x x x x x
metalworking processes)
Awareness of preventive √ x x
maintenance
Management and supervision √ x x x x
Notes: a √ shortcoming related to and investigated at the particular educational level; x relationship (either direct or indirect)
b
VTC, vocational training centre; TSS, technical secondary school; TC, technical college.
c
t, teachers; s, students; f,c, family, community; sm, school management.
d
f, facilities; c, costs; o,c, organization, contents; m, method.
229
230 The Industrial Experience of Tanzania
concerned, there is no teacher training course: all teachers either hold a BSc
degree or are graduates from a technical college (advanced diploma in en-
gineering). Most of them, however, have ample working experience in
industry, and often combine teaching with working.
Their experience, therefore, compensates for the lack of a formal teacher
training course, and their mastery of English as well as their knowledge of
metalworking technologies appear to be sufficiently high.
Facilities and costs. The educational facilities to teach the various kinds of
knowledge and skills (books and machines) are desperately lacking in most
institutes, and this is a direct result of a lack of financial means. At the VTC
level, private institutes, depending fully on foreign donor organisations,
have more financial means available per student than their public counter-
parts, which rely exclusively on government support. At the TSS level,
donor organisations are hardly involved and here the private institutes are
worse off than the public ones.
Contrary to the VTC level, at the TSS level private institutes in general
hardly receive any donor funding and they can only partially compensate
for this by student fees, since these are limited to a maximum. With respect
to the per capita amount of money, private TSSs are at a disadvantage com-
pared to public ones. At this level, the research included the examination
results of 10 073 students at 8 public institutes, and 1749 examination
results at 4 private institutes. The results of the comparison are presented in
Tables 9.3 and 9.4.
Although the average pass rates are higher for private VTCs than for
public ones, the differences are not significant (–z95%<zt<z95% for all pass/fail
combinations). Average pass rates for all subjects shown are significantly
higher for public TSSs than for private ones (zt<–z95%). In view of the exam-
ination results and the financial constraints the respective institutes face, it
stands to reason to assume a direct and positive relationship between the
amount of money spent per student and his or her examination results.
Theory Practical p q n p q n
Basic mathematics .603 .397 3.855 1.488 2,123 .136 .864 4.788 .372 382 22.777 –22.798
English .796 .231 3.686 .946 2,108 .468 .532 4.403 .516 380 11.685 –16.869
Technical subjects .630 .370 3.746 1.592 5 842 .502 .498 4.175 .788 987 7.485 –6.393
the teaching material), which in turn has its effect on the performance of
students. Changing the medium of instruction from English to Swahili
would seem to be a logical decision. However, there are a number of argu-
ments against such a step. (1) Most (if not all) literature is in English.
Cooperation with foreign countries in production requires a mastery
of English. Introducing Swahili would frustrate this cooperation.
(2) Introducing Swahili would make it necessary to translate almost all lit-
erature. Apart from the fact that this would take considerable time, the
costs would be prohibitive. (3) Although Swahili is a complex and precise
language it unfortunately has an inadequately developed technical vocab-
ulary and developing a suitable technical vocabulary would again take
considerable time and be costly. There are, however, also arguments
against the present use of English as a medium for instruction: (1) teachers
and students have insufficient mastery of the language, and, as a result,
(2) the teaching methodology remains mainly expositive. Considering the
wide use of English in science and technology, and the importance of
these fields with respect to industrial growth and development, and con-
sidering the immense problems to overcome when switching to Swahili,
the Tanzanian government should stimulate the use of English, inside as
well as outside the educational institutes, much more intensively.
The financial means of the educational institutes come from three differ-
ent sources (government, donor organisations and school fees), and
influence the quality of the teachers and the availability and state of
books and machines, of which the latter is directly related to the most
appropriate teaching methodology. Furthermore, the comparison of
examination results of public and private institutes indicated that stu-
dents at institutes which have ample financial means obtain better results
than students at institutes which are financially less well off. At the voca-
tional level, the private institutes have more ample spending possibilities,
while at the secondary level the public institutes have more financial
means than their private counterparts. For the latter these differences are
aggravated by the fact that, with respect to selection of students, primary
school leavers with the highest examination results are selected by the
government to enter public secondary schools, while others (that is, with
lower examination results) are free to enter private institutes. Also, with
respect to the selection of teachers, private schools are at a disadvantage:
teachers at public secondary schools are appointed, trained and allocated
by the government, whereas teachers at private institutes have to apply
on their own initiative – their employer (private school) and/or the teach-
ers themselves having to pay for the bulk of their training. Since also at
the teacher training college persons with the highest results at formal
educational levels are allowed to enter, the private schools, unless they
have ample financial means, are left with the less qualified teachers to
teach the less bright students.
234 The Industrial Experience of Tanzania
Notes
* Section of Technology and Development Studies, Faculty of Technology
Management, Eindhoven University of Technology.
Part III
Environmental and Energy Aspects
of Industrialization
This page intentionally left blank
10
Industry and Environment:
Methodologies for Environmental
Assessment in Data-Poor Situations
Lex Lemmens*, Peter Scheren**, Harro Zanting*, Gregory Njau***
and David van Horen*
1 Introduction
The key environmental issues for Tanzania are: land degradation, lack of
accessible water supply and poor water quality, deterioration of aquatic
systems, loss of wildlife habitats and biodiversity, deforestation and envir-
onmental pollution (Division of Environment, 1995). These environmental
problems have evolved over a long period of time throughout the country.
The contribution of industry to these problems is relatively recent.
Industry can be a cause for all of the problems mentioned, but is most
often associated with environmental pollution. Indeed the impact
of industry in the other problem areas is often through pollution.
The determination of the relative importance of industry as a threat to
the quality of the environment is one of the topics of this paper. The
major problem when monitoring or assessing the environment
in Tanzania is the structural lack of data (DANIDA, 1989; Lemmens
et al., 1996).
It is, therefore, of utmost importance to develop methodologies that
provide the best possible environmental assessments in view of the limited
information available. This article presents the rapid assessment methodol-
ogy as a tool for making an environmental pollution inventory in data-
poor situations. The tool was first developed, evaluated and applied to the
case of Lake Victoria. The results of this research will be presented first.
Then it will be shown that the methodology can also be used to get a first
insight into the environmental situation of the entire country using data
from the national census. To complete the available information on indus-
try and environment, the last sections focus on industrial pollution in the
two major cities, Dar es Salaam and Tanga, and on pollution caused by
some mining activities.
237
238 The Industrial Experience of Tanzania
Table 10.1 Functional variables and pollution intensities for the principle
pollution sources
L − K× D
= exp (10.3)
L0 86.4 ×
0–10 1.00
10–50 0.90
50–100 0.75
100–150 0.60
150–250 0.45
242 The Industrial Experience of Tanzania
mean factor of 0.20 is selected as the most likely value. The performance of
wetlands in removing nutrients from waste loads depends highly on the
average nutrient loading per wetland surface unit. Using studies of several
(North American) wetlands, Richardson and Nichols (1985) present graphi-
cal relations between P and N removal rates as a function of average loading.
These relations have been applied to estimate nutrient filtration by swamps
in the Lake Victoria basin. Selected values are presented in Table 10.3.
2 0.4
10 0.3 0.6
50 0.7 0.75
500 0.9 0.8
Industry and Environment 243
With a surface area of 68 800 km2, Lake Victoria is the largest lake in Africa.
Its adjoining catchment area (Figure 10.1) of 194 000 km2 stretches out
over five countries: Tanzania, Kenya, Uganda, Rwanda and Burundi. Only
the first three border on (and share a part of) the lake.
The relatively shallow lake (average depth is 40 m) serves its riparian
population in the provision of fresh water and fish, as well as a depository
for waste from different sources. The only outflow is the Victoria Nile,
leaving the lake at Jinja in Uganda. Evaporation and precipitation domi-
nate the water balance of Lake Victoria.
The features of the catchment area of Lake Victoria that determine
the functional variables for the waste load assessment are discussed in the
following sections.
Figure 10.1 The catchment area of Lake Victoria (source: Crul, 1993)
Victoria
Nile
Victoria
Nile
Lake
Kagera Victoria
Bukoba Mara
Musoma
RWANDA
Muanza
BURUNDI
TANZANIA
244 The Industrial Experience of Tanzania
3.2 Population
With a population of about 30 million, the Lake Victoria catchment area is
one of the most densely populated parts of Africa. Annual population
Notes: a Sources: Oerlemans (1985); Calamari (1994); Ministry of Planning and National
Development (1994); Kirugara and Nevejan (1996), selected industrial site visits.
b
Sources: Droruga (1990); Ministry of Finance and Economic Planning (1993); Ministry of
Natural Resources (1995); selected site visits.
c
Source: Scheren et al. (1995).
Industry and Environment 245
growth is 2–4 per cent in most parts of the lake basin, but urban popula-
tion growth is over 5–10 per cent per year in most of the larger towns.
Table 10.5 summarizes the population statistics of the lake basin. For each
of the 40 towns larger than 10 000 inhabitants (held as a threshold for
urban characteristics) in the Lake Victoria basin, information on sanitary
conditions, domestic waste-water treatment, wetland drainage and distance
from the lakeshore was gathered.
3.3 Agriculture
There are many densely cultivated areas in the Lake Victoria basin, espe-
cially in Kenya, Rwanda and Burundi. Important staple crops are maize
and bananas. The main cash crops are sugar cane, coffee and tea.
Application of artificial fertilizers is significantly higher in Kenya than in
other parts of the catchment area. Agricultural characteristics are summa-
rized in Table 10.6. Most data in Table 10.6 are extrapolations based on
surveys from the early to mid-1990s. An annual growth rate of 1 per cent
is assumed for the area of cultivated arable land (Ministry of Water
Development, 1992). Wetland penetration factors have been applied for
areas draining through wetlands, requiring considerable judgement of the
investigators.
3.4 Rainfall
Average rainfall over Lake Victoria is 1450 mm yr–1 (±10 per cent), equiva-
lent to approximately 100 km 3 yr–1. Regional rainfall varies between
895 mm yr–1 in Musoma (Tanzania) and 2216 mm yr –1 in Kalangala
(Uganda) (Crul, 1993).
Catchment land area (1 000 ha) Livestock population (1 000 heads) Fertilizer use (t y–1)
Notes: a Sources: Lake Basin Development Authority (1987); Kenya Grain Growers Cooperative Union (1991); Ministry of Water Development (1992);
Kirugara and Nevejan (1996); Ministry of Natural Resources (1996).
b
Source: Ministry of Agriculture (1992, 1995); World Bank (1993e), Bank of Uganda (1995).
c
Source: Scheren et al. (1995).
d
Source: Bullock et al. (1995); United Nations (1995).
Industry and Environment 247
3.5 Wetlands
The largest wetland area in the Lake Victoria basin is in western Uganda.
The entire drainage of the Katonga and Ruizi-Kibali river basins is through
river channels that used to be more rapidly flowing in the past. Now, these
channels are ridden with swamps and small swampy lakes (Beadle and
Lind, 1960). Other important wetlands in Uganda are the small swamps on
the lakeshore, receiving the domestic and industrial waste-water from
Kampala and Jinja (Ogaram and Kalema, 1995). Furthermore, the Kagera
river basin, draining through Rwanda, Burundi, Tanzania and Uganda, is
covered with large areas of wetlands. Important swamps on the Kenyan
side are at the mouths of the rivers Yala and Nyando. For all these wet-
lands, information on size and average nutrient loads where combined to
determine the nutrient penetration factors on which Table 10.3 is based.
TN 0.80
TP 0.09
248 The Industrial Experience of Tanzania
lake, the estimated generated BOD loads were corrected for purification in
treatment plants, rivers and wetlands. Table 10.12 presents the results.
The probability intervals presented in Table 10.8 and 10.9 result from an
error analysis (Lemmens et al., 1998b). They account for the roughness of
the data in presenting, more or less, a best- and worst-case scenario. The
authors have chosen the values in such a way that they represent the most
likely case.
Breweries, sugar-cane factories and soap and oil factories display the
largest amount of industrial BOD load. The most likely BOD loads are pre-
sented in Figure 10.2. The conclusion is that domestic pollution accounts
for most of the BOD load. An interesting note here: 75 per cent of Uganda’s
domestic BOD load originates from its capital, Kampala, while in Kenya
50 per cent originates from Kisumu. Further calibration of the Kenyan
results shows that, if all the present treatment plants could perform opti-
mally and meet the theoretical reduction efficiency, BOD loads could be
brought down to 50 per cent of the presented most likely value.
Table 10.9 Results of the Lake Victoria nutrient loading assessment (t yr–1)
Industrial
Domestic
8000
7000
6000
BOD (ton/year)
5000
4000
3000
2000
1000
0
Kenya Uganda Tanzania
Figure 10.3 Most likely nitrogen loads, by source, and agricultural nitrogen loads,
by country
Burundi 4%
Uganda 13%
Kenya 32%
The rapid assessment methodology presented above can also be used to get
a first impression of the environmental situation on a national scale. The
data necessary can be extracted from national censuses or more specific
surveys when available.
In this study we attempted to analyse the contribution of different indus-
trial sectors to the total environmental pressure on aquatic systems. The
data used are those of the 1989 Industrial Census, referring to establish-
ments with ten or more persons employed. As with the analysis in the Lake
Victoria region, we limit ourselves in this study to organic, nitrogen and
phosphorus pollution.
As a departure for the analysis, we start with the 29 ISIC three-digit
sectors as presented in Table 10.10, together with their total and propor-
tional contribution to the value added in 1989. According to WHO
Table 10.10 Basic data extracted from the Tanzanian industrial census, 1989
ISIC Sector Value added Proportional Value added Value added % of value Value added % of value added
code (000 TSh) contribution of polluting of polluting added omitted industries for taken into account
industries industries because of which WHO in the assessment
(000 TSh) with complete incomplete conversions
data (000 TSh) data are available
(000 TSh)
311–12 Food 58 526 087 12.81 57 995 567 56 222 869 3.06 43 363 464 74.77
313 Beverages 3 213 597 0.70 3 213 597 2 925 230 8.97 1 885 467 58.67
331–32 Wood/wood 32 161 429 7.04 1 760 820 821 185 53.36 97 774 5.55
products/furniture
341–42 Paper/paper 6 531 744 1.43 3 795 996 3 738 471 0.57 3 721 712 98.98
products/printings
publishing
351/354 Chemicals/ 6 003 916 1.31 5 218 226 4 160 039 20.28 1 085 634 20.80
petroleum/petroleum
& coal products
355–6 Rubber/plastic 3 855 234 0.84 2 366 671 2 350 033 0.70 2 234 182 94.4
371–2,381 All metal and 134 708 655 29.47 5 305 363 4 024 467 24.14 4 024 467 75.86
metal products
Source: Bureau of Statistics, data files of the 1989 Census of Industrial Production (10+ establishments).
251
252 The Industrial Experience of Tanzania
311–12 Food 34 041 272 0.44 97 654 796 16.35 5 908 84.90 1 207 100.00
313 Beverages 1 525 445 0.02 845 081 0.14 0 – 0 –
331–2 Wood/wood products/ 17 954 <0.005 17 516 <0.005 1 051 15.10 0 –
furniture
341–2 Paper/paper products/ 7 603 316 996 97.56 446 015 025 74.70 0 – 0 –
printing/publishing
351/354 Chemicals/petroleum/ – – 1 902 827 0.32 0 – 0 –
petroleum & coal products
355–6 Rubber/plastic 64 142 <0.005 693 <0.005 0 – 0 –
371–2 381 All metal and metal products 154 296 491 1.98 50 664 519 8.49 0 – 0 –
Total 7 793 262 300 100.00 597 100 457 100.00 6.959 100.00 1 207 100.00
253
254 The Industrial Experience of Tanzania
255
Grand Total 14 395.71 28 095.48 2 222.5 212.4 3 320.7 478 2.42 0.03 0.004 48.2 0.2
256 The Industrial Experience of Tanzania
BOD0 SS Oil N P
BOD0 SS Oil N P Cr Zn Fe
Grand total 1 239.56 904.73 51.55 7.54 1.579 0.8 0.4 1.9
257
258
Table 10.15 (continued)
Grand total 87 227.1 1 502.7 11 328.6 154 327.3 106 489.2 7.0 43.1 0.2 0.1
Table 10.16 Sample of stationary and mobile air emissions Tanga municipal
B: Mobile sources
Land transport 52.8 69.1 253.0
mercury (mean 2.6 ng ml–1). Soil around the mining areas had an average
mercury concentration of 3.4 ug g–1. This in contrast to the regional back-
ground concentration of 0.05–0.06 ug g –1. The study showed low mean
levels of mercury in fish (8.9 ppb) in Lake Victoria.
According to the WHO environmental health criteria (WHO, 1976),
mercury concentrations in freshwater fish from non-polluted areas are
commonly in the range of 10–100 ppb.
7 Conclusions
printing/publishing sector for organic pollution and the food and furniture
sectors for their contribution to the eutrophication problem.
An interesting outcome of the case study is that an environmental policy
directed at reducing nutrient loading, in order to eradicate eutrophication,
should not concentrate on domestic liquid waste reduction. Rather, policies
should be developed to reduce the leaching of nutrients from agricultural
land (for example, the introduction of improved manure management and
enhanced farm systems). Also, the exact causes for the observed high con-
centrations of nutrients in rainwater should be investigated (for example,
forest and biomass burning, volatilization of N compounds in manure and
wind erosion), and policies should be directed to counteract these.
Although uncertainties remain, the outcome of the rapid assessment
methodology has proven very useful. Also, the introduction of a basic error
analysis, and several verification steps, has increased the method’s credibil-
ity. In order to further reduce the probability interval, several steps in the
methodology should be investigated. In the case of Lake Victoria, research
on waste loads from pit latrines and septic tanks, on run-off and leaching
of nutrients from different types of cultivated and non-cultivated land, and
on nutrient concentrations in rainwater would be the major focal points.
Also, the evaluation of some other potential nutrient sources, such as that
of dry deposition and nitrogen fixation by cyanobacteria, could be intro-
duced. Especially the latter should be considered for the case of Lake
Victoria, where algal blooms of cyanobacteria occur frequently, introducing
a large potential source of nutrient input. Further evaluation would,
however, require more advanced modelling techniques, incorporating also
other nutrient processes such as nitrification and denitrification. However,
such would be outside the scope of rapid assessment studies in general.
Available data for industries in Dar es Salaam and Tanga justify the con-
clusion that urban industry contributes considerably to pollution. Gold
mining is an important and dangerous source of point pollution.
Notes
* Centre for Technology for Sustainable Development, Eindhoven University of
Technology
** UNIDO, Côte d’Ivoire
*** Environmental Association of Tanzania
1. The settling velocity is calculated by applying: VSA = rS*SA/CA. The average values
for the sedimentation rate rS (g m–2 yr–1) and the concentration of nutrient A in
the sediment SA (mg g–1) (for both N and P) were taken from Hecky (1993).
11
Energy Conservation in the
Industrial Sector in Tanzania
Frank van der Vleuten*, Lex Lemmens**, Otto Bos***, Caspar
Samplonius***, Dick Toussaint*** and Michel Yhdego****
1 Introduction
The Energy Policy of Tanzania (MWEM, 1992) clearly specifies the impor-
tant role energy plays in the development process: ‘Energy is a prerequisite
for the proper functioning of nearly all subsectors of the economy. It is an
essential service whose availability and quality can determine the success or
failure of development endeavors’ (par. 2). ‘There cannot be sustainable
development and the satisfaction of basic needs of society without
sufficient and efficient supply and use of energy’ (par. 26). Since Tanzania
does not (yet) have at its disposal many indigenous energy sources, the use
of energy, however, is a heavy burden on the balance of payments and thus
on the national economy as a whole. While the long-term strategy of the
National Energy Policy is the reduction of dependence on external energy
sources and the exploration and rational management and utilisation of
the country’s own resources, the short- and medium-term strategies
include, among others, more efficient use of energy in the transport and
industry sector.
The Tanzanian industrial sector offers ample scope for energy efficiency
improvement, in particular in case studies involving the cement and beer
industries. However, in general opportunities have also been identified to
‘leapfrog’ to a cleaner production path, where industrial development can
be better tuned to the available resources and the natural environment.
This contribution discusses the energy dimension of the industrial sector of
Tanzania and shows where and how energy efficiency can be enhanced.
262
Energy Conservation in Industry 263
energy sources, energy carriers (or fuels) and conversion technologies. The
efficiency of the energy system should be enhanced in such a way that the
same energy service can be supplied with less primary energy source/fuel,
or that more energy services can be supplied with the same quantity of
fuel.
The cement industry, which will be discussed in more detail in Section 5,
illustrates the various approaches towards energy efficiency. Table 11.1 pre-
sents the figures for an average cement plant in Africa (van der Vleuten,
1995). The cement industry is one of the most widespread energy-intensive
industries in Africa. For the production of cement, raw materials are ground
and burned in a kiln to produce clinker. The clinker is ground with
gypsum, and occasionally other additives, to produce cement. The con-
sumption of primary heat to produce one ton of cement not only depends
on the efficiency of the technology and the management of the clinker
manufacturing process, but also on the quantity of clinker used to produce
cement. This illustrates that the energy and the materials used in produc-
tion have strong interactions.
From Table 11.1 it is apparent that:
• significant energy saving can be achieved by better process management
and good housekeeping without significant cost
• energy efficiency of a cement plant not only depends on the energy
efficiency of the kiln, but to a large extent on the materials used in
grinding; this underlines that material efficiency (of energy-intensive
materials) can be very important to attain energy efficiency.
Primary
heat 4200 MJ 3600 MJ 3600 MJ 3000 MJ
Process 4 000 MJ/t cement 3 429 MJ/t cement 2 667 MJ/t cement 2 857 MJ/t cement
efficiency
4 Beer industry
Table 11.2 Specific fuel oil, electricity and water consumption at TBL
1992 1993
Fuel oil
l/hl brewed beer 8.0 8.8
l/hl bottled beer 10.5 9.4
l/hl sold beer 11.0 10.1
Electricity
KWh/hl brewed beer 23.1 27.4
KWh/hl bottled beer 30.4 28.9
KWh/hl sold beer 31.6 31.4
Water
hl/hl brewed beer 34.9 44.8
hl/hl bottled beer 46.1 47.4
hl/hl sold beer 47.5 51.2
Brewery Beer output (hl/year) Specific fuel oil Specific electricity Specific total Specific water
consumption consumption energy consumption consumption
(l/hl beer) (kWh/hl beer) (MJ/hl beer) (hl/hl beer)
267
268 The Industrial Experience of Tanzania
Table 11.4 Cost attribution of direct production costs per crate of beer
(TBL, 1994)
subsidiaries of TBL rather than for the Dar es Salaam branch only.
Table 11.5 shows that a net reduction of energy costs of 30 per cent causes
a gross profit increase of 43 per cent.
It is interesting to compare the specific energy costs of TBL with those of
Western breweries. The average specific energy consumption and costs for
German breweries in 1991 are used as a benchmark (Mayer, 1992). These
breweries had a production between 100 000 and 500 000 hl per year. The
specific energy consumption is based on 1993 figures, but to get a more up-
to-date profile energy prices of January 1994 are used. Figure 11.1a shows
that specific energy costs for TBL exceed those of Western breweries by a
factor of two. Energy prices for the reference breweries, however, are
higher, and when this is accounted for the energy costs of TBL exceed
those of Western breweries by a factor four. See Figure 11.1b.
All this shows that the reduction of energy use by conservation is impor-
tant from a financial point of view.
Cost reduction by Increase of value added (%) Increase of gross profit (%)
energy conservation (%)
10 2 14
20 4 29
30 6 43
Energy Conservation in Industry 269
Figure 11.1 Specific energy costs at TBL in comparison with Western breweries
Western
Western brewery
brewery US$ 3.1/hl
US$ 3.1/hl extra
at TBL extra
+ US$ 3.5/hl at TBL
+ US$ 8.7/hl
a: b:
at equal energy prices at equal energy prices
The brewing section had other problems. The refrigeration system was
not able to keep all sections at the prescribed temperature. The main
reasons were old age and bad maintenance of the refrigeration system. In
particular, the compressors often broke down. Other problems related more
to the performance quality. Filled bottles often contained too much air,
reducing the shelf life of the beer and hence provoking customer com-
plaints. Improvements requiring investments were often not possible
because of a continuous liquidity problem. In addition, as can be seen in
Table 11.2, beer disappeared during the production process.
The conclusion with respect to this case study is that energy is an inher-
ent aspect of the functioning of the entire plant. Energy use is closely
linked to use and handling of water, steam and the product. The waste of
energy in the brewery can, therefore, be largely attributed to the perfor-
mance of the management. The major improvements that took place soon
after South African Brewers took over the management support this conclu-
sion (see also Chapter 15).
5 Cement industry
400
tonnes as ( x1000)
300
200
100
0
66 68 70 72 74 76 78 80 82 84 86 88 90 92
67 69 71 73 75 77 79 81 83 85 87 89 91 93
through the Bank of Tanzania and the aid from the Danish development
organization DANIDA. Performance consequently dropped to 55 per cent
of the rated capacity and further to 26 per cent in 1983. Early in 1983 the
last installed kiln (kiln 3) suffered a major breakdown. A solution for these
problems was found in contracting Cementia International AB (nowadays
Scancem) from Sweden and foreign exchange support from the Swedish
development aid organization SIDA. The capacity utilization slowly
improved to about 80 per cent in the early nineties. The present energy
analysis was made during the second half of 1993.
Table 11.8 The heat balance for the kilns compared to two reference kilns in
kJ/kg clinker
Input
Fuel
From sensible heat 21 20 21 13 14
From combustion 4 396 4 208 4 360 3 150 3 050
Raw meal
From sensible heat 116 116 116 54 92
From sensible heat of water 2 2 2 – –
Combustion air
From sensible heat of primary air 5 5 5 6 8
Output
Heat of formation 1 750 1 750 1 750 1 750 1 789
Evaporation of water from raw meal 12 12 12 13 17
Exhaust gas sensible heat 1 854 1 295 1 353 636 595
Dust sensible heat – 23 23 18
Incomplete combustion – – – –
Clinker sensible heat 62 62 62 63
Cooler exhaust gases – 0 0 423 520
Losses radiation & conv. kiln 862 903 759 297 243
Losses radiation & conv. cooler – 65 62
Losses radiation & conv. Pre-heater – 241 483
Rest 23
by a factor of two higher than in the reference kilns. The losses through
radiation and convection are two to three times higher. The heat remain-
ing in the clinker is comparable for all kilns. The reference kilns are a kiln
with a four-stage pre-heater and a grate cooler with a capacity of 2000–
3000 tpd (tons per day) and a kiln with a four-stage pre-heater, a pre-
calciner and a grate cooler with a capacity of 2800 tpd. The difference in
size of these reference kilns, compared to those of the TPCC, accounts for
part of the difference in radiation and convection losses. However, consid-
erable savings can be expected from reducing heat losses in this part of the
cement production process.
In the course of processing, the factory loses energy. Much more energy
is lost when cement production has to be interrupted. Whenever the kiln
has to be shut down, heat is wasted in the product since this has to be
rejected because it has not gone through the whole process. In addition
heat is wasted in starting the process up again. Production will remain low
until temperatures throughout the whole system are brought up to
normal.
It takes from thirty minutes to several hours to balance the system after
an interruption. Factors that cause kiln interruption can be internal or
external. For the TPCC, which gets its power from TANESCO, the external
factor which causes interruption is power supply. These power interrup-
tions affect not only energy efficiency but also production capacity use. In
1993 the kilns were down for about 600 hours because of power failure or
power reduction, which accounted for the major part of the inefficiency.
Internal factors that cause interruptions are unplanned maintenance and
lack of raw materials. These were not so important, but since they can be
directly affected by the management they provide important energy-saving
opportunities. In Table 11.9 the most important reasons for unplanned
stops (except power cuts) are presented.
Loss of feed 8 7 8
Cyclone blockage 6 8 5
Waste gas fan tripped 3 8 4
Energy Conservation in Industry 275
Total 11 460
Variable costs:
Energy costs 175 920 3 340 550 45 5 030
Other 100 40 155 505 780 1 580
plant of 6 MWh would cost 1500 million Tsh. The anticipated pay-back
period is five years. In view of the high energy prices of TANESCO, an
additional advantage could be a lower kWh price for the self-generated
electricity.
6 Conclusions
Both cases confirm the high potential for energy conservation suggested by
earlier energy audits in Tanzania. Both in energy-intensive industries and
less energy-intensive plants, energy conservation presents an opportunity
to enhance the profits of the industry, to improve the manufacturing
process and to improve the environmental performance.
At the technical level, the main constraints with respect to more energy-
efficient industrial operations include the following:
The previous outcomes of energy audits, as well as the case studies, under-
line the large potential for financially and environmentally profitable
investments in energy efficiency existing in the industrial sector in
Tanzania. This potential can be seen as a resource, the exploitation of
which is in line with the sustainable development of Tanzania.
Notes
* Free Energy Europe
** Centre for Sustainable Development, Einhoven University of Technolog
*** Section of Technology and Development Studies, Faculty of Technology
Management, Eindhoven University of Technology
**** Environmental Resource Consultants, Dar es Salaam
Part IV
Lessons from Past Experiences,
Economic Reform, Prospects for
the Future
This page intentionally left blank
12
Industrialization of Tanzania: Can
Tanzania Learn from European
History?
Paul Lapperre*
283
284 The Industrial Experience of Tanzania
3 Industrialization in Tanzania
Table 12.1 Population, GDP per capita and economic structure for
133 countries, by income category
Tanzania 30 170 84 5 11 58 17 24
4.1 Politics
Achieving internal integration and external adaptation are the two funda-
mental functions of the institutional sphere of politics in any society, func-
tions that can be summarized as ‘how can society, in face of a multitude of
internal contradictions, conflicting interests and threats from outside, be
kept together?’. Internal integration and external adaptation are achieved
through system maintenance, conversion and government functions
(Bertholet and Lapperre, 1991). The system maintenance function refers to
aspects such as socialization, recruitment, mobility and retrenchment of
the members of the political system – or, in other words, to the ways in
which the allocation of positions takes place. The conversion function
refers to aspects such as the articulation and aggregation of interests, and to
communication. The government function refers to aspects such as rule
making, rule application and rule adjudication. A number of aspects of the
functions mentioned have played an important role in the transition
process from agricultural to industrial society.
In traditional European societies it was particularly aspects related to the
allocation of positions and the articulation and aggregation of interests that
hampered the transition towards an industrial society. The medieval
European political system, embodied in the Three Estates (nobility, clergy
and common citizens), was characterized by a relatively low social mobil-
ity. Position allocation almost exclusively took place on the basis of kinship
ties: ‘one’s birth was more important than personal merits. By the end of
the 18th century the political system gradually ceased to obstruct social
mobility, and a new trading, commerce and manufacturing elite gained a
say in the affairs of the state. In addition, be it reluctantly at first, the state
also allowed alliances between various new groups active in the economic
sphere (Jansen, 1991).’
The degree of political development of a society can be judged on the
basis of a number of political system capacities (Lapperre, 1997). The extrac-
tive capacity of the traditional agrarian societies – the extent to which the
political system is capable of mobilizing material and human resources as
inputs – was relatively low. For an industrial take-off, the state has to be
increasingly capable to manipulate taxes, subsidies and markets in favour
of trade, commerce and industry, and to extract the necessary services. This
was exactly what was achieved in a number of European countries,
Lessons from History 287
including Britain, in the early 19th century. The regulating capacity – the
extent to which the thinking and acting of the members of a social system
are governed by power and coercion, or the threat of coercion – in tradi-
tional societies was also low. A lot of coercion was usually necessary to keep
society together. When transition gets on its way, the political system
should be able to enforce its will without unduly alienating economically
important groups and thus rendering them less valuable. For this purpose it
is important that the symbolic capacity of the political system is high. The
latter was achieved by most of European nations during the early stages of
industrialization. As far as the internal and external response capacities – the
measure and way in which the system serves its interests and the measure
in which the political system succeeds to adapt to external pressures and to
maintain its internal integrity – are concerned, all political systems
throughout history have always tried to find the most favourable balance.
At the onset of the industrialization process, the political system should be
able to maintain its territorial integrity and to solve conflicting interests
with the least possible damage to the emerging new economic system. In
the late 18th century these prerequisites were already fulfilled in Britain.
With respect to the distributive capacity – the way in which the system
allocates the extracted goods, services, power, prestige and opportunities to
persons and groups in society – the traditional situation was one in which
the goods and services produced were allocated almost exclusively to the
kinship-based elite. At the take-off of the transition, the political system
has to be able to affect a significant transfer of extracted goods and services
to new economically important individuals and groups. This was exactly
what happened in Britain and Germany.
In traditional agrarian societies, the productive capacity – the measure in
which the system is effective in mobilizing and regulating the economic
sphere for its own political aims – was relatively low. For the take-off of the
industrialization process, it is imperative that the production capacity
increases. For this purpose the political system has to provide, among
others, collective facilities in the fields of transport, communication, trade,
commerce and education infrastructure. It has to do away with obstruc-
tions such as, for example, toll barriers and unfavourable tariffs. It has to
remove monetary and fiscal bottlenecks, and to allow a free movement of
labour. In addition, it is important that it is able and willing to act as entre-
preneur when no other possibilities exist. Already at the onset of the indus-
trial revolution, Britain, Germany and France had taken extensive measures
to improve their productive capacity by the means mentioned. The pre-
requisites for industrialization in the political sphere mentioned so far
were all structural.
The cultural prerequisites, among others, concern new skills and tech-
nologies, a high level of knowledge, individualistic, utilitarian and merito-
cratic ways of thinking and acting, thinking in terms of social classes
288 The Industrial Experience of Tanzania
4.2 Economy
In a situation where four peasants can hardly maintain one person not
involved in agriculture (Grigg, 1982) and shortages of staple foods are a
recurrent phenomenon, there is little money available to buy all but the
most essential manufactured commodities. The overwhelming majority of
the population is bound to the land, and struggles to wrest a living from it
and to satisfy the demands of land owners. For the industrialization process
to take off, a productive agriculture and low staple food prices were among
the most important prerequisites that were regionally fulfilled by the end of
the 18th century and even much earlier (More, 1997).
In the cottage industry that had already developed in the 14th century,
production of, for example, textiles took place jointly with agricultural
work. When making hay or at harvest time, spinning and weaving just
took second priority.
When production eventually started to concentrate in places where
much was invested in machinery, buildings and infrastructure, part-time
labour, as in the cottage industry, no longer sufficed, and manufacturers
required and attained a steady and guaranteed supply of labour of sufficient
quality.
Capital is understood to mean produced means of production – capital
goods – and the power of disposition, expressed in money, over these
goods. In a metaphorical sense, it even means the disposition over people.
When, at the end of the 18th century, the demand for manufactured con-
sumer goods increased, this created problems of supply for the traditional
manufacturing sectors, and brought about a fertile environment for tech-
nological change (Cameron, 1991). To finance innovations for mass pro-
duction, in the long run, additional capital was needed over and above
family capital and profit from one’s own enterprise. New forms of
banking systems and enterprises developed, and were made legitimate by
new laws and government regulations (Rosenberg and Birdzell, 1986;
More, 1997).
Mass production uses immense amounts of raw materials and energy,
and it is more than convenient if raw materials and energy sources are at
hand locally. If they are not, there should be a good transport and trade
infrastructure. Before the end of the 18th century, prices of raw materials,
agricultural produce and handicrafts were strongly affected by the distance
to be covered between producers and consumers.
Lessons from History 289
Prices of grain, for example, tripled with every 150 km of transport over
land (Thirsk, 1982). Manufactured goods, in so far as they were not made
locally, were only available to the elite. Mass production of consumer
goods, on the other hand, relies on a steady demand and this demand, cer-
tainly in the take-off phase of the industrialization process, depended to a
large extent on relatively low prices. To achieve these low prices, a well-
established network of roads, canals, harbours and, later, railroads was
established and maintained.
Goods, services and money can, however, only be made really mobile by
a good financial infrastructure. By 1720 some 300 years of slowly expand-
ing markets had been accompanied by a corresponding expansion of agri-
cultural and craft production. The large-scale cottage industry and
ironworks, which first emerged in the 15th century and expanded in the
17th, were indicative of the growing pressure expanding markets exerted
on traditional production processes. In the cottage industry the entrepre-
neurs were usually merchants, and they supplied the materials to cottages
and workshops, bought and distributed the final products, and, sometimes,
provided the necessary equipment. In the iron industry the entrepreneurs
were often connected to the landed gentry. In both the cotton and iron
industries entrepreneurship had its roots in trade, commerce and agricul-
ture (Pieterson, 1987). Although these traders and gentry were keen on
profit and avidly looked for market opportunities, they were generally not
quite prepared for a new situation in which they had to perform a host of
functions at the same time. Products had to be designed and tested, and
raw materials acquired and tooled. Parts had to made and assembled to
yield a final product.
Large numbers of labourers had to be trained in factory discipline, new
regional, national and international markets had to be found and distribu-
tion networks established. By the end of the 18th century a class of entrepre-
neurs had emerged with the realization that manufacturing was not simply
equivalent to efficiently operating a small workshop or factory, but involved
the creation of changes – in product, distribution and organisation – that
would increase the margins between costs and revenues (Righart, 1991).
The transition from agrarian to industrial, of course, also required a
demand for industrial products. Such a demand, however, could only
develop in an environment in which there was enough purchasing power
and in which consumers were willing and able to change their consump-
tion patterns. When the first industrial revolution took off, new skills and
technologies were required to make the change from cottage industry to
the mass production of consumer and capital goods. Around 1760 the
level of knowledge and skills was already high in many fields. Science,
however, did not yet play a significant role in the industrialization
process. In the beginning this process was predominantly fed by craft
skills and knowledge.
290 The Industrial Experience of Tanzania
Even in the late Middle Ages the world was still conceived as relatively
static. The notion prevailed that God had given every human being its
place in the world already at birth, and that the Three Estates were a God-
given order. Private property and interest were guilt-ridden concepts
(Evans, 1985). Work was considered more a curse than a blessing.
Renaissance, Reformation and Enlightenment eventually contributed to the
emergence of individualistic and meritocratic ways of thinking and acting,
thinking in terms of classes instead of estates, thinking in terms of the right to
own private property, the pursuit of profit and a more positive approach to
work, cultural aspects in the economic sphere which can all be considered
prerequisites for industrialization. Last but not least, in an environment which
became increasingly complex, entrepreneurs had to think in terms of econ-
omic systems. They learned how to view the world as something which can
be empirically researched and manipulated on the basis of research findings.
4.3 Kinship
Both overall population growth and urban population growth in an envir-
onment of declining rural populations were stimuli for developments that
created preconditions for industrialization. Particularly urban population
growth in an environment of declining rural populations, a common phe-
nomenon in parts of 16th and 17th century Europe, brought about innova-
tive behaviour of small farmers and increased their productivity (de Vries,
1984). In view of the importance of the size of the market for manufac-
tured products, countries with a relatively large population and a high and
evenly distributed population density had an advantage over countries
which did not have these features.
At a certain point in the development process, ascriptive attitudes were
replaced by more individualistic and meritocratic ways of thinking and
acting, and the paramount importance of kinship as the integrating
element in society diminished (Hillebrand and Jansen, 1991). Estate
society, with its reliance on ascription, gave way to forms of class society
(Appelbaum and Chambliss, 1995). The production-oriented extended
family gradually became a consumption-oriented nuclear family.
4.4 Religion
The zeal to convert the infidels to Christianity and religious faith in general,
in different ways, had their role in the ‘slow revolution’ that brought about
industrial society in Europe. When, during the crusades, European rulers
tried to rid the Holy Land of the Arabs, this also meant the re-establishment
of trade links with the Middle East – trade which contributed, in turn, to the
revival of the economy in the 13th century (Runciman, 1994). Religious zeal
was one of the motives behind the European expansion in the 15th century,
and during the religious wars, in the 16th and 17th centuries, religion was
instrumental to the coming into being of a variety of independent and
Lessons from History 291
competing nation states with their own social identity (Wright, 1984).
Religion, and particularly Calvinism, it is argued by a number of scholars,
was instrumental in the development of labour ethics which were particu-
larly suited for industrial production (Romein, 1954).
4.5 Education
At the onset of the European transition from a traditional agrarian society
to an industrialized one, education had been fully institutionalized in
almost all European countries. Primary education was available to many
but certainly not all, and literacy rates generally remained low. Practical
training in crafts and early manufacturing usually took place during (long)
apprenticeships and on the job.
There were hundreds of universities in which a great variety of subjects
could be studied. With a few exceptions, however, technology, industry
and science were not yet linked in joint endeavours to increase productiv-
ity. The availability of skilled artisans, trained in apprenticeships and on
the job, in almost all fields of economic activity was probably the only
direct significant prerequisite for the transition to take place.
The identified prerequisites for the early European transition in the institu-
tional spheres of politics, economy, kinship, religion, education and health
care can now be checked against relevant aspects of the same institutional
spheres in contemporary Tanzania. The Tanzanian aspects are derived from
literature and research findings of 14 postgraduate students of the
Department of Technology and Development Studies of the Faculty of
Technology Management, Eindhoven University of Technology, in the
Netherlands. All these students conducted their research in Tanzania
between 1991 and 1997.
5.1 Politics
Table 12.2 presents an overview of identified prerequisites for the early
European (E) transition in the institutional sphere of politics, and checks
these against the state of relevant aspects in contemporary Tanzania (T).
292 The Industrial Experience of Tanzania
Prerequisites E T
Structural
State able and willing to:
• give the new (economic) elite a say in the political system + –
• let the old and new (economic) elite form alliances + –
• manipulate taxes, subsidies and markets in favour of trade,
commerce and industry + –
• enforce its will without unduly alienating economically important
groups + +
• maintain its territorial integrity and solve conflicting interests with
the least possible damage to the emerging new economic system + +
• affect a significant transfer of extracted goods and services to new
economically important individuals and groups + +
• provide collective provisions in the field of transport,
communication, trade and education infrastructure + –
• do away with obstructions such as toll bars and unfavourable
tariffs + +
• remove monetary and fiscal bottlenecks + +
• allow free movement of labor + +
• act, when necessary, as entrepreneur + +
Cultural
State able and willing to promote:
• new skills, technologies and a high level of appropriate know-how + –
• individualistic, utilitarian and meritocratic ways of thinking and
acting + –
• thinking in terms of classes in stead of estates or other rigid
entities + +
• thinking in terms of private property and profit + +
• thinking in terms of economic systems + +
In the institutional sphere of politics, Tanzania does not (fully) fulfil the
structural prerequisites of a system able and willing (1) to give the new
economic elite a say in politics, (2) to let the new economic elite form
strategic alliances, (3) to manipulate taxes, subsidies and markets in favour
of trade, commerce and industry and (4) to provide collective provisions in
the field of transport, communication, trade and education infrastructure.
Certainly before the economic reforms of the 1980s, but also thereafter,
Tanzania’s political system lacked the necessary flexibility to fully incorpor-
ate economically important individuals and groups, notably from the
Asian community, in the various institutes of power. Although the state in
theory advocates the free development of political alliances of (economi-
cally) important groups, in practice the possibilities are limited by the
biased way in which parties are funded and the unequal access to the
media. Both aspects are detrimental to the national economy, since they
Lessons from History 293
push actors into the illegal informal sector from which it is difficult to
extract revenues (Schulte, 1996). Mainly because of its bureaucratic charac-
ter and a widespread culture of bribery – in spite of the efforts to curb the
latter – Tanzania’s political system lacks the strength to fully exploit an
effective system of levying taxes in favour of important economic activities
such as, among others, the provision and maintenance of an adequate
infrastructure in the fields of transport, communication, trade and, to a
lesser extent, education (Newberry and Stern, 1987). Contrary to many
other African countries, Tanzania has been remarkably successful in its
peaceful endeavour to forge a nation state from what in essence is a multi-
ethnic society. It therefore meets the prerequisite of a system able and
willing to enforce its will without alienating important groups. In the past
decades it has also met the prerequisite of a system able and willing to
maintain its territorial integrity with the least possible damage to the eco-
nomic system, and a system able and willing to solve internal conflicting
interests with the least possible damage. An exception was the military
intervention in Uganda in the early 1980s, which caused prolonged grave
harm to the economy.
A cultural prerequisite that is not (fully) fulfilled is a state willing and
able to promote (1) new skills, technologies and a high level of appropriate
know-how, and (2) individualistic, utilitarian and meritocratic ways of
thinking and acting.
Although, in theory, the state promotes new skills, technologies and high
levels of appropriate know-how, in practice the situation is less favourable.
The budget limitations to enhance education and technological develop-
ment are severe, and the actual priorities set are often at odds with formal
policies. In the field of education, for example, vocational secondary tech-
nical education is high on the priority list but receives inadequate funds to
fulfil the demand (Duijsens, 1996). Since the mid-1980s, educational pro-
grammes, among others, promote individualistic, utilitarian and merito-
cratic ways of thinking and acting (see also Section 5.3 on kinship) but the
attitude of the state remains ambivalent, in the sense that the African
Socialism ideologies (‘Ujamaa’) from the 1960s and 1970s are still pervasive
in many government laws and regulations.
5.2 Economy
Table 12.3 presents an overview of identified prerequisites for the early
European (E) transition in the institutional sphere of the economy, and
checks these against the state of relevant aspects in contemporary Tanzania
(T). In the institutional sphere of the economy, Tanzania does not (fully)
fulfil the structural prerequisites of (1) a highly productive agricultural
system, (2) relatively low real prices of staple foods, (3) a continuous and
guaranteed supply of labour, (4) labour with sufficient skills and knowl-
edge, (5) capital over and above family capital, (6) new banking systems,
294 The Industrial Experience of Tanzania
Prerequisites E T
Structural
Highly productive agricultural systems + –
Relatively low prices of staple foods + –
Continuous and guaranteed supply of labour + –
Labour with sufficient skills and knowledge + –
Capital over and above family capital + –
New banking systems + –
New forms of organization of production + +
Locally available raw materials + –
Locally available energy + –
Good trade infrastructure + –
Good network of roads, canals and harbours + –
Entrepreneurs with the conviction that mass production is not
simply equivalent to efficiently operating a small workshop or
factory, but concerns the creation of changes – in product,
distribution and organization – that increase the margins between
costs and revenues + –
Demand for industrial products + +
Sufficient purchasing power + –
Environment in which consumers are able and willing to change
their consumption patterns + +
Cultural
New skills, new technologies and a high level of know-how + –
Individualistic, utilitarian and meritocratic ways of thinking and
acting + –
Thinking in terms of classes instead of estates or other rigid
units + +
Thinking in terms of private property and profit + +
Positive view on work + +
Thinking in terms of economic systems + +
(7) locally available raw materials, (8) locally available energy, (9) a good
trade infrastructure, (10) a good network of roads and harbours, (11) a class
of innovative and cost-conscious entrepreneurs and (12) sufficient purchas-
ing power of the majority of the people.
The absence of a highly productive agricultural system – with the related
relatively high prices for staple foods and their irregular supply – is one of
the most notable bottlenecks in Tanzania for industrialization to proceed
(FAO Yearbooks). This is particularly significant since the European experi-
ence shows that countries which seriously fell behind in the productiveness
of their agriculture in the 18th and 19th centuries – for example Russia and
to a lesser extent France – were also late to industrialize. A highly productive
agricultural system releases labour for industrial employment and guarantees
Lessons from History 295
a regular supply of staple foods at affordable prices. Since less money has to
be spent on food, more money can be spent on manufactured commodities.
Although unemployment in Tanzania is high, the prerequisite of a con-
tinuous supply of labour with sufficient skills and knowledge, essential for
industrial enterprises, is not always met. Most workers have extensive
(extended) family obligations and, forced by the low wage levels, hold
more than one job at a time. Absenteeism, therefore, is high and endangers
the necessary continuity of work (ILO, 1992). Nationwide there is an
insufficient supply of well-educated workers, particularly at the technical
vocational levels (Duijsens, 1996). If the recent acceleration of the innova-
tion rate in almost all sectors of the economy in industrialized countries is
also going to affect developing countries, the availability of well-trained
workers will become ever more significant.
The lack of sufficient infrastructure in the fields of banking, trade, roads,
railways, harbours and energy supply – although recently significant
improvements have become visible – is probably the second most impor-
tant bottleneck for further industrial development. Although part of the
banking system has been modernized and liberalized recently, loans and
bank guarantees at reasonable conditions are hard to get for those without
‘connections’ (Zijl and Lassche, 1997).
Roads, railways and harbours have been significantly improved during
the last decade, but funds for proper maintenance are lacking (Fleisheuer,
1994; Gnoth, 1994; Koenders, 1994; van der Ven, 1996). Hence, these facil-
ities are deteriorating again. Hydropower-generated electricity is available
in all major towns, but the supply suffers from frequent and prolonged
cuts, notably in the dry season (Kers, 1997).
Tanzania has a great variety of important raw materials, ranging from
good-quality coal to different types of ore. Owing to a lack of capital and
expertise, and transport difficulties, these resources have so far not yet been
(fully) exploited.
Tanzania has a small class of innovative and cost-conscious entrepre-
neurs, many of whom originate from the Asian community. The majority
of indigenous African entrepreneurs have their roots in small-scale manu-
facturing and artisanal enterprises. Generally, and for a variety of reasons,
they find it difficult to increase their margins between costs and revenues
by means of innovations in product, organisation and distribution. They
have little access to capital, lack managerial and technical skills, and carry
the burden of extended family ties and communal attitudes connected
with the long period of African Socialism (Zijl and Lassche, 1997).
Last but not least, the per capita income of the overwhelming majority of
the Tanzanian population is so low (US$170 in 1995) that the overall pur-
chasing power is seriously impaired.
Cultural prerequisites that are not (fully) fulfilled are (1) new skills, tech-
nologies and a high level of appropriate know-how and (2) individualistic,
296 The Industrial Experience of Tanzania
5.3 Kinship
Table 12.4 presents an overview of identified prerequisites for the early
European (E) transition in the institutional sphere of kinship and checks
these against the state of relevant aspects in contemporary Tanzania (T). In
the institutional sphere of kinship, Tanzania does not fulfil the prerequi-
sites of (1) relatively high population densities, (2) a population relatively
equally distributed over the territory, (3) an urban population growth in an
environment of a declining rural population, (4) consumption-oriented
nuclear families and (5) individualistic, utilitarian and meritocratic ways of
thinking and acting.
Tanzania’s overall relatively low population densities and the unequal
distribution of this population over the country are likely to impede indus-
trialization. A transport and communication infrastructure – among others
effectively linking market centres for specialized agricultural produce and
manufactured products – is (too) costly to establish and particularly to
maintain. The same applies to electricity grids and the electronic commu-
nication infrastructure. Although modern computer systems are available
in Tanzania, the degree of their interconnectedness is extremely low, and
hence the role of modern information technology in industry is small.
Urban population growth in an environment of a declining rural popula-
tion – bringing about agricultural innovation, increased agricultural pro-
ductivity and an impetus to the development of ‘cottage industries’ –
appears to be a specific European phenomenon, and is of little consequence
for contemporary Tanzania.
Prerequisites E T
Structural
Relatively large populations + +
Relatively high population densities + –
Population relatively equally distributed over the territory + –
Population growth + +
Urban population growth in an environment of declining rural
populations + –
Some kind of class society + +
Consumption-oriented nucleus families + –
Cultural
Individualistic, utilitarian and meritocratic ways of thinking and
acting + –
Lessons from History 297
5.4 Religion
Table 12.5 presents an overview of identified prerequisites for the early
European (E) transition in the institutional sphere of religion, and checks
these against the state of relevant aspects in contemporary Tanzania (T).
The prerequisites in the institutional sphere of religion appear to be specific
for the European transition from the agrarian to industrial society, and do
not apply to contemporary Tanzania. From a historical point of view, they
are probably coincidental.
5.5 Education
The influence of a high level of education and of science on the European
transition has often been overestimated. In the first phase of the transition,
298 The Industrial Experience of Tanzania
Prerequisites E T
Structural
Religious competition (contributing to the establishment of
independent nation states) + –
Cultural
Zeal to convert + –
‘Calvinist’ labour ethics + –
it was the artisans that led the way, and science did not play an important
role as yet. In the institutional sphere of education, the lack of skilled arti-
sans in key areas is, therefore, the only prerequisite not met.
alliances. Even the present multi-party state is still dominated by the old
elite. The state, for a variety of reasons, is not able to exploit the existing
taxation system efficiently and effectively. This reduces the possibilities to
manipulate tax revenues in favour of trade, commerce and industry, and to
provide adequate collective provisions in the field of transport, communi-
cation, trade and education infrastructure. Additional political reform
would appear to be necessary for further economic development.
In the sphere of kinship, Tanzania’s development is hampered by rela-
tively low population densities and an unequal distribution of people over
the territory. This makes the development and maintenance of an adequate
infrastructure in almost all spheres prohibitively expensive. In view of the
low income per capita, consumption-oriented families are few, and the
local market for manufactured goods remains relatively small. Extended
family obligations weigh heavily on individual households, reduce their
saving capacity and require significant investments in social time, all
aspects that do not really enhance further economic development.
Communal ways of thinking and acting – having their roots in the tradi-
tional East African subsistence peasant societies and re-emphasized in the
philosophy of ‘traditional ujamaa living’ in the 1950s and 1960s – still
linger on and, by and large, have not been replaced by more individualis-
tic, utilitarian and meritocratic ways of thinking and acting. Cultural
changes of this type are hard to enforce by political means, and a further
change will take considerable time.
In the sphere of the economy, Tanzania lacks a highly productive agri-
cultural system and staple food prices that reflect the benefits of such a
system. Further economic development is hampered by a large class of sub-
sistence peasants who have great difficulties producing more than their
basic family requirements. Reviewing the situation of this sizeable part of
the population, and taking appropriate measures, would appear to be an
important political priority. In addition, credit facilities are limited, essen-
tial raw materials are not (yet) locally available, energy production and dis-
tribution are inadequate, and a good trade and communication
infrastructure is lacking. Various extended-family characteristics and the
concurrent communal ways of thinking and acting do not facilitate further
economic development, and will be difficult to change in the short term.
Notes
* Section of Technology and Development Studies, Faculty of Technology
Management, Eindhoven University of Technology
13
Macroeconomic Policy and
Performance of the Manufacturing
Sector in Tanzania: Has
Liberalization Helped? An
Econometric Approach
A. V. Y. Mbelle*
1 Introduction
300
Macroeconomic Policy and Manufacturing Performance 301
2.1 Performance
Table 13.1 below presents data on the performance of the sector in the
1990s.
Real growth (%) 4.1 1.9 –4.1 –0.6 –0.2 1.6 4.8 5.7 8.1
Real mfg export 13.3 –27.7 –8.7 –19.0 48.1 41.9 1.4 –5.7 –42.6
growth (%)
Share in GDP 8.8 8.7 8.2 8.2 8.1 7.9 8.0 8.1 8.4
Share in total exports 23.8 19.6 16.2 11.8 14.8 16.0 14.6 14.5 8.8
Average capacity 50 50 38 50 46 46 48 60 55
utilization
Share in NT exportsb 48.9 43.7 38.8 28.4 42.2 36.5 35.3 29.2 18.2
Employment (000) 123 121 126 126 130 131 129 132 137
changes in the management of the exchange rate, though exports ran out
of steam by 1996. The share of manufactured exports in total exports
declined from 23.8 per cent in 1990 to a mere 8.8 per cent in 1998. The
dominance of manufacturing exports in non-traditional exports was greatly
eroded by 1998. The factors accounting for this drastic decline include
trade sanctions against Burundi, unrest in the Great Lakes region, and dis-
ruption of transport network caused by El Niño rains, thus hindering deliv-
ery of industrial exports to neighbouring countries, as well as leading to
reduced inputs for agro-industries.
2.1.4 Employment
Trends in manufacturing employment were on the increase between 1991
and 1998. When compared to the sector’s real growth, the increase in
employment was more rapid than is warranted by performance of the
sector. Further analysis of the employment data indicates most employ-
ment is found in food and beverages (ISIC 31) and textiles and leather (ISIC
32). For example, in 1998, these two categories accounted for 67.1 per cent
of total manufacturing employment: ISIC 31, 37.3 per cent, and ISIC 32,
29.8 per cent.3 For the food and beverages group, this reflects increased
investments, especially after 1995, with an average of 31 per cent of all new
industrial licences issued between 1995 and 1998.4
Variable 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999d
Growth in money
supply (M2) 45.4 26.1 38.5 28.8 32.5 26.2 11.6 11 11.1 7.9
Inflationb 32 34 24 24 27 34 15.4 15.4 15.2 7.8
Forex Premiumc 66.7 43.8 36.0 5.3 1.9 <1 <1 <1 <1 <1
Savings rate 1 yr av. 26 26 26 24 24 15 11 7.3 8.7 10
Lending rated 31 31 31 39 39 40 38 32 24.3 25
GDP growth (real) 5 6 4 4 3 4 4.2 3.3 4.0 n.a.
Notes:
a
As of 30 June, 1999.
b
Headline inflation.
c
Defined as the difference between the parallel exchange rate and official exchange rate
expressed as proportion of the official rate.
d
Maximum, medium and long term.
Sources: Bank of Tanzania, various issues; Mbelle (1998).
Macroeconomic Policy and Manufacturing Performance 305
3.1.2 Discussion
Prudent macroeconomic policies are supposed to provide an overall
environment which is benign for economic agents to take advantage of.
Tanzania has gone a long way towards improving the macroeconomic
environment (for example, liberalizing the exchange rate towards
a market-determined rate). However, the extent of growth of GDP
does not reflect the intensity and extensiveness of the policies being
pursued.
(a) Introduction of value added tax (VAT) on 1 July 1998, replacing sales
tax and stamp duties for firms registered in the VAT system. This will
simplify taxation procedures (depending on co-operation of agents and
good record-keeping).
(b) Depositing duty draw-back funds in a special account instead of in the
general account of the government where the money had been used for
other purposes, thus discouraging exporters.
(c) Further liberal tariff reforms with only four bands: a low tariff rate of
5 per cent to be imposed on raw materials, replacement parts and
investment goods; a 10 per cent tariff for semi-processed inputs; a 20
per cent tariff for fully processed inputs and a top rate of 25 per cent for
consumer goods. A zero rate for strategic and leading investment
sectors continues to apply as well.
(d) Pre-shipment inspection to be introduced for textile materials, in order
to curb illegal imports.
(e) Reduction of duties for COMESA (Common Market for Eastern and
Southern Africa) goods to 20 per cent.6 By June 1999 the rate averaged
7 per cent.
An informed assessment of the benefits is not easy, given the short time
span. However, casual observation of economic performance in 1998 would
point to a positive impact of these measures.
(a) expensive credit, implying that the liberalisation of the banking sector
has not helped; in the survey by the Regional Programme on Enterprise
Development (RPED) 64 per cent of the industries ranked this as the
most severe constraint
(b) ‘expensive’ foreign exchange – accessibility is mainly guaranteed to
those with connections to foreign investors
(c) ‘excessively’ restricted circulation of money in the economy
(d) markets swamped by cheap imports from abroad, which have, for
example, adversely affected the textile industry
(e) increases in general unemployment brought about by retrenchments in
the civil service and in the manufacturing sector (though for this sector
it doesn’t show so much), as privatisation of public enterprises is taking
its course (market forces)
(f) still too many taxes and too high levels of taxation.
3.2.1 Discussion
Looking at Tanzania’s history since independence, one sees differences
in the reliance on macroeconomic policies for improved production
(see the overview in Chapter 1). While in the 1970s sectoral policies
were given equal weight (for example, in agriculture and industry in
the Basic Industry Strategy see Skarstein and Wangwe, 1986; Musonda,
1992; Ndulu and Semboja 1994, 1996; Parker et al., 1995), the 1980s
witnessed an absence of such meso-policies following up on macro-
policies with concrete targets. The SIDP for Tanzania points out this fact.
In the introductory remarks it is acknowledged that one of the reasons
for launching SIDP 1996–2020 is that the Basic Industry Strategy
(1975–95) had expired (in 1995). In fact it had already expired in late
1970s, when the economic crisis unfolded. It is only now that a vision
for the sector, and one for the country as a whole, is being formulated
again.
Against this background, the following section presents an econometric
investigation of the impact of macroeconomic policies on the manufactur-
ing sector in Tanzania.
Macroeconomic Policy and Manufacturing Performance 307
where:
where:
Notes: a The critical values are –3.556, –3.575 respectively for the, DF and ADF. All
variables are in log levels.
Macroeconomic Policy and Manufacturing Performance 311
Table 13.4, we cannot reject the null of cointegration in all three cases. We
then proceed to estimate a static regression involving all the variables and
the residual generated therefrom. One period lag is included in the general
autoregressive distributed lag (ADL) model as the error correction term.
After estimating the general model, 10 we went through a series of
simplification procedures. We proceeded principally by eliminating those
variables with t-values of less than one in absolute terms, the efficiency of
which was assessed at each stage using the F-statistic, the information crite-
rion and the model diagnostic tests, until we arrived at a preferred and a
certainly more parsimonious characterization of the data.
By reducing the model, the F-statistic increased from 3.1555 to 5.4806 in
the general and preferred models respectively. The standard error of the
regression declined from 0.965761 in the general model to 0.903656 in the
preferred model. The information criterion11 also showed a decline, starting
with the SC from 0.903278 to 0.480687, HQ from 0.44081 to 0.183386,
and FPE from 1.39904 to 1.07907 in the general to the preferred model
respectively.
The model diagnostic tests are insignificant at the conventional test
levels starting, with the AR test for higher order autocorrelation. The ARCH
test for autoregressive conditional heteroscedasticity, the Bera–Jaque test
for residual normality, Whiter test for heteroscedasticity and the REST test
for functional form misspecification were carried out. There is nothing to
suggest that our model is misspecified in any way.
The error correction term (Res4_1) has the expected sign and is
significant at the 1 per cent test level. This is a further confirmation that
the variables CMAE, CMAO, TOT, and RERM indeed co-integrate.
In addition, we ran the estimation excluding dummy variables to test for
parameter constancy. The insignificant Chow test value and the forecast
test value all point to parameter constancy over the period 1988 to 1996.
The model can thus be used for forecasting.
The preferred model results are presented in Table 13.5.
Note: R2 = 0.697666; F(8,19) = 5.4806 [0.0011]; s = 0.903656; DW = 2.25; RSS = 15.51529265 for
9 variables and 28 observations.
While reporting these results we will not repeat the procedures outlined in
the estimation of the first model. 12 Only a few tests are given (Tables 13.6
and 13.7). The rejections in Table 13.7 may point to problems with the
Macroeconomic Policy and Manufacturing Performance 313
Table 13.6 Test for unit roots: the manufacturing investment equation
Notes:
a
Critical values at 5% level are: DF –3.567; ADF –3.587; SBDW 1.72.
b
Lags were used in the auxiliary equation for ADF.
testing statistic when the error correction term is also rejected. With this
assumption we then proceeded to include the error correction term in the
general model and re-estimated the equation. If the ECT turns out to be
significant and bearing a negative sign, then our assumption is correct. The
results of the general model confirmed the correctness of our assumption. 13
In Table 13.8 we present the final results.
Note: R2 = 0.534871; F(7,20) = 3.2855 [0.0172]; s = 0.414869; DW = 1.74; RSS = 3.442327352 for
8 variables and 28 observations.
5 Concluding remarks
Appendix
Notes
* Department of Economics, University of Dar es Salaam
1. The shares for manufacturing, mining, construction and utilities being 8.4%,
2.0%, 4.4% and 1.7% respectively (URT, 1999a).
2. ‘Mkakati wa Taifa wa Kusimamia Utekelezaji wa Sera za Kuondoa Umaskini’
(National Strategy for Poverty Alleviation) (URT, 1997).
3. Computed from URT (1999a), table 60, p. 150.
4. Computed from URT (1999a), table 64, p. 154.
5. Introduction of new products, higher expenditure in R&D and product
development.
Macroeconomic Policy and Manufacturing Performance 317
1 Introduction
318
Urban Informal Manufacturing Sector 319
noted in the debate. The advocates of the so-called sectoral approach (among
others, Souza and Tokman, 1976; van Dijk, 1980; Sethuraman, 1981;
Liedholm and Mead, 1987; De Soto, 1989) still demonstrate little consen-
sus, not only with regard to the number of sectors that should be distin-
guished (two or more), but also with regard to the criteria on the basis of
which such distinctions should be made. Neo-Marxist authors (for example,
Lebrun and Gerry, 1975; Davies, 1979; Guerguil, 1988) make a distinction
between different modes of production within one capitalist system, and
try to prove that ‘petty production’ is subordinated to – but also integrated
in – the capitalist mode of production. This approach is not very convinc-
ing in the light of the (scanty) empirical evidence. The third theoretical
approach, the so-called continuum approach (see, for example, Bromley and
Birbeck, 1984; Tezler and Molenaar, 1989), seems more promising. In this
literature, the assumption is made that it makes sense to distinguish two
sectors (formal and informal). However, instead of supposing that the two
sectors operate essentially independently of each other, it seems more prob-
able and realistic according to these authors that the two sectors interact
with each other in a continuous fluctuating manner, and that parts of the
informal sector can be dominated or even created by parts of the formal
sector.
Since the introduction of the formal–informal dichotomy (Hart, 1971)
some progress has certainly been made in empirical research. For example,
empirical research has demonstrated quite convincingly that it is incorrect
to declare the informal sector synonymous with the urban poor. In most
cases, paid workers in the formal sector earn considerably less than entre-
preneurs in the informal sector (see Table A.27). Nevertheless, one may
legitimately conclude that still relatively little is known about the urban
informal sector in developing countries.
The following major and often related research topics are frequently
mentioned in the literature:
• What exactly belongs to the urban informal sector? What are the
specific characteristics of this sector? To what degree do these character-
istics differ from those of the formal sector?
• What is the effect of the informal sector on income distribution and
poverty alleviation? To what degree and how does absorption of labour
take place? What are the advantages of the informal sector for the urban
economy or, put in more general terms, what is the relevance of the
urban informal sector for socioeconomic development?
• What kinds of relationships exist between the formal and the informal
sector? Do the two sectors function in relative isolation from each
other? Do they compete with each other or are they complementary to
each other? Should one speak of exploitation or mutual interest?
• What are the constraints on further growth of the informal sector?
322 The Industrial Experience of Tanzania
4 Methodological problems
Our answers to the above questions can only be approximate and prelimi-
nary, mainly because of the methodological problems facing urban infor-
mal sector research. We will briefly discuss the main methodological
problems with regard to urban informal sector research as observed in most
of the case studies collected (see References). The scientific and the social
value of research results is mainly determined by their degree of comparabil-
ity, generalization and replicability.
The first issue is concerned with conceptual definitions, units of research
and theoretical frameworks. A large majority of case studies typically does
not give any definition at all of the concept of the urban informal sector or
provides only a vague description. Furthermore, different units of research
are used: sometimes individuals, sometimes households, and sometimes
production units. A theoretical framework, indicating what exactly is to be
investigated and how these elements are expected to be interrelated to each
other, is hardly ever given.
The second issue is mainly concerned with sampling. The large majority
of the case studies typically does not give any information at all about the
applied sampling method. Furthermore, samples are usually very small. The
population distribution according to economic activity is very seldom
taken into account, resulting in substantial bias that is mentioned only
very rarely by the researcher.
The third issue is especially related to the operational definitions. Hardly
any of the case studies provide the research instrument used, indicating
how different concepts actually have been measured.
In conclusion, much of the research on the urban informal sector in
developing countries is still of a predominantly pre-scientific nature. As a
result – as stated above – many of the relevant major research questions
can so far only be answered partially and in a preliminary fashion.
Although quite a few case studies have been executed, these studies are
largely of a pre-scientific nature. The lack of generalisation power and the
scant possibilities for comparison and replication seriously hamper theory
formation.
5 Empirical evidence
On the basis of these findings, one may cautiously conclude that the infor-
mal-sector organisation does have specific characteristics, diverging sub-
stantially from those of the formal-sector organisation and that, therefore,
the distinction formal–informal is a meaningful one, as stated by protago-
nists of the sectoral approach (see Section 3).
5.2 The relations between the informal sector and the formal
sector
Although employees and apprentices on average earn less than the legal
monthly minimum wage in the formal sector, it should be noted that the
income distribution within the informal sector is much more equal than in
the formal sector of developing countries. In Tanzania on average employ-
ees in the informal sector even seem to earn more than the legal monthly
minimum wage (Table A.25).
Furthermore, the training of informal-sector entrepreneurs, workers and
apprentices is overwhelmingly done within the urban informal sector at
ridiculously low costs, again in comparison with the formal sector. Thus, in
Tanzania 91 per cent of the informal-sector entrepreneurs received their
training in the informal sector, either ‘on the job’ (35 per cent) or via an
apprentice system (57 per cent). The latter system seems predominant in
Africa (Table A.22).
Summarising, the available data clearly demonstrate the relevance of the
urban informal sector for socioeconomic development. The urban informal
sector generates substantial amounts of employment, income and training,
absorbs migration and provides a relatively more equal income distribu-
tion, all this at only minimal capital costs. Thus, on the basis of our survey
of the available empirical evidence, we may conclude that the urban infor-
mal sector – much more than its formal counterpart – can create a substan-
tial middle class, ensuring socioeconomic development, not only for
the happy few in the formal sector, but for larger segments of the whole
population.
important questions have not been asked: How much credit? All at once or
in phases? How much in kind and how much in money? For which activi-
ties? For which enterprises? Credit for what (which equipment, which raw
materials, which semi-manufactured products, which tools, which
premises)? How must credit be supervised and be repaid?
In short, specification of the measures according to activity and
phasing in time is lacking. Furthermore, the mutual relationships
between the measures are often obscure, and alternatives are seldom for-
mulated. It is even more rare for alternatives to be compared with each
other on the basis of explicit criteria, as is necessary for deciding on
policy priorities.
When a government has only limited resources – a rather typical situa-
tion not only for most governments in developing countries – it seems
obvious that it should distribute these scarce resources as optimally as pos-
sible, in the sense that the expected contribution to future growth and
development is as high as possible. However, this will only be possible with
the help of a detailed selection procedure based on questions as outlined
above, resulting in alternative congruent packages of mutually supporting
measures for specific sectors and subsectors. As long as such efforts are not
made, only global policy formulations will be possible, leading to a subop-
timal allocation of scarce resources.
Appendix
X R
Agriculture 8 5 10–3
Manufacturing 24 15 28–10
Food 2 5 8–2
Clothing 5 6 10–3
Wood 6 1 7–<1
Non-metal 2 1 3–<1
Metal 1 1 4–1
Other 8 1 10–0
Construction 7 6 11–4
Trade 51 53 68–46
Transport 2 6 8–1
Services 6 15 20–6
Sources: Tanzania: References 1, 2, 3, 6, 7; (case studies) 19–21, 26–28, 32, 34–36, 39, 44–48,
53–54, 57–58, 63–64, 67–68, 75.
Notes: (Tables A.1–A.27); x = average of the case studies; (x) = variable included only when or
more case studies provide data; R = range of the case studies.
328 The Industrial Experience of Tanzania
Sources: Tanzania: References 2, 4, 5; Africa: References, 20, 25, 34, 36, 53, Asia: References
44–45, 48, 54, 57–58, 63, LA: References 19, 21, 27, 32, 47, 51, 67, 75.
Sources: Tanzania: as Table A.2; Africa: as table A and 37, 39 50, 66, 70; Asia: References 42, 48,
54, 57, 58, 63–64; LA: as Table A.2, and 25, 55, 58, 62, 65.
Sources: Africa: as Table A.3, Asia: References 24, 44–45, 48, 54, 57, 58, 53, 64, 72; LA: as
Table A.3.
Sources: Tanzania: References 2, 4; Africa: 20, 26, 33, 34, 35, 36, 46, 49, 50, 54; Asia: References
44, 45, 48, 58, 54, 57, 75; LA: References 24, 27, 38, 47, 55, 67.
Urban Informal Manufacturing Sector 329
Sources: Africa: References, 20, 34, 35, 36, 37, 40, 49, 50, 53; Asia: References 44, 45, 48, 54, 57,
63; LA: References 19, 21, 24, 51, 62, 65.
Source: Africa: References 2: 20, 34, 36, 49, 50, 53; Asia: References 2: 48, 58, 54, 59, 63;
LA: References 2: 17, 18, 21, 27, 41, 67, 71.
Source: Africa: References 20, 34, 36, 39, 50, 53; Asia: References 48, 58, 54, 59, 63;
LA: References 17, 18, 21, 27, 41, 67, 71.
Source: Africa: References 20, 34, 36, 49, 50, 53; Asia: References 48, 58, 54, 59, 63;
LA: References 17, 18, 21, 27, 41, 67, 71.
330 The Industrial Experience of Tanzania
Source: Tanzania: References 16, 18, 19; Africa: References 20, 25, 34, 35, 36, 37, 49, 50, 53, 70;
Asia: References 26, 38, 48, 58, 54, 57, 59, 63; LA: References 15, 16, 17, 18, 19, 21, 32, 57, 67,
71, 75.
Source: Tanzania: References 16, 18; Africa: References 20, 25, 34, 35, 36, 50, 53, 70;
Asia: References 26, 38, 43, 48, 58, 54, 57, 59, 63; LA: References 15, 16, 17, 18, 19, 21, 32, 57,
67, 71, 75.
Source: Tanzania: References 16, 17, 18; Africa: References 20, 22, 25, 28, 34, 35, 36, 39, 49, 66,
70; Asia: References 26, 38, 42, 43, 48, 58, 54, 57, 59, 63; LA: References 15, 16, 17, 18, 19, 21,
24, 27, 32, 38, 55, 71, 75.
Urban Informal Manufacturing Sector 331
% of entrepreneurs that experience much competition Africa (x) Asia (x) LA (x) R
with
Source: Africa: References 25, 34, 37, 53, 60, 66; Asia: References 42, 48, 58, 59, 62;
LA: References 15, 16, 17, 18, 19, 21, 24, 27.
Source: Africa: References 20, 25, 28, 34, 35, 39, 40, 46, 53, 68; Asia: References 48, 58, 54, 57,
59, 63; LA: References 17, 18, 19, 21, 38, 71, 75.
Source: Tanzania: References 16, 17, 18; Africa: References 20, 22, 28, 34, 35, 36, 46, 49, 54;
Asia: References 26, 43, 44, 45, 48, 54, 57, 59, 63, 64; LA: References 17, 18, 21, 57, 65, 75, 73.
Source: Africa: References 20, 28, 34, 35, 39, 40, 49, 68; Asia: References 26, 48, 54, 57, 59, 64;
LA: References 17, 18, 21, 32, 67, 71, 75.
332 The Industrial Experience of Tanzania
Source: Africa: References 20, 28, 34, 35, 52, 53, 68; Asia: References 34, 48, 54, 57, 59, 64.
Source: Tanzania: References 15, 16, 18, 19, 20–23; Africa: References 20, 28, 33, 34, 35, 36, 37,
39, 40, 46, 49, 52, 53, 60, 61, 68, 70; Asia: References 34, 38, 42, 44, 45, 48, 54, 57, 63, 64, 72,
74; LA: References 15, 16, 17, 18, 19, 21, 32, 57, 67, 71, 75.
Source: Tanzania: References 16, 18, 23; Africa: References 20, 22, 28, 34, 35, 36, 37, 39, 40, 46,
53, 68; Asia: References 26, 44, 45, 48, 54, 57, 63, 64; LA: References 15, 16, 17, 18, 19, 21, 32,
57, 67, 71, 75.
Source: Tanzania: References 16, 18, 23; Africa: References 20, 25, 34, 35, 36, 37, 39, 50, 53, 66,
68, 70; Asia: References 38, 43, 44, 45, 48, 58, 54, 57, 59, 63, 74; LA: References 15, 16, 17, 18,
19, 21, 67, 71, 75, 73.
Urban Informal Manufacturing Sector 333
Tanzania Africa
Source: Tanzania: References 16, 18; Africa: References 20, 28, 34, 35, 39, 40, 53.
Source: Tanzania: References 16, 18, 20; Africa: References 20, 22, 28, 34, 35, 36, 39, 49, 50, 53,
68; LA: References 15, 16, 17, 18, 19, 57, 71.
Africa(x) R
Overall 76 37 64 98–14
International 7 0 3 13–0
Inter-regional 32 10 11 42–6
Intra-regional 36 27 50 60–4
Source: Africa: References 20, 36, 37, 53; Asia: References 48, 54, 57, 63, 64; LA: References 17,
18, 32, 57, 75.
Source: Tanzania: References 16, 18, 19; Africa: References 20, 34, 35, 36, 37, 53;
Asia: References 43, 48, 54, 57, 49, 63; LA: References 15, 16, 17, 18, 19, 21, 55, 71, 73.
Africa(x) Asia(x)
Trade 1 1
Services 1.3 1.8
Construction 4.9
Manufacturing 2.4
Source: Africa: References 20, 34, 35, 36, 37, 53; Asia: References 43, 48, 54, 59, 63.
Urban Informal Manufacturing Sector 335
Source: Tanzania: References 16, 18; Africa: References 20, 34, 35, 36, 37, 53; Asia: References 43,
48, 53, 54, 57, 59, 63; LA: References 15, 16, 17, 18, 19, 21, 55, 71, 73.
Tanzania (%)
1. Capital or credit 41
2. Lack at equipment/spare parts 16
3. Finding premises 12
4. Market/customers 11
5. Government regulations 6
6. Raw materials 6
7. Transport 3
8. Skilled workers 1
9. Service, e.g. water 1
10. Other 3
Source: (Tables A.28–A.31) Planning Commission, Ministry of Labour and Youth Development
(1991).
336 The Industrial Experience of Tanzania
Tanzania (%)
1. Credits 32
2. Non-payment debts 20
3. Unavailability of capital equipment 15
4. Raw materials 7
5. Transport 5
6. Space 4
7. Management skill 3
8. Regulations/law 2
9. Taxes/licence fees 2
10. Thieves 2
11. Spare parts 1
12. Skilled personnel 0
13. Other 8
Tanzania (%)
1. Loans 61
2. Permanent site 16
3. Raw materials 6
4. Transport 5
5. Marketing 4
6. Modern technology 4
7. Government regulations 2
8. Training, self 2
9. Training, workers 0
10. Other 0
Tanzania (%)
Notes
* Section of Technology and Development Studies, Faculty of Technology
Management, Eindhoven University of Technology
1. The list of Tanzanian studies in the References (1–14) is complete and updated.
The list of case studies in the References (15–75) contains only two references
after 1990. Thus the comparative database refers to studies from the period
1975–90.
References
12. Bienefield, M. (1975), ‘The Informal Sector and Peripheral Capitalism: The Case
of Tanzania’, Institute of Development Studies Bulletin, 6, 3, 33–73.
13. ILO (1993), The State and the Informal Sector in Tanzania: Report of a Study on the
Development Policies and Institutional Environment for the Informal Sector in
Tanzania, Addis Ababa: Jobs and Skills Programme for Africa.
14. Vitta, P.B. (1985), ‘The Informal Sector in Eastern Africa: Selected Policy-Related
Issues’, Africa Development, 10, 4, 59–71.
32. Defranco, S.D. (1980), Employment and the Urban Informal Sector: The Case of
Managua, Ann Arbor, Michigan: Xerox Univ. Microfilms.
33. Diemer, G. and W. van der Laan (1981), ‘The Informal Sector in Historical
Perspective: The Case of Tunis’, Cultures et Développement, 13, 1–2.
34. Dijk, M.P. van (1980), ‘De Informele Sector van Ouagadougou en Dakar’,
Amsterdam: PhD Thesis, Vrije Universiteit.
35. Effron, L. (1980), The Informal Sector in Abidjan (Ivory Coast) in: S. Sethuraman
(ed.) (Reference 69).
36. Fapohunda, O.J. (1981), ‘Human Resources and the Lagos Informal Sector’ in
S. Sethuraman (ed.) (Reference 69).
37. Fowler, D.A. (1981), ‘The Informal Sector in Freetown: Opportunities for Self-
Employment’, in S. Sethuraman (ed.) (Reference 69).
38. Gelder, P. van (1985), Werken onder de Boom: Dynamiek en Informele Sector: de
Situatie in Groot-Paramaribo, Dordrecht: Foris Publications.
39. Gerry, C. (1974), ‘Petty Producers and the Urban Economy: A Case Study of
Dakar’, Working Paper, World Employment Programme Research, no. 8, Geneva:
International Labour Office.
40. Gerry, C. (1978), ‘Petty Production and Capitalist Production in Dakar: The
Crisis of the Self-Employed’, World Development, 6, 9/10, 1127–60.
41. Gonzalez de la Rocha, M. (1984), ‘Domestic Organisation and Reproduction of
Low-Income Households, the Case of Guadalajara, Mexico’, Manchester: PhD
Thesis, University of Manchester.
42. Guisinger, S. and M. Irfan (1978), ‘Pakistan’s Informal Sector: The Case of
Rawalpindi’, Journal of Development Studies, 412–426.
43. Harris, J., ‘Character of an Urban Economy; Small-Scale Production and Labour
Markets in Coimbatore’, Economic and Political Weekly, 17, 23, 945–54, 24,
993–1002.
44. ILO (1974), Calcutta, see A. Bose (Reference 26).
45. ILO (1975), Djakarta.
46. ILO (1975), Abidjan.
47. ILO (1976), Sao Paulo.
48. Jurado, G.M. et al. (1981), ‘The Manilla Informal Sector: In Transition’, in
S. Sethuraman (Reference 69).
49. Kennedy, P. (1981), ‘The Role and Position of Petty Producers in a West African
City (Accra, Ghana)’, Journal of Modern African Studies, 565–94.
50. King, K.J. (1974), ‘Kenya’s Informal Machine-Makers: A Study of Small Scale
Industry in Kenya’s Emergent Artisan Society’, World Development, 2, 4, 9–28.
51. Laenen, A. (1988), Dinamica y Transformacion de la Pequena Industria en
Nicaragua, Dordrecht: Foris Publications, series: CEDLA Latin American Studies,
42.
52. Livingstone, J. (1991), ‘A Reassessment of Kenya’s Rural and Urban Informal
Sector’, World Development, 19, 6.
53. Mabogunje, A.L. and M.O. Filani (1981), ‘The Informal Sector in a Small City:
The Case of Kano’, in S. Sethuraman (Reference 69).
54. Marga Institute (1981), ‘Informal Sector without Migration: The Case of
Colombo’, in S. Sethuraman (Reference 69).
55. Merrick, T.W. (1976), ‘Employment and Earnings in the Informal Sector in
Brazil: The Case of Belo Horizonte’, Journal of Developing Areas, 10.
56. Middleton, A. (1981), ‘Petty Manufacturing, Capitalist Enterprises and the Proces
of Accumulation in Ecuador’, Development and Change, 12, 4, 505–24.
340 The Industrial Experience of Tanzania
57. Moir, H. (1981), ‘Occupational Mobility and the Informal Sector in Jakarta’, in
S. Sethuraman (Reference 69).
58. Mosselman, L. (1992), Kleine Ondernemers in Kerala (India), Rotterdam: Erasmus
Universiteit Rotterdam.
59. Naponen, H. (1987), ‘Organizing Women Petty Traders and Home-Based
Producers: A Case Study, India’, in Singh A. et al. (eds), Invisible Hands, Women
and Home-Based Production, New Delhi: Sage Publications.
60. Nihan, G. and R. Jourdan (1978), ‘The Modern Informal Sector in Nouakchott’,
International Labour Review, 117, 6, 709–719.
61. Nihan, G., et al. (1979), ‘The Modern Informal Sector in Lomé’, International
Labour Review, 118, 5, 631–44.
62. Pansters, W.G. (1985), De Dialektiek van Mikro en Makro, een Onderzoek naar het
Funktioneren van Kleinschalige Bedrijvigheid in Ciudad Juarez, Mexico,
Rijksuniversiteit Utrecht: Geografisch Instituut, Diskussiestukken van de vak-
groep Sociale Geografie van Ontwikkelingslanden, nr. 30.
63. Papola, T.S. (1978), The Informal Sector in an Urban Economy: A Study in
Ahmedabad, New York: PhD Thesis, New York University.
64. Papola, T. (1981), Urban Informal Sector in a Developing Economy, New Delhi:
Vikas.
65. Pfeiffer, P. (1987), Politische, Sozio ökonomische und Urbanistische Aspecte der
Favelas und ihre Sociale Organisation in Rio de Janeiro, Stuttgard: PhD Thesis,
University of Stuttgard.
66. Regt, M. de (1990), Tapijtknoopsters in Rabat-Salé (Marokko), Nijmegen: PhD
Thesis, University of Nijmegen.
67. Sanchez, C.B. et al. (1981), The Informal and Quasi-Formal Sectors in Cordoba in
S. Sethuraman (Reference 69).
68. Sethuraman, S.V. (1977), ‘The Urban Informal Sector in Africa’, International
Labour Review, 116, 3, 343–352.
69. Sethuraman, S. (ed.) (1981), The Urban Informal Sector in Developing Countries,
Employment, Poverty and Environment, Geneva: ILO.
70. Simon, D. (1984), ‘Urban Poverty, Informal Sector Activity and Intersectoral
Linkages: Evidence from Windhoek (Namibia)’, Development and Change, 15, 4.
71. Smith, S. (1988), Een Studie van de Ontwikkelingsmogelijkheden van de Informele
Kleding- en textielnijverheid in Aguascalientes (Mexico), Nijmegen: Katholieke
Universiteit Nijmegen.
72. Sit, V.F.S. et al. (1980), ‘Ambulatory Labour in Hong Kong’, International Labour
Review, 119, 4, 505–514.
73. Suarez, R. (1988), Informal Sector, Labour Markets and Returns to Education in Peru,
Washington D.C.: World Bank.
74. Tom, J. (1987), De Inschakeling van Vrouwenarbeid in de Zijde-Industrie in Zuid-
India: de Dynamiek van de Informele Sector Belicht, Amsterdam: Vrije Universiteit:
Geografische en Planologische Notities, no. 51.
75. Tosta Berlinck, M. et al. (1981), ‘The Urban Informal Sector and Industrial
Development in a Small City: The Case of Campinas’, in S. Sethuraman
(Reference 69).
15
The Impact of Reforms in Tanzania:
The Case of Privatized
Manufacturing Industries
Humphrey P.B. Moshi*
1 Introduction
2 Conceptual framework
341
342 The Industrial Experience of Tanzania
• There is robust evidence that state enterprises and mixed enterprises are
less profitable and less efficient than private corporations (Boardman
and Vining 1989, p. 17).
• There is virtually universal consensus that privatisation improves
efficiency (Boyco, et al., 1993, p. 1).
• Without exception, the empirical findings indicate that the same level
of output could be provided at substantially lower costs if output were
produced by the private rather the public sector (Bennett and Johnson,
1979, p. 59).
(ESAP, 1989/90– 1991/92). The overall performance of NESP and SAP had been
very poor because these programmes failed to improve the domestic produc-
tion capacity of the economy. The critical weakness of the pre-1986 reforms
was their failure to address fundamental constraints inherent in a highly con-
trolled and state dominated socioeconomic system. In the subsequent reform
efforts, however, these impediments did receive adequate attention.
In an attempt to improve the performance of the industrial sector two
policies were adopted. The first policy shift was in favour of financing
improvements in utilization of existing capacities rather than in creation of
new capacities. This shift was informed by the realization that the speed at
which capacity was being created had outstripped the ability to finance
recurrent inputs, especially imported inputs (Wangwe, 1992b). Therefore,
resources were directed towards rehabilitation of productive sectors as well
as of the infrastructure. To that effect donor initiatives took the form of
balance of payment support, import support and commodity import
support. The overall impact of this policy shift is reflected in fairly modest
real growth rates of investment.
The second policy shift, which was adopted in the early 1990s, was that
of privatisation of public enterprises as a way of turning around the perfor-
mance of the economy in general. At this juncture it was realized that there
was a need to redefine the roles of the state and the private sector in the
national economy in order to achieve an appropriate balance, supportive of
growth and efficiency.
The response of the private sector to the new policy environment has
been quite positive as far as investment is concerned. The share of private
investment in total investment increased from about 50 per cent in the
mid-1970s to 60 per cent in 1986, and further to 70 per cent in the early
1990s. This partly reflects the slow-down of investment in the parastatal
sector and partly the private-sector response to economic liberalization and
an improved government attitude towards private enterprise.
An analysis of the investment behaviour of both the private and public
sector for the 1967–95 period suggests that public investment reduced its
contribution to GDP to 8.1 per cent during 1981–85 and rose slightly to
10.3 per cent in the reform period. In the case of private investment, a dif-
ferent pattern is observed. Its share of GDP depicts a rising trend, rising
from a low 11.3 per cent in 1967–80 to 17.6 per cent during the reform
period (Likwelile, 1997).
The public-sector trend observed with respect to GDP is also reflected in
the case of gross fixed capital formation. Having contributed significantly
to GFCF during the pre-reform period (54.1 per cent), the public sector
share fell to 32.8 per cent (1981–85) and then slightly rose again to 37.6 per
cent (1986–95), the average for the entire period being 41.5 per cent. The
contribution of the private sector to GFCF rose from 45.9 per cent
(1967–80) to 62.4 per cent during the reform period, with an average of
344 The Industrial Experience of Tanzania
53 per cent for the whole period. The decrease in public investment should
be celebrated with some caution, however. Whether or not public invest-
ments crowd out private investment depends very much on the nature of
those public investments. For instance, investment in infrastructure tends
to complement private investment (Moshi and Kilindo, 1995). This points
to the need for public investment to concentrate on those projects which
demonstrate positive relationships between private and public investment.
On the privatization front, evidence shows that the privatization exercise
has not been very successful in terms of speed and numbers of enterprises
involved. So far only about 68 per cent (270 out of 395) of the firms target-
ted for divestiture have actually been divested. Further, the privatization
exercise has not been able to attract the envisioned massive amounts of
foreign direct investment (FDI).
The performance of the industrial sector shows some improvement.
Overall manufacturing output expanded as access to foreign exchange
improved, and trade and exchange-rate reforms reduced the prevailing
anti-export bias. Capacity utilisation has increased to an average of 50 per
cent. Nevertheless, the share of manufacturing GFCF has almost remained
the same, at 22.2 per cent of GDP for the last five years. Further, in the
recent past (1991–95) the growth of industrial output declined to around
two per cent (Wangwe, Musonda and Kweka, 1997).
• Improved access to finance via foreign direct investment for the compa-
nies which were sold to multinationals. This accounts for about a dozen
cases. The establishment of private-sector financial institutions in
Tanzania has further improved the accessibility of finance.
• Acquisition of modern technology, which improves both labour and
capital productivity and the quality of product(s). Massive investment
by the new owners of breweries (US$ 48.5 million) and of TANELEC
(US$ 2.5 million) facilitated the acquisition of modern machinery and
equipment.
• Competent and accountable management that is cost conscious and
flexible, and adheres to a ‘management by objectives’ approach. The
post-privatisation retrenchments are indicative of management’s quest
for profits through cost-reduction initiatives.
Impact of Reforms in Tanzania 347
• Improved workers’ welfare. This includes better pay packages, which boost
the morale of the labour force and consequently improve productivity.
• Targeted training of employees as a way of enhancing their capabilities by
exposing them to modern organisational and production technologies.
• An improved policy environment that is more friendly to the private
sector, coupled with wide-ranging complementary reforms. That is, a
market-oriented environment.
• Near monopoly positions of the firms studied. For example, TANELEC
controls 90 per cent of the domestic market as far as transformers are
concerned. In the case of switch gears it has a total (100%) monopoly.
Likewise, TBL is a giant in the beer industry, having a market share of
about 85 per cent in 1998. Though there are small competing breweries,
a monopoly exists at the level of brands. They belong exclusively to a
particular brewery. In the case of TBL the brands are Safari, Ndovu,
Kilimanjaro and Bingwa. The Associated Breweries has a monopoly in
the production of Serengeti and kick brands.
5 Concluding remarks
It is evident from the two cited case studies that the privatisation of public
enterprises in the context of ongoing macroeconomic policy and
348 The Industrial Experience of Tanzania
Notes
* Economic Research Bureau, University of Dar es Salaam.
1. Source: TBL data files.
16
Economic Reforms, Industrialization
and Technological Capabilities in
Tanzanian Industry
Samuel M. Wangwe*
1 Introduction
349
350 The Industrial Experience of Tanzania
2 Performance of industry
4.3 Weak linkages between industry and other sectors and support
institutions
Linkages between industry and related sectors were not articulated, and
complementary investments were not made to ensure integrated develop-
ment of major sectors. For example, investments in mining, agriculture
and infrastructure were not articulated in line with the BIS (Wangwe,
1993).
Although technology institutes were established in the late 1970s and
early 1980s, they did not function in ways which served the needs of
industry. For example, in their study of TIRDO, Bongenaar and Szirmai
Economic Reforms and Technological Capabilities 359
(1999) found that out of twelve projects, ten were initiated by TIRDO
officials themselves or by one of its governing bodies, rather than by enter-
prises, which are the main clients. Entrepreneurs were not involved in the
project selection process, and insufficient attention was paid to the transfer
of technologies to the users. For these reasons TIRDO could not pay ad-
equate attention to the appropriateness of technologies to the Tanzanian
environment. In general industrial linkages between TIRDO and domestic
industry were found to be weak (Bongenaar and Szirmai, 1999, and
Chapter 7 of this volume).
earners. In the case of Tanzania, 92 per cent of the poor live in rural areas
where the majority of them depend on agriculture for a living. Industrial
development will need to have a positive influence on agricultural
incomes, if poverty is to be reduced considerably.
6 Conclusion
Notes
* Executive Director, Economic and Social Research Foundation, Dar es Salaam.
17
Highlights of the Sustainable
Industrial Development Policy in
Tanzania, 1996–2020
A.K. Maziku*
1 Introduction
In his foreword to the new policy document, the Minister for Industry and
Trade writes: ‘Two major factors accounted for the need to prepare a new
policy for the industrial sector. One is the expiry of the Basic Industry
Strategy in 1995; the other is the government’s decision to phase itself out
of investing directly in productive activities and let the private sector be
the principle vehicle to the role’ (URT, 1996a).
The second factor described by the Minister is in fact one of the implica-
tions of the wide-ranging reform programme that the country has been
implementing since the mid-eighties. So far, Tanzania has successfully
transformed its political system from a one-party system to a multiparty
system. It has opened up its financial sector and implemented other
367
368 The Industrial Experience of Tanzania
One of the major constraints to the implementation of the BIS was the
lack of foreign exchange to finance imports of required machinery, spares
and raw materials. Therefore, the SIDP calls for a balanced approach
between import-substitution and export-oriented industrialization.
Regional and rural–urban disparities will also be targeted so as to bring
about equitable development. The current drive to develop physical and
economic infrastructure across the country, the provision of incentive
packages to large-scale industrial investors in less developed areas and the
promotion of small and micro-enterprises are expected to contribute to
such an equitable development.
In the SIDP the government role will change from that of being a direct
and indirect investor, through the parastatal enterprises, to a ‘facilitating
and supportive’ role by way of putting in place an enabling environment in
the form of appropriate macroeconomic policies, legal frameworks and
efficient infrastructures. On the other hand, the private sector will be
expected to invest in a manner that will contribute to the realization of the
objectives of the SIDP. This requires a very close working relationship or
partnership between the government and the private sector, based on a
convergence of purpose, interest, trust and transparency, all of which ulti-
mately depend on each side playing its role effectively. On the one hand,
government policies must be transparent, realistic, practical and predictable
and sustainable, while on the other, the private sector must be capable of
meeting the new challenges. There are three essential preconditions for this
to happen:
3.2.1 Trade
(i) A new trade policy will be formulated to streamline export develop-
ment, cross-border trade, import management and business licensing
(ii) further measures will be undertaken to implement the 1994 Fair Trade
Practices Act
(iii) special export incentives will be provided under the export promotion
strategy, which includes the establishment of export processing zones
(EPZs).
The new trade policy will chart Tanzania’s strategy in subregional eco-
nomic groupings, and other international and global programmes in the
context of the post-Uruguay Round agreements under the WTO. Measures
geared towards opening up the economy to external competition will con-
tinue to be taken, but selective protection within the framework of WTO
agreements will be applied.
There are four key players in implementing the SIDP. First, the government
will continue with the consolidation of the democratization process, the cre-
ation of a smaller, more streamlined and more efficient bureaucracy, and the
protection of individual property rights. Second, the ministry responsible for
industry will be in overall charge of the operationalization, monitoring and
review of the SIDP at the national level, in collaboration with the ministries
responsible for finance and planning. Third, the industrial support organiza-
tions, which offer essential services in the field of standardization, quality
assurance and certification, research and development, and environmental
management, will continue to play that role jointly with the government.
The fourth and last key player is the private sector. In order to be able to play
its role effectively, it will not only need to be developed (especially the local
private sector), but will need to reorganize itself so that it develops a viable
institutional network which can enter into dialogue with the government
and engage in productive joint planning on a regular basis.
The most important departure from the old plan implementation
modality will be the partnership approach adopted in the SIDP. This
approach is required for the full involvement of the private sector in the
realisation of the objectives of industrial policy. Based on the fundamental
assumption of a commitment by the partners to industrialization in
Sustainable Industrial Development Policy 373
(i) A preparatory phase that will involve putting in place a legal frame-
work within which the NCPI will operate; setting up the national co-
ordinating body; organising and staffing the technical and
management support groups; identifying industrial subsystems and
setting up their corresponding WCGs.
(ii) A study phase to establish base line data that will enable the new
approach to take off and provide the basis for detailed future studies,
including:
• diagnostic studies of the identified industrial subsystems
• analysis of the industrial structure with a view to establishing critical
gaps that need to be filled
• situation reports on the informal sector: micro- and small-scale
enterprises
• assessment of natural resource endowments for industrialization
• assessment of the performance of industrial support organizations
and the services they provide.
5 Concluding remarks
Note
* Ministry of Industry and Trade.
Bibliography
Adam, C. (1994), ‘Privatization and Structural Adjustment in Africa’, in: W. van der
Geest (ed.), Negotiating Structural Adjustment in Africa, London: James Curry.
Agrawal, N., Z. Ahmed, M. Mered and R. Nord (1993), Structural Adjustment, Economic
Performance, and Aid Dependency in Tanzania, Policy Research Working Paper
WPS1204, Washington: World Bank.
Alange, S. (1987), Acquisition of Capabilities through International Technology Transfer:
The Case of Small Scale Industrialization in Tanzania, published PhD thesis, Sweden:
Chalmers University of Technology.
Alcorta, L. (1995), The Impact of Industrial Automation on Industrial Organisation:
Implications for Developing Countries’ Competitiveness, submitted as Background
Paper no. 7 to the UNIDO Global Forum on Industry.
Appelbaum, R.P. and W.J. Chambliss (1995), Sociology, New York: Harper Collins.
Ark, B. van (1993), ‘The ICOP Approach – Its Implications and Applicability’, in: A.
Szirmai, D. Pilat and B. van Ark (eds), Explaining Economic Growth: Essays in Honour
of Angus Maddison, North Holland: Elsevier, 375–98.
Arkadic, B. van (1986) (ed.), External Finance and Policy Adjustment in Africa, London,
Sage.
Arrow, K. (1962), ‘Economic Welfare and the Allocation of Resources for Inventions’,
in: R.R. Nelson (ed.), The Rate and Direction of Inventive Activity, Princeton:
Princeton University Press.
Bagachwa, M.S.D. (1992), ‘Choice of Technology in Small and Large Firms: Grain
Milling in Tanzania’, World Development, 20 (1), 97–107.
Bagachwa, M.S.D. (1993a), ‘Impact of Structural Adjustment on the Small Enterprise
Sector in Tanzania’, in: A.H.J. Helmsing and T. Kolstee (eds), Small Enterprises and
Changing Policies, London: Intermediate Technology Publications, 91–113.
Bagachwa, M.S.D. (1993b), Export Competitiveness and Firm Level Capabilities in
Selected Tanzanian Industries, Economic Research Bureau, University of Dar es
Salaam.
Bagachwa, M.S.D. (1994), Export Competitiveness and Firm Level Capabilities in Selected
Tanzanian Industries, Research Paper, ERB, University of Dar es Salaam.
Bagachwa, M.S.D. and A.V.Y. Mbelle (eds) (1993), Economic Policy under a Multiparty
System in Tanzania, University of Dar es Salaam, DUP.
Bagachwa, M.S.D. and A.V.Y. Mbelle (1995), ‘Tanzania: Export Manufacturing’, in:
S.M. Wangwe (ed.), Exporting Africa: Technology, Trade and Industrialization in sub-
Saharan Africa, London Routledge.
Bank of Tanzania (1970 [1973, 1977, 1978, 1981, 1982]), Economic and Operations
Report June 1982 [1973, 1977, 1978, 1981, 1982], Dar es Salaam: Bank of Tanzania.
Bank of Tanzania (1975), Economic Bulletin, 7, 1.
Bank of Tanzania (1981), Tanzania: Twenty Years of Independence (1961–1981), A
Review of Political and Economic Performance, Dar es Salaam: Bank of Tanzania.
Bank of Tanzania (1984), Tanzania: Twenty Years of Independence (1961–1981): A
Review of Political and Economic Performance, Dar es Salaam.
Bank of Tanzania (1988), Economic and Operations Report 1988, Dar es Salaam: Bank of
Tanzania.
376
Bibliography 377
Bank of Tanzania (1989), Economic and Operations Report for the Year ended 30th June
1989, Dar es Salaam: Bank of Tanzania.
Bank of Tanzania (1992), Economic and Operations Report 1992, Dar es Salaam: Bank of
Tanzania.
Bank of Tanzania (1995), Economic and Operations Report 1995, Dar es Salaam: Bank of
Tanzania.
Bank of Tanzania (1996a), Economic Bulletin, 24 (4).
Bank of Tanzania (1996b), Economic Trends in Figures, June 1966–June 1996.
Bank of Tanzania (1997), Economic Bulletin, 30 Sept.
Bank of Tanzania (1998a), Economic Bulletin, 27 (4).
Bank of Tanzania (1998b), Monthly Economic Review, May 1998.
Bank of Tanzania (various), Economic and Operations Report.
Bank of Uganda (1995), Importation of Fertilizers in Uganda (1990–1994), Agricultural
Secretariat.
Barker, C.E., M.R. Baghavan, P.V. Mitschke-Collande and D.V. Wield (1986), African
Industrialisation: Technology and Change in Tanzania, Vermont: Gower.
Barro, R.J. and Jong-Wha Lee (1993), ‘Losers and Winners in Economic Growth’,
Proceedings of the World Bank Annual Conference on Development Economics, 267–98.
Washington DC: The World Bank.
Barros, A.R. (1993), ‘Some implications of New Growth Theory for Economic
Development’, Journal of International Development, 5 (5), 532–58.
Bates, R. (1981), Markets and States in Tropical Africa, Berkeley: University of
California Press.
Beadle, L.C. and E.M. Lind (1960), ‘Research on the Swamps of Uganda’, Uganda
Journal, 24, 84–98.
Beaulac, M.N. and K.H. Reckhow (1982), ‘An Examination of Land Use – Nutrient
Export Relationships’, Water Resources Bulletin, 18, 1013–24.
Bell, M., B. Ross-Larsen and L.Westphal (1984), ‘Assessing the Performance of Infant
Industries’, Journal of Development Economics, 16, 101–28.
Benjamin, J.R. and C.A. Cornell (1970), Probability, Statistics and Decision for Civil
Engineers, New York: McGraw-Hill.
Bennett, J.T. and M.H. Johnson (1979), ‘Public versus Private Provision of Collective
Goods and Services’, Public Choice, 34:55–64.
Bertholet, C.J.L. and P.E. Lapperre (1991), The Development of Societies in Global
Perspective, Eindhoven: lecture notes, Eindhoven University of Technology, Mimeo.
Biersteker (1991), ‘Reducing the Role of the State in the Economy: A Conceptual
Exploration of IMF and World Bank Prescription’, International Studies Quarterly,
34, 477–92.
Board of External Trade (1989), Ten Years of the Board of External Trade: 1979–1989,
Dar es Salaam: Tanzania Litho Ltd.
Boardman, V.E. and Vining (1989), ‘Ownership and Performance in Competitive
Environment: A Comparison of the Performance of Private, Mixed and State
Owned Enterprises’, Journal of Law and Economics, 32 (April), 136.
Bol, D. (1995), Employment and Equity Issues in Tanzania, A paper presented at an ERB
workshop on Beyond SAP in Tanzania, Dar es Salaam.
Bongenaar, B. (1997), ‘Analyzing Technology Development, Part I: Evaluating the Role of
the Tanzania Research and Development Organization’, MSc thesis, Eindhoven:
Technology and Development Studies, Eindhoven University of Technology.
Bongenaar, B. and A. Szirmai (1998), The Role of a Research and Development Institute
in the Development and Diffusion of Technology: the case of TIRDO, paper presented at
378 Bibliography
Bureau of Statistics (1982), The Economic Survey 1981, Dar es Salaam, Government
Printer.
Bureau of Statistics (1983), Statistical Abstract, 1982, Dar es Salaam: Bureau of
Statistics, Ministry of Economic Affairs and Development Planning.
Bureau of Statistics (1985a), National Accounts of Tanzania, 1976–1984, Sources and
Methods, May.
Bureau of Statistics (1985b), Census of Industrial Production, vol. 3, Dar es Salaam:
Bureau of Statistics, Ministry of Planning and Economic Affairs, March.
Bureau of Statistics (1986), Input–Output Table of Tanzania, 1976, a Background on
Sources and Methods, Dar es Salaam: Bureau of Statistics, Ministry of Finance,
Economic Affairs and Planning.
Bureau of Statistics (1988a), Industrial Commodities, Quarterly Report 87–1 [1985:4,
1986:1–4, 1987:1–4, 1988:1–2, 1989:1–4, 1990:1–2, 1990:1–4, 1992: 1–2, 1991:1–4,
1992:1–4, 1993:1, 1993:3, 1993:4, 1994:1, 1995:3], Dar es Salaam: Bureau of
Statistics, Ministry of Finance, Economic Affairs and Planning.
Bureau of Statistics (1988b), Population Census: Preliminary Report of the Bureau of
Statistics, Ministry of Finance, Economic Affairs and Planning, Dar es Salaam:
Government Printer.
Bureau of Statistics (1990), National Accounts of Tanzania, 1976–1989, Dar es Salaam:
Bureau of Statistics, President’s Office, Planning Commission.
Bureau of Statistics (1992), National Accounts of Tanzania, 1976–1991, Dar es Salaam:
Bureau of Statistics, President’s Office, Planning Commission.
Bureau of Statistics (1993a), Census of Industrial Production (1989), vol. V, Methodology
Report, Dar es Salaam: Bureau of Statistics, President’s Office, Planning
Commission.
Bureau of Statistics (1993b), National Accounts of Tanzania, Dar es Salaam: Bureau of
Statistics, President’s Office, Planning Commission.
Bureau of Statistics (1994a), Census of Industrial Production (1989), vol. IV (Part 1),
Comprehensive Report 10+, Dar es Salaam: Bureau of Statistics, President’s Office,
Planning Commission.
Bureau of Statistics (1994b), Environment Statistics in Tanzania Mainland, Dar es
Salaam: Bureau of Statistics, President’s Office, Planning Commission.
Bureau of Statistics (1994c), National Accounts of Tanzania, 1976–1993, Dar es Salaam:
Bureau of Statistics, President’s Office, Planning Commission, August.
Bureau of Statistics (1994d), Statistical Abstract, 1992, Dar es Salaam: President’s
Office, Planning Commission, Government Printer, February.
Bureau of Statistics (1994e), Selected Statistical Series, 1951–1991, Dar es Salaam:
President’s Office, Planning Commission.
Bureau of Statistics (1995a), Revised National Accounts of Tanzania, 1976–1990, Dar es
Salaam: Bureau of Statistics, President’s Office, Planning Commission.
Bureau of Statistics (1995b), National Accounts of Tanzania, 1976–1994, Dar es
Salaam: Bureau of Statistics, President’s Office, Planning Commission.
Bureau of Statistics (1995c), Selected Statistical Series, 1951–1992, Dar es Salaam:
Bureau of Statistics, President’s Office, Planning Commission.
Bureau of Statistics (1995d), Workshop on Consumer Price Index, 6 to 8 February 1995,
Mkonge Hotel, Tanga.
Bureau of Statistics (1995e), Selected Statistical Series, 1951–1994, Dar es Salaam:
Bureau of Statistics, President’s Office, Planning Commission.
Bureau of Statistics (1996a), Producer Price Index (Manufacturing Sector), Quarterly Report
1995:3, Dar es Salaam: Bureau of Statistics, President’s Office, Planning
Commission.
380 Bibliography
Clark, W.E. (1978), Socialist Development and Public Investment in Tanzania, Toronto:
University of Toronto Press.
Clerides, S.K., S. Lachs, and J.R. Tybout (1998), ‘Is Learning by Exporting Important?
Micro-Dynamic Evidence from Colombia, Mexico, and Morocco’, Quarterly Journal
of Economics, 113, 903–48.
Cody, J., H. Hughes and D. Wall (1980), Policies for Industrial Progress in Developing
Countries, Oxford: Oxford University Press.
Cody, J., R. Kitchen and J. Weiss (1990), Policy Design and Price Reform in Developing
Countries: Guidelines with Special Reference to Industry, New York: Harvester
Wheatsheaf.
Collier, P. and J. Gunning (1999), ‘Explaining African Economic Performance’,
Journal of Economic Literature, 37, 64–111.
Cooper, C. (1991), Are Innovation Studies on Industrialized Economies Relevant to
Technology Policy in Developing Countries?, UNU/INTECH, Working Paper no. 3,
Maastricht.
Cooper, C. (1995), Technology, Manufactured Exports and Competitiveness,
UNU/INTECH Discussion Paper no. 9513, December.
Cooper, C.M. (1993), Technology and Innovation in the International Economy, London:
Edward Elgar.
Cooper, C., et al. (1994), Import Liberalization, Industrialization and Technological
Capability in Africa: An Outline of a Research Project, Oxford.
COSTECH (1990), Tanzania Commission for Science and Technology, Dar Es Salaam.
Coulson, A. (ed.) (1979), African Socialism in Practice, Nottingham: Spokesman.
Coulson, A. (1982a), ‘The State and Industrialization in Tanzania’, in: M. Fransman
(ed.), Industry and Accumulation in Africa, London: Heinemann, 60–76.
Coulson, A. (1982b), Tanzania: A Political Economy, Oxford: Clarendon Press.
Crul, R.C.M. (1993), Limnology and Hydrology of Lake Victoria, Monographs of the
African Great Lakes, UNESCO/IHP.
CS I (September 1982–February 1987), Caustic Soda Project File I, collection of approx-
imately 150 notes, reports and documents, TIRDO.
CS II (March 1987–July 1993), Caustic Soda Project File II, collection of approximately
180 notes, reports and documents, TIRDO.
Dahlmann, C., B. Ross-Larsen and L.E. Westphal (1987), ‘Managing Technological
Development: Lessons from Newly Industrialising Countries’, World Development, 15.
DANIDA (1989), Country Strategy for Strengthening Environmental Considerations in
Danish Development Assistance to Tanzania.
Dao, G.C. (1989), ‘The Developing World and the Multi-Fibre Arrangement’, in: J.
Whalley (ed.), Developing Countries and the Global Trading System, vol. 2,
Basingstoke: Macmillan Press – now Palgrave.
Dar, U. (1990), ‘Management of Technological Change: An Overview’, in: UNIDO,
Management of Technological Change: Issues and Case Studies from India, chapter XIV,
Vienna, UNIDO.
Dar es Salaam Technical College (1995), unpublished internal documents.
David, P. (1975), Technological Choice, Innovation and Economic Growth, Cambridge:
Cambridge University Press.
Davies, R. (1979), ‘Informal Sector Versus Subordinate Mode of Production: A
Model’, in: R. Bromley and C. Gerry (eds), Casual Work and Poverty in Third World
Cities, Chichester: Wiley, 87–109.
DeBresson, C. (1989), ‘Breeding Innovation Clusters: A Source of Dynamic
Development’, World Development, 17 (1), 1–16.
382 Bibliography
Surveys on the National Opportunities for the Application of New Bamboo Construction
Technologies in the Housing Sector in Tanzania, Egomond-de Wilde de Ligny, E.C.L.
van, Gaillard, H. and Janssen, J.J.A. (ed.): Centre for International Co-Operation
Activities, Eindhoven University of Technology, Eindhoven, The Netherlands and
NCC, Dar es Salaam, Tanzania, February 1996.
Egmond-de Wilde de Ligny, E.C.L. and Rijkenberg, S. (1998), The Construction
Industry in Tanzania: Challenges to Contribute to Tanzanian Socioeconomic
Development? Paper presented at the Conference on Industrial Performance of
Tanzania, Eindhoven, The Netherlands, 25–27 June, 1998.
Egmond-de Wilde de Ligny, E.L.C. van (2000), Technology Mapping for Technology
Management, PhD Thesis, Delft University of Technology.
Ehrhorn, T.H. (1982), Integrierte Energieversorgung im Brauereien, Berlin, Ehrhorn.
Elbadawi, I. (1998), ‘Real Exchange Rate Policy and Export Competitiveness in sub-
Saharan Africa’, paper presented at a UNU/WIDER Workshop on the Impact of
Liberalization on Key Markets in sub-Saharan Africa, Addis Ababa, Ethiopia,
March.
Engblom, A.C. and V. Veres (1998), Capacity Utilization and the Impact of Structural
Adjustment: The Case of the Tanzania Textile Industry unpublished MA thesis,
Gothenburg University.
Engelen, D.M. van (1996), Public Policy Making and the Rise and Fall of the
Tanzanian Manufacturing Sector, MSc thesis, Eindhoven: Technology and
Development Studies, Eindhoven University of Technology.
ESCAP (1989), A Framework for Technology Based Development: An Overview of the
Framework for Technology-Based Development, Bangalore: United Nations.
ESRF (1997), Multi-Country Comparative Study of Private Enterprise Development (MCC-
PED): A Tanzania Country Report, (S.M. Wangwe, H.H. Semboja, J.P. Kweka and
H.B.P. Moshi), Dar es Salaam, Tanzania.
ESRF (1998), Labour Market Survey for VETA, Dar es Salaam, May.
ESRF-PSRC (1996), The Impact of Divestiture and Privatization Process on Industrial
Performance: The Case of Six Divested Companies, a report submitted to the Parastatal
Sector Reform Commission (PSRC), Dar es Salaam.
ESRF-VETA (1998), Labour Market Survey in the Dar es Salaam Region, final draft report.
Evans, J. (1985), The Flowering of the Middle Ages, London: Thames and Hudson.
Co-penhagen.
Fei, J.C.H. and G. Ranis (1964), Development of the Labour Surplus Economy, Theory and
Policy, Homewood, Ill: Richard D. Irwin.
Fei, J.C.H. and G. Ranis (1976), ‘A Model of Growth in the Open Dualistic Economy:
the Case of Korea and Taiwan’, Journal of Development Studies, May.
Fereira, M.L. (1993), Tanzania: A Poverty Profile, Washington DC: World Bank.
Fereira, M.L. and C.C. Griffin (1995), Human Resource Development Survey, Tanzania,
Washington DC: World Bank, I, Annexes I & II.
Fleisheuer, M. (1994), Road Construction in Tanzania: Problems, Constraints and
Opportunities. A Case Study on the Execution of Road Construction in Tanzania,
MSc thesis, Eindhoven: Technology and Development Studies, Eindhoven
University of Technology.
Foster, A.D. and M.R. Rosenzweig et al. (1995a), Learning by Doing and Learning From
Others: Human Capital and Technical Change in Agriculture, University of
Pennsylvania.
Foster, A.D. and M.R. Rosenzweig et al. (1995b), Technical Change and Human Capital
Returns and Investments: Evidence from the Green Revolution, University of
Pennsylvania.
384 Bibliography
Helleiner, G.K. (ed.) (1992), Trade Policy, Industrialization and Development, Oxford:
Clarendon Press.
Helleiner, G.K. (1994), ‘From Adjustment to Development in sub-Saharan Africa:
Consensus and Continuing Conflict’, in: G.A. Cornia and G.K. Helleiner, From
Adjustment to Development in Africa: Conflict, Controversy, Convergence, Consensus,
London: Macmillan Press – now Palgrave.
Hemming, R. and K. Miranda (1991), ‘Privatization’, in: Ke-young Chu and R.
Hemming (ed.) Public Expenditure Handbook, IMF, Washington, DC 139–45.
Henley, J.S. and G.B. Assaf (1993), ‘Privatization in an African Context: The Case of
the United Republic of Tanzania’, United Nations Industrial Development
Organisation, Industry and Development, 32.
HIID (1998), Africa Competitiveness Report, Boston: Harvard,
Hillebrand, H. and H. Jansen (1991), ‘Voorwaarden voor industrialisering’, in:
J.H. Righart (ed.), De Trage Revolutie: Over de Wording van Industriele Samenlevingen,
Meppel: Boom.
Husain (1993), ‘Comment on Barro and Wha’, Proceedings of the World Bank Annual
Conference on Development Economics, 305–310, Washington DC: The World Bank.
Hyden, G. (1982), Beyond Ujamaa in Tanzania: Underdevelopment and an Uncaptured
Peasantry, London: Heinemann.
ICRAF (International Centre for Research in Agroforestry) (1995), ‘Rain Sample
Analysis’, in: A. Braun (ed.), Nutrient Flows in Fallow Systems and Maize in Western
Kenya, 163.
ILO (1972), Employment, Incomes and Equality: A Strategy for Increasing Productive
Employment in Kenya, Geneva: International Labour Organization.
ILO (1982), Basic Needs in Danger, Addis Ababa.
ILO (1992), Increasing Productive Employment in East Africa, Geneva: ILO.
Iwaarden, A. van (1997) Production and Application of Timber in the Construction Sector,
Tanzania, MSc thesis, Eindhoven: Technology and Development Studies,
Eindhoven University of Technology.
Iwugo, K.O. (1990), Basic Concepts of Environmental Quality Control in Relation to
Baseline Studies on Land-Based Sources of Pollution, lecture presented at the
IOC–SAREC–UNEP–FAO–IAEA–WHO workshop on regional aspects of marine pol-
lution, Mauritius, 20 October–9November, Mimeo.
James, J. (1995), The State, Technology and Industrialization in Africa, Basingstoke:
Macmillan Press – now Palgrave.
James, J. (1998), Public Choice, Technology and Industrialization in Tanzania,
paper presented at the conference on The Industrial Performance in Tanzania,
Eindhoven, 25–27 June.
Jansen, H. (1991), ‘De sterke en Flexibele Staat: Politiek-Institutionele Voorwaarden
Voor Industrialisatie’, in: H. Righart, (ed.), De Trage Revolutie: Over de Wording van
Industriële Samenlevingen, Meppel: Boom.
Jedruszek, J. (1979), Development in Employment and Productivity in Tanzania, 1967–77,
paper 78.5/78.6 Economic Research Bureau, University of Dar es Salaam.
Jolliffe, D. (1995), Cognitive Skills, Schooling and Income: Evidence from Ghana,
manuscript.
Jørgensen, S.E. (1980), Lake Management, Oxford: Pergamon.
Kahama C.G. (1982), Tanzania Professional Centre 2nd Inter Professional Conference on
Priorities in National Development: Executive Summary, Dar Es Salaam.
Kaplinsky, R. (1984), ‘Third World Industrialisation in the 1980s: Open Economies
in a Closing World’, Journal of Development Studies, 21 (1).
Kaplinsky, R. (1990), The Economies of Small, London: IT Publications.
386 Bibliography
Krugman, P. (1987), ‘The Narrow Moving Band, the Dutch Disease, and the
Competitive Consequences of Mrs. Thatcher – Notes on Trade in the presence of
Dynamic Scale Economies’, Journal of Development Economics, 27, 41–55.
Kulindwa, K.A. (1994), Energy Consumption Patterns in the Manufacturing Sector of
Tanzania, PhD thesis, Dar es Salaam: University of Dar es Salaam.
Kumar, N. and M. Saqib (1994), Firm Size, Opportunities for Adaptation, and In-House
R&D Activity in Developing Countries: The Case of Indian Manufacturing, UNU-
INTECH Working Paper no. 13.
Kyessi, A.G. (1995), Overview of the Human Settlements Development and
Management in Tanzania, conference paper presented at the National Workshop on
Launching Habitat II for NGOs, CBO and other actors, Dar es Salaam: CHS,
ARDHI.
Kyessi, A., A.A. Mwaiselage and B. Seleki (1995), Shelter Situation in Tanzania, a paper
presented during the National Workshop on Prefabrication in Housing
Construction, Dar es Salaam: CHS, ARDHI.
Lake Basin Development Authority – Republic of Kenya (1984), Technical Course on
Water Quality Monitoring and Analysis, Kakamega, January, Final Report.
Lake Basin Development Authority – Republic of Kenya (1987), The Study of the
Integrated Regional Development Master Plan for the Lake Basin Development Area, 7
vols.
Lall, S. (1982), Developing Countries as Exporters of Technology: A First Look at the Indian
Experience, London: Macmillan Press – now Palgrave.
Lall, S. (1985), ‘Trade in Technology by a Slowly Industrializing Country: India’, in:
N. Rosenberg and C. Frischtak (eds), International Technology Transfer: Concepts,
Measures and Comparisons, New York: Praeger.
Lall, S. (1987), Learning to Industrialize: The Acquisition of Technological Capability by
India, London: Macmillan Press – now Palgrave.
Lall, S. (1990), Building Industrial Competitiveness in Developing Countries, Paris: OECD.
Lall, S. (1992a), ‘Technological Capabilities and Industrialization’, World
Development, 20 (2).
Lall, S. (1992b), Alternative Development Strategies in sub-Saharan Africa. New York: St.
Martin’s Press – now Palgrave.
Lall, S. (1995), ‘Structural Adjustment and African Industry’, World Development, 23,
2019–31.
Lall, S. (1996), Opening Up, and Shutting Down?: Import Liberalization and Industrial
Technology in sub-Saharan Africa, Study for the United Nations Institute for New
Technologies (INTECH), QEH, Oxford.
Lall, S. (1999), ‘Technological Change and Industrialization in the Asian NIEs:
Achievements and Challenges’, in: A. Szirmai, J. Halman, and B. Verspagen (eds),
Innovation in Theory and Practice, papers of the ECIS Opening Conference,
Eindhoven: Eindhoven University of Technology.
Lall, S. and S.M. Wangwe (1997), ‘Industrial Policy and Industrialization in sub-
Saharan Africa’, A paper prepared for discussion at the AERC Research Workshop,
25 May.
Lall, S. et al. (1997), Import Liberalisation and Technological Capabilities Building in
Tanzania, Kenya and Zimbabwe, Research Report, n.p.
Lapperre, P.E. (1992), Man, Technology, Society and Development, from the Hunters and
Gatherers to the First Industrial Revolution, 3.5 Million Years Ago until AD 1760,
Helmond: Wibro Press.
Lapperre, P.E. (1993), Problems of Developing Countries in Institutional Perspective,
lecture notes, Eindhoven: Eindhoven University of Technology.
388 Bibliography
M.S.D. Bagachwa and A.V.Y. Mbelle, Economic Policy under a Multiparty System in
Tanzania, Dar es Salaam: Dar es Salaam University Press.
Little, I.M.D., R.N. Cooper, W.M. Corden and S. Rajapatirana (1993), Boom, Crisis and
Adjustment: The Macroeconomic Experience of Developing Countries, Washington:
World Bank.
Lofchie, M. (1988), ‘Tanzania’s Economic Recovery’, Current History, May edition.
Luvanga, N. (1994), Growth of Small and Medium Size Enterprises under Liberalized
Markets: The Tanzania Experience, paper prepared for the ICGE, June 1994,
Harare.
Lynch, D. (1995), ‘Does Financial Sector Development Matter to Investment?’,
Savings and Development, 1 (19).
Maddison, A. (1995) Monitoring the World Economy, Paris: OECO Development
Centre.
Maddison, A. and B. van Ark (198), Comparisons of Real Output in Manufacturing,
Washington: World Bank, Working Paper, no. 5.
Maddison, A. and B. van Ark (1994), International Comparisons of Real Output and
Productivity, Research Memorandum GD-b. Groningen Growth and Development
Centre.
Mageni, O.S. (1994), The Role of Government on Industrial Research and Development
Institutions, TEMDO/ UNDP Regional Workshop on ‘Roles, Relevance, and
Capabilities of Industrial Research and Development Institutes: Optimising for the
21st Century’, Arusha.
Malecki, E.J. (1991), Technology and Economic Development – The Dynamics of Local,
Regional and National Change, Harlow: Longman Science & Technical.
Maliyamkono, T.L. and M.S.D. Bagachwa (1990), The Second Economy in Tanzania,
London: James Curry.
Maro, W.E. et al. (1991), Construction Cost Trends and Control Mechanisms in Tanzania,
Economic Research Bureau/University of Dar es Salaam.
Masuha, J., B. Mutagahywa and B. Nyichomba (1992), Phase I of Development of
Improved Production Systems (DIPS): the TDP Project, Dar es Salaam: University of Dar
es Salaam, Faculty of Engineering.
Mattikalli, N.M. and K.S. Richards (1996), ‘Estimation of Surface Water Quality
Changes in Response to Land Use Change: Application of the Export Coefficient
Model Using Remote Sensing and Geographical Information System’, Journal of
Environmental Management, 48, 263–82.
Maxwell, P. (1977), Learning and Technical Change in the Steel Plant of Acindar S.A. in
Rosario, Argentina, Research Program Working Paper no. 4, Buenos Aires: Economic
Commission for Latin America.
Mayer, A.F. (1992), ‘Richtzahlen und Ihr Sinnvoller Einsatz in der Brauerei’, II,
Brauwelt, 46, 2427–9.
Mbatia, K.S. (1993), ‘Constraints to the Growth of Non-Traditional Exports:
Incentive Schemes’, Tanzania Economic Trends, 5, 3–4.
Mbelle, A.V.Y. (1988), Foreign Exchange and Industrial Development: A Study of
Tanzania, PhD thesis, Sweden: Gothenburg School of Economics and Commercial
Law.
Mbelle, A.V.Y. (1989), ‘Industrial Performance under the Economic Recovery
Programme in Tanzania’, Tanzania Economic Trends, 2 (1), 7–10.
Mbelle, A.V.Y. (1993), New Technologies and Export Orientation: The Case of
Tanzanian Textile Industries, Department of Economics, University of Dar es
Salaam.
390 Bibliography
Mongi, J.F.K. (1980), ‘The Development of Price Control in Tanzania’, in: K.S.P.
Rwegasira and L.A. Kanneworff (eds), Inflation in Tanzania, Dar es Salaam: Institute
of Finance Management.
More, C. (1997), The Industrial Age: Economy and Society in Britain, 1750–1995,
London: Longman.
Moshi, H.P.B. (1996), Fiscal Impact of Commercialization and Privatization of Public
Enterprises, Dar es Salaam: Mimeo.
Moshi, H.P.B. and A.A. Kilindo (1995), The Impact of Government Policy on
Macroeconomic Variables: The Case of Private Investment in Tanzania, Nairobi.
Mourik, C. van, E.M. van der Ouderaa and C.G. Hage (1991), Management van
Research en Development, Deventer: Kluwer.
Mukandala, R. (1985), ‘Trends in Civil Service Size and Income in Tanzania,
1967–1982’, Canadian Journal of African Studies, 2, 253–62.
Mukandala, R. (1988), The Political Economy of Parastatal Enterprise in Tanzania and
Botswana, unpublished doctoral dissertation, Berkeley: University of California.
Mumburi, U.P. (1998), Manufacturing Output Growth in Tanzania: An Empirical
Analysis, unpublished MA thesis, University of Dar es Salaam.
Musonda, M. (1992), Development Strategy and Manufactured Exports in Tanzania, PhD
thesis, Sweden: Department of Economics, University of Lund.
Musonda, M. and J.P. Kweka (1998), The Dynamics and Potential of Cluster Firms in the
Growth and Development of MSEs in Tanzania: The Case of Three Selected Clusters in
Tanzania, Dar es Salaam ESRF, Mimeo.
Mutagahywa, B. and P.N. Materu (1992), Academic–Industry Links, Development of a
Vital Partnership, Faculty of Engineering, University of Dar es Salaam.
Mwaiselage, A.A. (1992), Development of Local Construction Capacity in Building Works
through Small Scale Contractors in Tanzania, Dar es Salaam: Centre for Housing
Studies (CHS), Ardhi Institute.
Mwandosya, M. and M. Luhanga (1993), ‘Energy and Development in Tanzania,
Issues and Perspectives’, Energy Policy, May, 441–53.
MWEM (1992), The Energy Policy of Tanzania, Dar es Salaam: Ministry of Water,
Energy and Minerals.
MWEM (1995), Energy Balance of Tanzania in 1992, print-out, Dar es Salaam:
Ministry of Water, Energy and Minerals.
Mwihava, N. and A. Towo (1994), A Study and Assessment of Energy Projects and their
Effective Utilization in Tanzania, Dar es Salaam: COSTECH.
National Water and Sewerage Corporation – Republic of Uganda (1995), Statistics on
Sanitation in NWSC 9 Towns (unpublished), PO Box 7053, Kampala, Uganda.
NBAQS and BC (1992), List of Registered Architects, Quantity Surveyors and Building
Contractors Dec 1992, Dar Es Salaam.
NBS (National Bureau of Statistics), Statistical Publications, various issues, Dar es
Salaam.
NCC, Komba, S.C. (1992a), Identification & Promotion of Utilisation of Local Resources
and Locally Produced Materials for Construction Industry in Tanzania, Final report,
phase I, Dar es Salaam: National Construction Council.
NCC National Construction Council (1992b), Research and Development Issues in the
Construction Industry in Tanzania, Dar Es Salaam.
NCC National Construction Council (1995), Proposed National Construction Industry
development Policy, Dar es Salaam: National Construction Council, Ministry of
Works.
ND I (July 1981–May 1984), Natural Dyes Project File I, collection of approximately
100 notes, reports and documents, TIRDO.
Bibliography 393
ND II (May 1984–November 1990), Natural dyes project file II, collection of approx-
imately 70 notes, reports and documents, TIRDO.
Ndulu, B.J. (1993), Tanzania’s Economic Development: Lessons from the Experience
and Challenges for the Future, paper written for the International Conference on
Development Challenges and Strategies for Tanzania: An Agenda for the 21st
Century Dar es Salaam.
Ndulu, B.J. and J.J. Semboja (1994), ‘Trade and Industrialization in Tanzania: A
Review of Experiences and Issues’, in: G.K. Helleiner (ed.), Trade Policy and
Industrialization in Turbulent Times, London: Routledge.
Ndulu, B.J. and J.J. Semboja (1996), ‘The Development of Manufacturing for Export
in Tanzania’ in: G. Helleiner (ed.), Manufacturing for Export in the Developing World:
Problems and Possibilities, London and New York: Routledge.
Ndulu, B.J., W. Lyakurw, J.J. Semboja and A. Chaligha (1987), Import Tariff Study –
Tanzania, Dar es Salaam: unpublished report to the Ministry of Finance.
Nelson, R.R. (1990), ‘On Technological Capabilities and Their Acquisition’, in:
R.E. Evenson and G. Ranis, Science and Technology: Lessons for Development
Policy, Boulder: Westview Press.
Nelson, R.R. (1993), National Innovation Systems: A Comparative Analysis, Oxford:
Oxford University Press.
Nelson, R.R. and E. Phelps (1966), ‘Investment in Humans, Technological Diffusion,
and Economic Growth’, American Economic Review, 56, May 69–75.
Nelson, R.R. and S.C. Winter (1982), An Evolutionary Theory of Economic Change,
Cambridge MA Belknap Press of Harvard University Press.
Newberry, D.P. and N. Stern (1987), The Theory of Taxation for Developing Countries,
Oxford: Oxford University Press.
Niosi, J. and J. Rivard (1988), ‘Canadian Technology Transfer to Developing
Countries through Small and Medium-Size Enterprises’, World Development,
18, 11.
Niskanen, W. (1973), Bureaucracy: Servant or Master?, London: Institute of Economic
Affairs.
Njau, G.J. et al. (1997), Environmental Impact Assessment on the Proposed Thermal Power
Station in Tageta Area, Dar es Salaam, Prepared for Independent Power Tanzania
(IPTL) Ltd, May.
Nyerere, J. (1968a), Freedom and Socialism, London: Oxford University Press.
Nyerere, J. (1968b), Ujamaa: Essays on Socialism, Dar es Salaam: Oxford University
Press.
Ochumba, P.B.O. (1993), Measurement of Water Currents, Temperature, Dissolved
Oxygen and Winds in the Kenyan Lake Victoria, Contribution to the IDEAL
Symposium on the Limnology, Climatology and Palaeoclimatology of the East
African Lakes, Jinja, Uganda, 18–23 February.
Ochumba, P.B.O. and D.I. Kibaara (1989), ‘Observations on Blue-green Algal Blooms
in the Open Waters of Lake Victoria’, African Journal of Ecology, 27, 23–34.
OECD (February 1971), Latest Information on National Accounts of Less Developed
Countries, F/8, No. 5, Paris: OECD.
Oerlemans, A.M.H.M. (1985), Research on the Production of Sugar Cane in Kenya,
MSc thesis, Eindhoven: Technology and Development Studies, Eindhoven
University of Technology.
Ogaram, D.A. and W.S. Kalema (1995), Lake Victoria Environmental Management
Program, Final report on regional task No. 15: Critical Sources of Industrial and
Municipal Pollution.
394 Bibliography
Orlova, A.P. and L.V. Yaroshenko (1976), ‘Contents of Pesticide and Mineral
Fertilizers in Runoff Water of Uzbekistan Irrigation’, Formirovanie Kontrol Kachestva
Poverkhnostnykh Vod, 3, 106–9.
Pack, H. and L.E. Westphal (1986), ‘Industrial Strategy and Technological Change:
Theory Versus Reality’, Journal of Development Economics, 22.
Pack, H. (1987), Productivity, Technology, and Industrial Development, New York:
Oxford University Press.
Pack, H. (1993), ‘Productivity and Industrial Development in sub-Saharan Africa’,
World Development, January, 21, 1–16.
Pack, H. and C. Paxson (1998), Is African Manufacturing Skill Constrained?, paper
presented at the conference on The Industrial Performance of Tanzania,
Eindhoven, 25–27 June.
Page Jr., J.M. (1990), ‘The Pursuit of Industrial Growth: Policy Initiatives and
Economic Consequences’, in: M. Scott and D. Lal (eds), Public Policy and Economic
Development: Essays in Honour of Ian Little, Oxford: Clarendon Press.
Parker, R.L., R. Riopelle and W.F. Steel (1995), Small Enterprises Adjusting to
Liberalization in Five Countries, World Bank Discussion Papers, no. 271,
Washington, DC: The World Bank.
Pasinetti, L.L. (1981), Structural Change and Economic Growth, Cambridge: Cambridge
University Press.
Pasinetti, L.L. and R.M. Solow (eds) (1994), Economic Growth and the Structure of Long
Term Development, proceedings of the IEA Conference held in Varenna, Italy,
Basingstoke and London: St Martin’s Press – now Palgrave – in association with the
International Economic Association.
Perkins, F. (1980), Technological Choice, Industrialization and Development: the Case
of Tanzania, unpublished doctoral dissertation, University of Sussex, England.
Perkins, F. (1983), ‘Technology Choice, Industrialization and Development
Experiences in Tanzania’, Journal of Development Studies, 19 (2), 213–43.
Pieper, U. (1998), Openness and Structural Dynamics of Productivity and Employment in
Developing Countries: A Case of De-Industrialization?, Geneva: ILO.
Pieterson, M. (1987), Het Technisch Labyrint: een Maatschappijgeschiedenis van drie
Industriele Revoluties, Meppel: Boom.
Planning Commission & Ministry of Labour and Youth Development (1991),
National Informal Sector Survey, Dar es Salaam.
Planning Commission & Ministry of Labour and Youth Development (1995), The
Dar es Salaam Informal Sector Survey 1994/5 [only estimates for manufacturing (in
Excel Worksheet format) at our disposal], Dar es Salaam.
Porter, M.E. (1985), Competitive Advantage: Creating and Sustaining Superior
Performance, London: Collier Macmillan.
Porter, M.E. (1986), Competition in Global Industries, Boston: Harvard Business School
Press.
Porter, M.E. (1990), The Competitive Advantage of Nations, New York: Macmillan –
now Palgrave.
Prins, I.M. (1997), Manufacturing Statistics, Reconstructing Tanzanian Manufacturing
Value Added, 1995–1995, MSc thesis, Eindhoven: Technology and Development
Studies, Eindhoven University of Technology.
Prins, I.M. and A. Szirmai (1998), A Reconstruction of GDP and Employment in
Tanzanian Manufacturing, 1961–1995, Report to the Tanzanian Bureau of Statistics,
Eindhoven, December.
Psacharopoulos, G. (1991), ‘Vocational Education in Theory: Why Does Vocational
Education Fail?’, International Journal of Educational Development, 11, 3.
Bibliography 395
Rothwell, R. (1992), ‘Successful Industrial Innovation: Critical Factors for the 1990s’,
R & D Management, 22 (3), 221–31.
Rothwell, R. and W. Zegveld (1982), Innovation and the Small and Medium Sized Firm,
London: Frances Pinter.
Rothwell, R. and W. Zegveld (1985), Re-industrialization and Technology, Longman:
Harlow.
RPED (1994), Development and Growth of Industrial Enterprise in Tanzania, Center for
International Business Research, Helsinki School of Economics, Helsinki, Finland.
RPED (1995), Dynamics of Enterprise Development in Tanzania, Final Report on Round II
Survey Data, Helsinki School of Economics.
Runciman, S. (1994), A History of the Crusades, London: The Folio Society.
Ruttan, V.W. (1998), ‘The New Growth Theory and Development Economics: A
Survey’, Journal of Development Studies, 35, (2), December.
Rweyemamu, J.F. (1973), Underdevelopment and Industrialisation in Tanzania: A Study
of Perverse Capitalist Industrial Development, London: Oxford University Press.
Sagasti, F. (1979), ‘Towards Endogenous Science and Technology for another
Development’, Development Dialogue, No 1.
Samplonius, C. (1994), Energy Efficiency in the Tanzanian Industry: The Cement Industry
as a Case Study, Eindhoven University of Technology, the Netherlands.
Santschi, P.W. and P.W. Schindler (1977), ‘Chemical and Geochemical Studies of
Lake Biel’, Schweizer Zeitschrift für Hydrologie, 39, (2), 182–200.
Scheren, P.A.G.M. (1995), A Systematic Approach to Lake Water Pollution Assessment:
Water Pollution in Lake Victoria, East Africa, Eindhoven University of Technology.
Scheren, P.A.G.M., J.C. Bosboom, K. Njau and A.M.C. Lemmens (1995), ‘Assessment
of Water Pollution in the Catchment Area of Lake Victoria, Tanzania’, Journal of
Eastern African Research and Development, 25, 129–43.
Schulte, W. (1996), Technology and Quality Care in Muhimbili Medical Center, Tanzania:
Methodology to Study the Role of Technology from the View Point of Total
Quality, MSc thesis, Eindhoven: Technology and Development Studies, Eindhoven
University of Technology.
Schultz, T.W. (1975), ‘The Value of the Ability to Deal with Disequilibria’, Journal of
Economic Literature, Sept., 13, 827–46.
Schumpeter, J. (1912), Theorie der Wirtschaftlichen Entwicklung, published in English
as Theory of Economic Development (1961 edition by Redvers Opie), Oxford: Oxford
University Press.
Semboja, H.H. (1995), An Overview of the Industrial Sector and Selected Industries’
Performance During the 1980s–1990s, unpublished ESRF Technical Paper, Dar es
Salaam: ESRF.
Semboja, H.H. et al. (1996), Studies Comparing Small-Scale and Large-Scale Production in
LDCs.
Semboja, J.J. and I. Karamagi (1991), The Import Duty Tariff Study, a report prepared
for the Presidential Commission of Enquiry into Public Revenue and Taxation,
October 1991, Dar es Salaam.
Semboja, H.H. and J.P.Kweka (1997), Garment Industry, Dar es Salaam: Economic and
Social Research Foundation.
Semboja, H.H. and J.P. Kweka (1998), An Analysis of the Garment Sector, paper pre-
sented at the conference on Industrial Performance in Tanzania, Eindhoven, 25–27
June.
Semboja, H. and J. Kweka (forthcoming), Import Liberalization, Industrialization and
Technological Capability in sub-Saharan Africa: The Case of the Garment Industry in
Tanzania, ESRF Research Paper, Dar es Salaam.
Bibliography 397
Thomas, C.Y. (1974), Dependence and Transformations, New York: Monthly Review
Press.
Timmer, M.P (1996), On the Reliability of Unit Value Ratios in International
Comparisons, Groningen: Groningen Growth and Development Centre, Research
Memorandum GD 31.
Timmer, M.P. and A. Szirmai (1999), ‘Comparative Productivity Performance in
Manufacturing in South and East Asia’, Oxford Development Studies, 27, (1), 57–79.
TIRDO (1989), Ten Years of the Tanzania Industrial Research and Development
Organisation (TIRDO) 1979–1989, TIRDO.
TIRDO (1990), Semi-Detailed Energy Audit at Tanzania Breweries’ Dar es Salaam Plant,
TIRDO and AF-Energi Konsult.
TIRDO (1993a), Operational Guidelines and Procedures for the Extension Services –
Manual, TIRDO.
TIRDO (1993b), Draft TIRDO Five-Year Plan 1993/94–1997/98, TIRDO.
TIRDO (1993c), Incentive Scheme Package – 1992/93, TIRDO.
TIRDO (1995), Staff List as for 1-10-1995, TIRDO.
Todaro, M.P. (1997), Economic Development in the Third World, New York: Longman.
Toussaint, T.J.A. (1994), Energy Conservation in the Industry of Tanzania: Cost Reduction
by Energy Conservation at Tanzania Breweries Ltd, Eindhoven University of
Technology, the Netherlands.
TPC, Meetings of the Technical Programme Committee, vols 1–22, TIRDO.
TSG (1996), Roadmap Study of Four African Countries: Ghana, Namibia, Uganda and
Tanzania, Washington, DC: The Service Group.
UNESCO (1989), Statistical Yearbook 1989, Lanham: UNESCO Publishing & Bernan
Press.
UNESCO (1994), Statistical Yearbook 1994, Lanham: UNESCO Publishing & Bernan
Press.
UNESCO (1995), Statistical Yearbook 1995, Lanham: UNESCO Publishing & Bernan
Press.
UNIDO (1980), The Provision, Storage and Handling of Dyes and Chemicals for Textile
Dyeing, Printing and Finishing, UNIDO.
UNIDO (1982), Assistance to the Tanzania Industrial Research and Development
Organisation, UNIDO.
UNIDO (1988a), Industrial Restructuring Policies in Developing Asian Countries with
Particular Attention to the Textile and Garment Industry, Vienna: UNIDO, Limited
Distribution.
UNIDO (1988b), Regenerating African Manufacturing Industry: Approach and Programme,
Studies of the Rehabilitation of African Industry no. 1, PPD 101-29, December.
UNIDO (1990), Rehabilitation of African Industry, no. 12, PPD 168, 10 July.
UNIDO (1991), Manual for the Preparation of Industrial Feasibility Studies, Vienna:
UNIDO.
UNIDO (1993), Industry and Development Global Report, 1992/93, Vienna: UNIDO.
UNIDO (1995), New Materials and the Challenge for Developing Countries, Background
Paper no. 3, UNIDO Global Forum on Industry.
UNIDO (various), Industry and Development Global Report, Geneva.
United Nations (1961), Index Numbers of Industrial Production, Studies in Methods,
series F, no. 1, New York: United Nations.
United Nations (1968a), A System of National Accounts, Studies in Methods, series F,
no. 2, rev. 3, New York: United Nations.
United Nations (1968b), International Standard Industrial Classification of All Economic
Activities, Statistical Papers, series M, no. 4, rev. 2, New York.
400 Bibliography
USAID-CTI (1994), Industrial Policy Proposal, a report based on Joint USAID and
Confederation of Tanzania Industries Programme, Dar es Salaam.
Valk, P. de (1992), A General Framework for Evaluating the Performance of Textile
Enterprises in LDCs: With Application to Tanzania under Structural Adjustment, The
Hague: Institute of Social Studies.
Valk, P. de and A.V.Y. Mbelle (1990), Textile Industry under the Structural Adjustment
Programme in Tanzania, 1980–1988, Dar es Salaam.
Ven, Y. van der (1996), Privatization at Tanzania Harbors Authority, MSc thesis,
Eindhoven: Technology and Development Studies, Eindhoven University of
Technology.
Verlinden, P. (1996), Rwanda & Burundi, Country Series Amsterdam, The
Netherlands: Royal Tropical Institute (KIT).
Vighi, M. and G. Chiaudani (1987), ‘Eutrophication in Europe: the Role of
Agricultural Activities’, in: E. Hodgeson (ed.), Reviews in Environmental Toxicology,
3, 213–57.
Vleuten, F. van der (1995), Energy and Environment in the Global Cement Industry:
Strategies and Instruments to Promote Energy Efficiency in Developing Countries,
Working Paper 4, Report ECN-C–94-035, Petten: ECN.
Vleuten, F. van der and M. Yhdego (1996), Cleaner Production Technology for
Greenhouse Gas Reduction in Tanzanian Private Industry, Preinvestment Study, Dar es
Salaam: ERC.
Vocational Education and Training Authority (1994), unpublished internal docu-
ments, Dar es Salaam: VETA.
Vocational Education and Training Authority (1995), unpublished internal docu-
ments, Dar es Salaam: VETA.
Vries, J. de (1984), European Urbanization, 1500–1800, London: Methuen and Co.
Wangwe, S.M. (1979), Capacity Utilisation and Capacity Creation in Manufacturing
in Tanzania with Special Reference to the Engineering Sector, unpublished PhD thesis,
University of Dar es Salaam.
Wangwe, S.M. (1986), Technology Imports, Technological Learning and Self-Reliance in
Tanzania, Working Paper WEP 2–22/WP, Geneva: International Labor Office.
Wangwe, S.M. (1990), Reflections on the Basic Industry Strategy, professorial inaugural
lecture (unpublished), University of Dar es Salaam.
Wangwe, S.M. (1992a), ‘Building Indigenous Technological Capacity: a Study of
Selected Industries in Tanzania’, in: F. Stewart, S. Lall and S. Wangwe (eds),
Alternative Development Strategies in sub-Saharan Africa, Basingstoke: Macmillan
Press – now Palgrave – 265–93.
Wangwe, S.M. (1992b), Rehabilitations in the Manufacturing Sector on Tanzania, Dar es
Salaam.
Wangwe, S.M. (1993), ‘Implications of Changing External and Internal Conditions
for Industrial Restructuring in Tanzania’, in: M.S.D. Bagachwa and A.V.Y. Mbelle
(eds), Economic Policy Under A Multiparty System in Tanzania, Dar es Salaam
University Press.
Wangwe, S.M. (1994), ‘Technology Development in Tanzania: Some Challenges for
the 21st Century’, in: L.A. Msambichaka, H.P.B. Moshi and F.P. Mtatifikolo (eds)
(1994), Development Challenges and Strategies for Tanzania: An Agenda for the 21st
Century, Dar es Salaam: Economic Research Bureau/Department of Economic,
University of Dar es Salaam.
Wangwe, S.M. (ed.) (1995), Exporting Africa: Technology, Trade and Industrialization in
sub-Saharan Africa, London: Routledge.
Bibliography 403
Wangwe, S.M. (1998), The Vision of Industrialization, paper submitted to the Team of
Experts of the Development Vision, Dar es Salaam, ESRF.
Wangwe, S.M. and M.S.D. Bagachwa (1990), ‘Impact of Policies on Technological
Choice and Development in Tanzanian Industry’, in: F. Stewart, H. Thomas, and
T. de Wilde (eds), The Other Policy, London: Intermediate Technology Publications.
Wangwe, S.M. and N. Luvanga (1993), ‘Technological Transformation in the Third
World: The case of Tanzania’, in: S.J. Patel (ed.), Technological Transformation in the
Third World, vol. II, UNU/WIDER, Helsinki.
Wangwe, S.M. and H. Semboja (1997), Industrialization in Africa, paper contributed
to CODESRIA programme on Beyond Structural Adjustment, Dakar.
Wangwe, S.M., F. Musonda and J. Kweka (1997a), Policies for Manufacturing
Competitiveness: The Case of Tanzania, ESRF Discussion Paper, Dar es Salaam.
Wangwe, S.M., H.H. Semboja and J.P. Kweka (1997b), Multi-Country Comparative
Study of Private Enterprises Development in Africa (MCC-PED), A Draft Report for
Tanzania, Dar es Salaam.
Wangwe, S.M., H.H. Semboja, and H.B. Lecomte (1996), Exit Procedures and Economic
Transformation: The case of Tanzania, ESRF Discussion Paper Series no. 010, Dar es
Salaam.
Watanabe, S. (1987), ‘Technological Capability and Industrialization: Effects of Aid
and Sanctions in United Republic of Tanzania and Zimbabwe’, International Labor
Review, 125 (5).
Weaver, J.H. and A. Kronemer (1981), ‘Tanzanian and African Socialism’, World
Development, 9 (9/10).
Weiss, C., Jr. (1990), ‘Scientific and Technological Constraints to Economic Growth
and Equity’, in: R.E. Evenson. and G. Ranis (eds), Science and Technology: Lessons for
Development Policy, Boulder: Westview Press.
Weiss, J. (1988), Industry in Developing Countries: Theory, Policy and Evidence, New
York: Croon Helm.
Welch, F. (1970). ‘Education in Production’, Journal of Political Economy, 38, 35–59.
Wells, L. (1975), ‘Economic Man and Engineering Man: Choice of Technology in a
Low-Wage Country’, in: C.P. Timmer et al. (eds), The Choice of Technology in
Developing Countries, Cambridge: Harvard Studies in International Affairs, 69–74.
Westphal, L.E. et al. (1990), ‘The Development of Technological Capability in
Manufacturing: A Macroscopic Approach to Policy Research’, in: R.E. Evenson and
G. Ranis (eds), Science and Technology: Lessons for Development Policy, Boulder:
Westview Press.
Westphal, L.E., L. Kim and C. Dahlman (1985), Reflections on the Republic of Korea’s
Acquisition of Technological Capability, New York: Praeger.
Westphal, L.E. and R.E. Evenson (1993), Technological Change and Technology Strategy,
Development Economics Seminar, paper no. 93–1/4, Den Haag.
White, G. (1984), ‘Developmental States and Socialist Industrialization in the Third
World’, Journal of Development Studies, 21 (1).
Williams, D. (1975), National Planning and the Choice of Technology: the Case of Textiles
in Tanzania, Economic Research Bureau, paper no. 75.12, University of Dar Es
Salaam.
Williams, D. (1976), Choice of Technology and National Planning: the Case of
Tanzania, unpublished doctoral dissertation, Harvard University.
Winston, G. (1979), ‘The Appeal of Inappropriate Technologies: Self-Inflicted Wages,
Ethnic Pride and Corruption’, World Development, 7 (8/9), 835–45.
Wissema, J.G. and L. Euser (1988), Samenwerking bij Technologische Vernieuwing – De
Nieuwe Dimensie van het Management van Innovatie [Cooperation in Technological
404 Bibliography
405
406 Index
Institute for Production Innovation (IPI) labour productivity 87, 92–8, 104–5
169 new indices of production 81, 82–5,
International Comparisons of Output 87, 88–91
and Productivity (ICOP) project new insights 98; changes in
92 structure 100, 102; level
International Labour Organisation (ILO) adjustments in nominal value
326 added 100, 101; trends in real
International Monetary Fund (IMF) 34 growth 102–4
post-independence 352–4
Jedruszek, J. 353 pre-independence 351–2
reconstruction of nominal 10+ GDP
Kenya 57, 58–61, 70 76; conceptual errors 78–80;
Kimambo, R.H. 211 coverage 76–7; summary of
knowledge transfer 176–7 adjustments 81; treatment of
Krugman, P. 126 non-response in official statistics
77–8
labour productivity 41–4, 45, 87 manufacturing sector
basic data 106–11 background 300–1
1989 benchmark 92; comparisons exports 315
98; data sources 95–6; investment 316
methodology for comparison macroeconomic policies 303–6;
92–5; unit value ratios 96–8, analysis of effects 307–14
97 recent developments 301–2;
comparative 104–5 dynamics 303; signs of recovery
Malaysian pattern 118 302–3
trends 104 Masuha 360
Lake Victoria Mbelle, A.V.Y. 306
pollution 237, 239, 241–2, 247–50 and Bagachwa, M.S.D. 303
statistics 243–7 metalwork industry
Lall, S. 198 research into education shortcomings:
Lewis, A. 115, 116–17 analysis 231, 233–4;
Lewis dual economy model 116–17 background 224; facilities/
costs 230; family/community
macroeconomic policies 228; finance 230–1;
empirical analysis: methodological methodology 224–6;
issues 307–10; time series organization, contents, methods
310–14 230; results 226, 227, 228, 229,
impact on manufacturing sector 230–1, 232; school management
305–6 230; students 228; teachers
industrial development 303–5 228, 230
insufficient reform 355–6 Ministry of Land and Urban
manufacturing performance 73–5 Development 210
capacity utilization 302 Ministry of Science and Technology and
data on 352–4 Higher Education 210
data sources for time series on 10+ Ministry of Trade and Industries (MIT)
75 347, 348
employment 302 Ministry of Works 210
exports 301–2 Morogoro Canvas Mill 169
gross value added/employment Morogoro Shoe Company 346
99 Moshi Leather Industries 159
growth/share in GDP 301 Mukanda, R. 147
Index 409