Beruflich Dokumente
Kultur Dokumente
June 2019
Abstract
I use a two-sector search and matching model in order to study
the effects of a public-sector-based employment stimulus program in
Lebanon. Results show that a permanent deterministic shock on pub-
lic vacancies and wages crowds out the labour market raising by that
unemployment. This result holds when the TFP shock is positive and
is amplified when the economy is subject to a negative temporary TFP
shock. Most of the crowding out is caused by the increase in public
sector wages. These results clearly indicate that given the tax policy
in Lebanon, a government-based employment stimulus program is in-
effective and can even be harmful.
∗
I take this opportunity to thank Professor Julien Albertini without whom this work
would not have been possible. I would also like to thank all the Professors for this enriching
academic year. I dedicate this work to my parents and my brother.
1
Contents
1 Introduction 3
1.1 Literature Review . . . . . . . . . . . . . . . . . . . . . . . . 3
1.2 Research Paper Framework and Results . . . . . . . . . . . . 7
3 The model 12
3.1 General Setting . . . . . . . . . . . . . . . . . . . . . . . . . 13
3.2 Households . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
3.3 Workers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
3.4 Firms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
3.5 Nash Bargaining and Private Sector Wage Determination . . 19
3.6 Government . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
3.7 Equilibrium . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
4 Calibration 22
5 Results 24
References 34
7 Appendices 36
7.1 Steady State Tax Level Variation . . . . . . . . . . . . . . . . 36
7.2 Steady State Public Sector Wage Premium Variation . . . . . 37
7.3 Steady State Public Sector Matching Efficiency Variation . . 38
7.4 Steady State Bargaining Power Variation . . . . . . . . . . . 39
7.5 Transition Path of the Economy in Response to a Permanent
Deterministic Shock on Technology, Public Employment and
Public Wages . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
7.6 Transition Path of the Economy in Response to a Permanent
Deterministic Shock on Public Employment and Public Wages 41
7.7 Transition Path of the Economy in Response to a Permanent
Deterministic Shock on Public Wages . . . . . . . . . . . . . 42
7.8 Transition Path of the Economy in Response to a Permanent
Deterministic Shock on Public Vacancies . . . . . . . . . . . . 43
7.9 Transition Path of the Economy in Response to a Permanent
Deterministic Shock on Technology . . . . . . . . . . . . . . . 44
1
7.10 Transition Path of the Economy in Response to a Determin-
istic Temporary Shock on the Total Factor Productivity . . . 45
7.11 Transition Path of the Economy in Response to a Tempo-
rary Deterministic Shock on the Total Factor Productivity
Augmented by a Permanent Deterministic Shock on Public
Vacancies and Wages . . . . . . . . . . . . . . . . . . . . . . . 46
7.12 Transition Path of Unemployment in Response to Different
Shocks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
7.13 Transition Path of Private Sector Employment in Response
to Different Shocks . . . . . . . . . . . . . . . . . . . . . . . . 48
7.14 Transition Path of Public Sector Employment in Response to
Different Shocks . . . . . . . . . . . . . . . . . . . . . . . . . 49
7.15 Transition Path of Private and Public Employment as well as
Unemployment in Response to a Negative TFP Shock . . . . 50
2
1 Introduction
1.1 Literature Review
Previous neoclassical models failed to replicate fluctuations in hours worked.
Labour supply in these models is determined via utility maximization while
labour demand is determined via profit maximization. The intersection
of labour supply and labour demand determines equilibrium wages. This
prevents neoclassical models from predicting unemployment, and thus the
extensive margin of labour supply.
On the other hand, search and matching models, succeed in taking un-
employment into account, thus, better explain fluctuations in hours worked.
This is done by introducing frictions in the labour market. Such that; unem-
ployed workers need time to find a job and vacancies are not filled instantly.
Moreover, wages are no longer determined in a centralized market but via
bilateral bargaining.
The rationale behind adopting this new approach when dealing with
public expenditures lies in the fact that shocks to government purchases
and government employment do not have the same effects on the economy.
A positive government purchase shock results in increased investment, em-
ployment and output i.e. the shock is an expansionary impulse. While a
positive public vacancies shock is a contractionary impulse due to the crowd-
ing out effect.
3
cle. He considers a perfectly competitive labour market and divides
government spending into two components:
(a) Purchase of final goods and services. This represents the ap-
proach used in the traditional Keynesian IS-LM models and other
RBC models1 .
(b) Compensation to employees.
4
Quadrini and Trigari (2007) studied the effect of public sector em-
ployment on the private sector labour market. This was done by con-
structing a basic search and matching model where the private sector
employment is determined through profit maximization and private
sector wages are determined through Nash bargaining. In their model,
the government follows an exogenous rule for public wages and employ-
ment. They find that due to the higher wage premium in the public
sector and reduced job creation in the private sector, acyclical public
sector wages amplify the magnitude of negative shocks by crowding
out private sector jobs. The authors also show that procyclical em-
ployment reduces business cycle volatility due to crowding out effect
generated by the large public sector employment.
3. Pappa (2009) adds on the paper of Finn (1998) by allowing for nominal
rigidities. The model accommodates an unproductive government con-
sumption, a productive government investment as well as government
employment disturbances in both RBC and New Keynesian models.
5
Results show that an increase in public sector vacancies in the high or
medium regions crowds out private sector vacancies on a one-to-one
basis. The study also shows that, in the low public wage region, an
increase in the public sector wage increases the equilibrium level of un-
employment. An increase in public sector wages of the medium public
wage region implies a decrease in the average wage paid to employ-
ees in the private sector. The average private wage paid to employed
workers in the private sector is independent of an increase in wages
in the high public wage region. These results show that the average
private sector wage is non-monotonically related to the public sector
wage.
Afonso and Gomes (2014) build a search and matching model that
permits for a bidirectional relation between private and public wage.
In this context, productivity shocks first affect the private sector wage
then propagate to the public sector wage through direct and indirect
mechanisms;
(a) Directly, through a direct relationship between private and public
wage.
(b) Indirectly, through an error correction mechanism. The model
also allows the private sector wage to be influenced by an exter-
nal power such as fiscal tightening which is transmitted to the
private sector through the public sector. The paper deals with
endogenous variables problem by estimating each equation sepa-
rately and instrumenting all the endogenous variables.
6
wage growth increases private sector real wage growth by 0.3 per cent.
This represents the direct effect. A one per cent increase in public
sector employment increases the private wage by 0.2 per cent. Results
also show that the private sector wage is pro-cyclical.
Gomes (2014) studies the optimal public sector employment and wage
policy in the steady state and over the business cycle. This is done
by adopting a search and matching model in a two-sector dynamic
stochastic general equilibrium model. The author shows that in the
steady state, the optimal public sector wage policy should mirror the
differences in the probabilities of finding a job. Not following an opti-
mal policy would lead to higher equilibrium unemployment.
The author also shows that in order to avoid unemployed queues dur-
ing recessions and recruitment problems during periods of expansion,
public sector wage should follow the private sector wage.
7
The objective of the household is to maximize the expected lifetime util-
ity by choosing optimal levels of consumption and bond holdings. Firms
choose the number of vacancies they want to post as well as the number
of employees they want to hire in such a way to maximize the present dis-
counted value of profits. Both public and private sectors produce consump-
tion goods following a constant return to scale production technology, with
labour as the only input. Posting vacancies is costly, this cost is represented
in the production function where part of the produced goods goes to finance
it. These costs differ between sectors.
Unlike the private sector wage, which is determined via the Nash Bar-
gaining process, public sector wage follows the evolution of the private sector
wage. An error correction mechanism corrects, the differences between the
private sector wage and the public sector wage in such a way to maintain
the public sector wage premium.
In the model, technology, public vacancies and public wages are subject
to deterministic shocks. Previous literature has shown that these shocks
affect the private sector wage and thus private sector employment. The
mechanism in which this happens is the following: shocks to public sector
vacancies and wages crowd out private employment by increasing the num-
ber of unemployed searching for public sector jobs. Due to the diminishing
marginal productivity of labour, this crowding out effect increases marginal
productivity-raising by that private sector wages. The second mechanism
in which a deterministic shock on public sector vacancies or wages affects
private sector wages is the following: an increase in public sector wages or
public sector vacancies increases the value of being employed in the public
sector and thus the value of being unemployed and searching for a public
sector job. Since the latter is present in the bargaining solution, pressure
on private sector firms to increase wages emerges.
8
mance in Lebanon, that has been marked by years of stagnation. They also
try to replicate the increase in public sector wages and vacancies as in the
aftermath of the 2017 salary scale. Then I study the effect of both positive
and negative temporary TFP shocks on the economy.
Results also show that when combined with a public sector employment
stimulus plan, negative TFP shocks have a magnifying effect. Following
a negative TFP shock, if the government deploys an employment stimulus
plan to revive its economy, unemployment would raise even more dramati-
cally due to the magnified crowding out effect. The latter is caused by the
stimulus plan and the fact that a negative TFP shock increases unemploy-
ment. In the presence of an employment stimulus plan, a negative TFP
shock magnifies the crowding out effect. Moreover, employment stimulus
programs are ineffective even when the economy is hit by a positive TFP
shock. Despite the fact that unemployment is lower than in the case of neg-
ative TFP shock, the positive TFP shock is incapable of totally absorbing
the crowding out of the labour market.
9
2 Lebanon’s Economic Environment
2.1 Lebanon’s Financial Situation
Lebanon is one of the highly indebted countries in the world. Due to cumu-
lative fiscal deficits over the past 10 years, the country’s debt soared around
75 U.S. billion dollars, generating by that a debt to GDP ratio of 145%.
The accumulated debt is primarily used to cover the interest accumulation,
which makes the problem even worse. Since the start of the Syrian civil war
in 2011, the government’s revenues grew at 1% yearly while its expenditures
kept growing at 5%. The International Monetary Fund estimates that, fol-
lowing the same trajectory, the fiscal deficit might reach 15% of the GDP
and the debt to GDP ratio could hit 200% by the year 2030.
According to the World Bank, the Lebanese GDP was around 52 U.S.
billion dollars in 2016 while the Lebanese population was around 6 million.
The GDP per capita in 2017 was equal to 8.800 U.S. dollars. The government
expenditures represent around 27% of the GDP. This number could reach
40% in the near future due to the new salary scale law which was agreed
upon in September 2017 as well as the new reforms and development pro-
grams. Debt servicing represents 36% of governmental expenditures. While
the public sector wage bill represents 33% of governmental expenditures.
Official studies estimate that the new salary scale would cost the Lebanese
government 800 U.S. million dollars yearly. While other economic estimates
suggest that the yearly real cost of the salary scale would be 1.8 U.S. billion
dollars.
According to Jaoude (2015) the labour force participation rate and em-
ployment rates were 49.2% and 43.6% respectively. While the World Bank
estimates that the participation rate was around 47% for 2018.
10
On the other hand, there is no reliable estimation for the unemployment
rate. The World Bank report of 2012 sets the national unemployment rate
at 11%; the Central Administration of Statistics(CAS) sets the unemploy-
ment rate at 6%; the Ministry of Labour estimates it to be between 20 and
25% for the same period.
In their 2012 report Robalino and Sayed (2012), the World Bank esti-
mates that the public administration employs 30% of wage employees, the
majority of them with higher education. Between 1997 and 2009, the gross
domestic product expanded at an average rate of 3.7 per cent per year, yet
employment grew by only 1.1 per cent thus, the employment-growth elas-
ticity is as low as 0.2.
In 2018 and following the new salary scale, the Lebanese government in-
tended to post 5.000 vacancies in the public sector and open 3.000 vacancies
in security bodies.
The most powerful trade union in Lebanon accounts only for 2.5% of
the total labour force. This weak representation is mirrored by the weak
bargaining power of the union trades.
11
3 The model
I use a search and matching model with two sectors, private and public. Fol-
lowing Petrongolo and Pissarides (2001), the search and matching framework
introduces frictions that hinder the unemployed from finding jobs. This ap-
proach is governed by a Constant Return to Scale Cobb-Douglas matching
function. The search and matching model is suitable since, without these
frictions, the public and private wages would equate due to arbitrage.
I also rely on the assumption of directed search, i.e. each unemployed
searches in only one of the private or public sector. This assumption is
preferred over the random search one due to the increasing number of micro-
founded evidence that supports it. Notably, the paper of Blank (1985)2 and
Heitmueller (2006)3 .
The private sector wage is determined via Nash Bargaining with unequal
bargaining powers. Following the data for similar Middle Eastern countries,
a public sector wage premium is present. The public sector wage follows the
one of the private sector, with the help of an error correction mechanism
that tries to sustain the public sector wage premium.
2
The author relies on a reduced form two-way probit model that determines in which
sector an unemployed is more likely to search. He shows that: the sectoral choice is
influenced by wage comparison. This enables the public sector to control the number of
unemployed searching for a job in this sector by altering the public sector wage
3
The author gives evidence on the directed search and shows that wage differential
between sectors is an important driving force for sectoral assignment
12
Posting a vacancy in the private or public sector incurs costs. These
costs differ between the two sectors.
Firms exhibit a labour based production function, they chose the num-
bers of vacancies to post as well as the number of employees to hire in such
a way to maximize profits subject to the law of motion of employment in
the private sector.
The government also produces goods via a labour based production func-
tion. Distortionary taxes are set exogenously as have been the case in
Lebanon where these taxes have not changed since more than a decade.Public
sector vacancies are equal to their steady state level plus a shock.
Three types of shocks are taken into account: public sector wages shock,
public sector vacancies shock and technology shock.
ltp , ltg and ut stand respectively for the proportion of workers employed in
the private, public sector and unemployed.
13
matching function described below.
In the above matching function, mit stands for the number of new matches
in sector i and time t, m̄i stands for the matching efficiency in sector i, uit
represents the number of unemployed searching for a job in sector i = p, g,
η i represents the elasticity of the match, vti represents the vacancy rate in
sector i at time t.
3.2 Households
The household can consume two types of goods. Private sector goods ct and
public sector goods gt . The quantities consumed by the latter are taken as
constant, i.e. the consumer does not have the power to alter the consump-
tion of public goods.
14
unemployment risks and different levels of risk aversion, consumption would
not be equal among all members if they receive the same income.
Using the logarithmic utility function implies that the elasticity of in-
tertemporal substitution εu(c) = 1/σ = 11 = 1 6 . Thus the income effect and
substitution effect offset. The logarithmic utility function also satisfies some
desirable properties7 .
∂2f
(a) the function is concave on C, i.e. the Hessian matrix Hi,j = is
∂ci ∂cj
negative-semidefinite. Implying positive but decreasing marginal utility for
consumption ci ∂f (c)/∂ci > 0 ∂ 2 f (c)/∂c2i < 0
(b) the limit of the first derivative is positive infinity as ci approaches 0:
lim ∂f (c)/∂ci = +∞
ci →0
(c) the limit of the first derivative is zero as ci approaches positive infinity:
lim ∂f (c)/∂ci = 0
ci →+∞
15
one is presented in the following equation.
+∞
X
Lt = E0 β t [ln(ct ) + ζln(gt ) − λt (ct + Bt − (1 + rt−1 )Bt−1
{ct ,Bt }∞
t=0 t=0
−(1 − τt )wtp ltp − (1 − τt )wtp ltp − zt ut − Πt )](7)
3.3 Workers
The value to the household of each employed member differs based on
whether she works in the private or the public sector. It increases as the
job destruction rate decreases. The value to the household of an employed
member in sector i and period t is denoted as Wti and can be represented
as:
Wti = (1 − τt )wti + Et βt,t+1 [(1 − λi )Wt+1
i
+ λi Ut+1
i
], i = p, g. (9)
8 ∂Lt ∂Lt
∂ct
= 0 gives ∂ct
= β t ( c1t − λt ) = 0 which gives 1
ct
= λt
∂Lt ∂Lt
∂Bt
= 0 gives ∂Bt
= β t (−λt ) + β t+1 Et λt+1 (1 + rt ) = 0 substituting λt by 1
ct
and λt+1
1 ct
by ct+1 gives us the Euler equation β(1 + rt )Et [ ct+1 ]=1
16
As mentioned previously, (1−τt )wti stands for the net wage for a member
employed in sector i at time t. It is important to state that the value of an
employed member does not depend solely on the current net wage, but also
on the continuation value of being employed in sector i, (1−λi )Wt+1
i +λi U i
t+1
i i
where λ stands for the job destruction rate in sector i and Ut+1 stands for
the value to the household of an unemployed member searching for a job in
sector i at period t + 1. βt,t+1 stands for the stochastic discounting factor,
it is calculated as:
ct
βt,t+1 = βEt [ ] (10)
ct+1
Unemployed members also provide value to the household. This value
differs whether the unemployed is searching for a job in the private or public
sector. The value to the household of an unemployed member searching for
a job in sector i at time t is denoted as Uti .
Remember that we have made the assumption that the unemployed can
search for a job in only one of the two sectors, private and governmental.
This assumption is known as the directed search assumption. Thus, the
value to the household of an unemployed member depends on whether this
member is searching for a job in the private or governmental sector.
Where zt stands for the unemployment benefits plus the value of home
production and pit stands for the conditional job finding rate in sector i.
mpt Et [Wt+1
p
− Ut+1 ] mg Et [Wt+1
g
− Ut+1 ]
= t (13)
(1 − st ) st
17
Writing the optimality condition in this fashion enables us to determine
the share of unemployed searching for a job in the governmental sector st .
It is worth mentioning that the share of unemployed searching for a public
sector job st depends on the value of being employed in the public sector,
as well as the number of vacancies posted in each sector.
3.4 Firms
Private firms produce goods using labour as a production factor. Costs of
posting vacancies represent the production costs. Thus, part of the goods
produced go to finance these costs. The private consumption good is denoted
by yt . Its production function is the following:
p(1−α)
yt = at (lt − θp vtp ) (14)
+∞
p(1−α) p p p
X
max
p
Et βt,t+k [at+k (lt+k − θp vt+k ) − wt+k lt+k ](16)
{vt ,lt+k }∞
t=0 k=0
The firm maximizes the above equation, taking into account as a con-
straint the law of motion of employment in the private sector that is pre-
sented below:
p
lt+1 = (1 − λp )ltp + qtp vtp (17)
18
The Lagrangian for this constrained maximization problem becomes:
+∞
p(1−α) p p p
X
Lt = Et βt,t+k [at+k (lt+k − θp vt+k ) − wt+k lt+k +
p
{lt+k ,vt }∞
t=0 k=0
µt ((1 − λp )ltp + qtp vtp − lt+1
p
)](18)
for k = 0 we have:
+∞
p(1−α)
− θp vtp ) − wtp ltp +
X
Lt = E0 βt [at (lt
p
{lt+1 ,vt }∞
t=0 k=0
µt ((1 − λp )ltp + qtp vtp − ltp )](19)
This represents the solution to the firm problem, i.e. the maximization of
the present discounted value of profits subject to the law of motion of private
sector employment. The above equation carries an important concept; the
expected cost of hiring a worker, the left-hand term of the equation equals
the discounted expected profit of hiring a worker, the right-hand term of the
equation. The latter includes the marginal productivity of new workers (1 −
p
p−α wt+1
α)lt+1 , wages relative to technology in period t+1 at+1 and the continuation
p
value of the match (1 − λp ) qθp .
It is important to mention that, due to the
t+1
decreasing marginal productivity of labour, the crowding out effect decreases
the number of unemployed searching for private sector job, thus increases
the marginal productivity of the newly hired workers.
wtp = argmax
p
(Wtp − Ut )b (Jt )1−b
wt
at θ p
∂Lt
p
∂vt
= 0 gives ∂Lt
p
∂vt
= −at θp + µt qtp = 0 which gives at θp = µt qtp and µt = qt
p .
19
In the above equation, wtp stands for the private sector wage at period
t, Wtp and Ut follow the same definition as in previous sections, b stands for
workers’ bargaining power and Jt stands for the value of a marginal job for
the firm. It is defined in the equation below:
.
at (1 − α)ltp−α is obtained from the derivative of private sector consump-
tion goods production function with respect to private sector employment.
The solution to the bargaining problem is the following:
b(1 − τt )
(Wtp − Ut ) = (Wtp − Ut + Jt ) (23)
1 − bτt
The solution carries an important concept which can be represented in
the fact that an increase in distortionary tax levels τt lowers the share of the
surplus going to the workers. Without such a tax, the share of the surplus
received by the workers would equal their bargaining power b. Therefore,
following a distortionary tax raise, workers might accept lower wages in
order to avoid higher taxes going to a third party; the government.
3.6 Government
As in the private sector, the government produces public goods using labour
as the production factor. Moreover, part of the produced public goods go to
finance the production costs represented by the cost of posting vacancies in
the public sector. The production function of the public consumption good
is the following:
zt = zat (25)
20
The government sets its wages one period before setting vacancies. The
growth of public sector wages depends on the expected growth of private sec-
tor wages plus an error correction mechanism that re-establishes the public
sector wage premium in such a way it matches the target premium. Public
sector wages are subject to a permanent deterministic shock wt that takes
the public sector wage to a new steady-state level. The following equation
explains how wages in the public sector evolve.
g p g
wt+1 wt+1 wt−1
= E t [ ] + κ[ − 1 − Ψ] + w (26)
wtg wtp p
wt−1 t
Public sector vacancies are set to their steady state levels. As public
sector wages, they are subject to a permanent deterministic shock vt that
takes the number of vacancies to a new steady-state level. The following
equation depicts the evolution of public sector vacancies.
In the above equation vtg stands for public sector vacancies, and v̄ g stands
for steady state public sector vacancies level. It is worthy to mention that
in the model, the distortionary income tax is considered as an exogenous
parameter since the Lebanese government has not changed it since a decade
and because is not a major component of the government’s revenues.
3.7 Equilibrium
Taking public sector vacancies, wages and taxes {vtg , wt+1
g
, τt }∞
t=0 as given,
the competitive equilibrium is characterized as follows: (i) the household
chooses the sequence of consumption {ct }∞ t=0 that maximizes his lifetime
utility (equations 5 through 8), (ii) the household chooses the fraction of
unemployed searching for a job in the public sector s in such a way the
optimality condition holds and the value of being unemployed and search-
ing in one of the two sectors equates (equations 12 and 13), (iii) private
sector firms maximize their profits subject to the law of motion of private
employment (equations 16 through 20), iv) private sector wages wtp solve
the Nash bargaining condition (equations 21 through 23) and (v) the goods
market clears ct = yt + gt . The equilibrium results in a sequence of prices
{vtg , wt+1
g
}∞
t=0 that satisfy the above conditions.
21
4 Calibration
The steady-state level of unemployment, private sector employment as well
as public sector employment are set to match the data respectively at, 15%,
76.5%, and 8.5% of the total labour force. Using these figures, and as seen
in the data, public sector employment represents 10% of total employment.
The public sector labour market tightness is set at 0.0704. This figure
relies on the numbers of Jaoude (2012) for the year 2012; declared vacancies
in the public sector were 2164 while the unemployed searching for a job in
the public sector were 30756. The income tax rate is set at 2%, the cur-
rent income tax prevailing in Lebanon. Following the previous literature,
the capital share of production is set at 0.3, while the bargaining power is
set at 0.35. Very rare to no data is present on the bargaining power of the
employees in Lebanon, however, based on the fact that the most powerful
trade union in Lebanon accounts only for 2.5% of the labour force, I set the
steady state bargaining power at 0.35 while Afonso and Gomes (2014) set
it at 0.43 for the United Kingdom.
22
the North, where the presence of the public sector is more pronounced than
in the rest of the country.
23
5 Results
Steady State Tax Level Steady State Public Sector Wage Premium Level
Steady State Bargaining Power Level Steady State Public Sector Matching Efficiency Level
The left graph of the top panel shows that an increase in the steady-state
distortionary tax level would result in a sharp decrease in private sector em-
ployment as well as an increase in unemployment and public employment.
This increase in tax has two opposite effects. The first being the reduction of
the share of surplus going to the workers captured by the Nash Bargaining
solution: (Wtp − Ut ) = b(1−τ t) p
1−bτt (Wt − Ut + Jt ). This leads workers to save on
their tax bill by accepting lower wages. The second effect is the reduction
of the match surplus; (Wtp − Ut + Jt ), that obliges private employers to raise
wages in order to keep on recruiting new workers.
24
the decreasing marginal utility of labour, increases the marginal productiv-
ity of new employees, causing private sector wages to hike. This increase in
wages reduces private employment and increases by that unemployment.
The right graph of the top panel stands for the reaction of the labour
market to an increase in the steady-state public sector wage premium. It
exhibits the same patterns observed in the tax raise scenario. However, the
magnitude of change is much lower. In this case, the crowding out effect
does not occur due to a decrease in private sector vacancies, but due to a
disproportionate increase in the value of being employed in the public sector
caused by a disproportionate increase in the public sector wages.
The left graph of the lower panel studies the reaction of the labour force
to a change in the bargaining power. A slight decrease in both private and
public sector employment is observed with an increase in unemployment to
compensate for the decrease in employment. This can be explained by the
fact that higher bargaining power increases the share of the surplus going
to private sector employees which puts pressure on private firms, increas-
ing by that wages. However, the increase in wages is not as severe as in
the previous cases since, despite the bargaining power increase, the number
of unemployed members searching for a public sector job decreases. Thus,
in this case, the crowding out effect is not present. Since more unemployed
would want to work in the private sector, and given the diminishing marginal
return, the private sector wage decreases. These two opposite forces result
in a slight increase in the private sector wage and thus a slight decrease in
private employment.
The right graph of the lower panel shows the reaction of the labour
market to an increase in the public sector matching efficiency. Unlike the
25
previous cases, the increase in the matching efficiency slightly decreases un-
employment despite the large decrease in private employment. Once again,
the number of unemployed searching for a public sector job increases rapidly,
which creates a crowding out effect increasing by that private sector wages.
However, and unlike previous scenarios, the probability of filling a job in the
public sector increases sharply, absorbing by that the unemployment created
by the crowding effect and the increase of private sector wages.
Period Period
In the left graph above, we see the reaction of the labour market to
a permanent deterministic technology shock combined with a permanent
deterministic public sector vacancies and salaries shock. We notice that
subject to these shocks, private employment starts by decreasing then in-
creases slightly to stay below its initial level. On the other hand, public
sector employment increases steadily, and unemployment presents a hump
shape, starting by increasing then decreasing below its initial value.
The crowding out effect caused by the increase in the number of unem-
ployed searching for a public sector job due to the increase in public sector
wages and vacancies raises private sector wages and thus reduces private
employment. Also, the increase in the value of being employed in the public
sector increases the value of being unemployed and searching for a public
sector job. This puts pressure on the bargaining power increasing by that
private sector wages, the effect is amplified due to the fact that a greater
number of unemployed are searching for a public sector job.
However, and despite all these factors that play a role in increasing
private sector wages and thus decreasing private employment, the labour
market is capable of rebounding and coming back to its initial levels.
26
The right graph of the above panel shows the reaction of the labour
market to a deterministic permanent increase in public sector vacancies and
wages. In this graph, the decrease in private sector employment is more
speaking than in the previous case and is not followed by a rebound. This is
mainly due to the absence of the technology shock. The increase in public
sector employment is steady and slightly greater than in the previous case.
Interestingly, at the end of the exercise, unemployment is still above its ini-
tial level.
Despite the fact that wages grow at a slower pace when we do not con-
sider a permanent change in technology, the increase in wages combined
with the severe decrease in private sector vacancies caused by the perma-
nent increase of public wages and vacancies has a greater crowding out effect
in comparison to the case where technology increases permanently.
27
Negative TFP Shock : Employment Stimulus Plan Vs No Plan, Unemployment
Period
In the above figure, and following a negative TFP shock, the level of
unemployment is higher in the case where the government decides to adopt
an employment stimulus plan i.e. when it increases public sector wages and
vacancies. This result is very interesting, it shows that following a nega-
tive TFP shock, employment stimulus plans are ineffective, they actually do
more harm than good by further increasing unemployment and decreasing
by that the aggregate consumption. A negative TFP shock increases un-
employment as in the standard literature and as seen in the figure above,
however, if the government decides to adopt an employment stimulus plan
by increasing public sector vacancies and wages it would magnify the ef-
fect due to the increase in the value of being employed in the public sector
and thus the share of unemployed searching for a public sector job, this
would put pressure on the bargaining power and thus raise private wages
also the increase in the share of unemployed searching for a public sector
job increases the marginal productivity of the match raising by that the
private wage. However, as unemployment increases due to the TFP shock,
and as the share of unemployed searching for a public sector job increases
the crowding out effect would be very speaking and would raise unemploy-
ment significantly as seen above (the unemployment rate is at 25% after 20
quarters in the employment stimulus plan scenario).
28
Transition of Unemployment Positive TFP VS Negative TFP shocks Transition of Unemployment: The Effect of a Negative TFP Shock
Period Period
Transition of Unemployment Subject to Different Shocks Transition of Unemployment Subject to Different Shocks
Period Period
29
The left graph of the upper panel shows the impact of deploying a public-
sector-based employment stimulus program when the economy is hit by a
positive TFP shock versus when the economy is hit by a negative TFP shock.
Results show that despite the positive TFP shock if the government deploys
a stimulus program by raising public wages and posting more vacancies the
unemployment level would increase. However, and as shown above, this in-
crease is not as severe as in the case where the TFP shock is negative since a
positive TFP reduces the number of unemployed and thus absorbs some of
the crowding out effect caused by the increase in public wages and vacancies.
The right graph of the upper panel confirms that in the presence of a
public employment stimulus plan, a negative TFP shock increases unemploy-
ment via the magnifying effect; negative TFP shocks increase unemployment
and stimulus programs increase the share of unemployed searching in the
public sector, thus, further worsening the problem due to a magnified crowd-
ing out effect. This is not the case when there is no negative TFP shock,
despite the normal crowding out effect, the magnified crowding out effect is
not present.
The left graph of the lower panel shows that the impact of a positive TFP
shock depends on whether the government deploys an employment stimulus
plan or not. A positive TFP shock when not combined with an employment
stimulus program temporarily reduces unemployment due to the increase
in public and private employment it creates. However, when this positive
shock is combined with an employment stimulus plan, the unemployment
level increases. This shows that the TFP shock is not able to absorb the
crowding out effect caused by the increase in public wages and vacancies. On
the other hand, a permanent deterministic shock on technology, when not
combined with a stimulus plan, takes unemployment to a lower steady-state
level, this is not the case for a temporary TFP shock where unemployment
rebounds. However, technological progress rarely happens in this form, it
usually takes the form of a temporary shock as a TFP shock.
The right graph of the lower panel shows that when a permanent increase
in the steady-state level of technology is combined with an employment stim-
ulus plan, unemployment presents a hump-shape increasing at first then de-
creases as it converges to a new steady-state with lower unemployment rate.
However, following a deterministic permanent shock on technology, unem-
ployment rate converges quickly to a lower steady-state level without any
increase in unemployment as in the case of a stimulus plan. In the graph,
we also see that, without a permanent increase in the steady-state level of
technology, adopting an employment stimulus plan would have catastrophic
effects raising unemployment dramatically to a new steady state around
21%. This is due to the crowding out effect caused by the stimulus plan
30
that is not offset by a positive permanent increase in technology that ab-
sorbs part of the crowding out effect by lowering unemployment. Finally, a
deterministic permanent increase in the steady-state level of both public va-
cancies and wages causes crowding out effect and increases unemployment.
However, most of the crowding out happens due to the public sector wage
increase, since, as seen in the graph an increase in steady-state public wages
increases unemployment more sharply than an increase in the steady-state
number of public vacancies.
31
6 Conclusion and Discussion
The paper was motivated by the newly issued 2017 salary scale that concerns
all public sector employees in Lebanon. This is done by using a two-sector
search and matching model in a deterministic framework.
32
gram known as the Civil Service Council. This approach could be the cause
behind the very low public sector matching efficiency. Further research is
needed in order to better understand the impact of the sectarian approach
on the Lebanese labour market.
33
References
Afonso, A., Gomes, P. (2014). Interactions between private and public
sector wages. Journal of Macroeconomics, 39 , 97–112.
Algan, Y., Cahuc, P., Zylberberg, A. (2002). Public employment and labour
market performance. Economic Policy, 17 (34), 7–66.
Ardagna, S. (2007). Fiscal policy in unionized labor markets. Journal of
Economic Dynamics and Control , 31 (5), 1498–1534.
Blanchard, O. J., Diamond, P., Hall, R. E., Murphy, K. (1990). The
cyclical behavior of the gross flows of us workers. Brookings Papers on
Economic Activity, 1990 (2), 85–155.
Blank, R. M. (1985). An analysis of workers’ choice between employment
in the public and private sectors. ILR Review , 38 (2), 211–224.
Burda, M., Wyplosz, C. (1994). Gross worker and job flows in europe.
European economic review , 38 (6), 1287–1315.
Burdett, K. (2012). Towards a theory of the labor market with a public
sector. Labour Economics, 19 (1), 68–75.
Diamond, P. A. (1982). Aggregate demand management in search equilib-
rium. Journal of political Economy, 90 (5), 881–894.
Finn, M. G. (1998). Cyclical effects of government’s employment and goods
purchases. International Economic Review , 635–657.
Gohlke-Rouhayem, J., Melki, N., Weinmann, C. D. (2016). Employment
and labour market analysis (elma). Lebanon: GIZ.
Gomes, P. (2014). Optimal public sector wages. The Economic Journal ,
125 (587), 1425–1451.
Hansen, G. D. (1985). Indivisible labor and the business cycle. Journal of
monetary Economics, 16 (3), 309–327.
Heitmueller, A. (2006). Public-private sector pay differentials in a devolved
scotland. Journal of Applied Economics, 9 (2), 295–323.
Ireland, P. N. (2001). Technology shocks and the business cycle: On empir-
ical investigation. Journal of Economic Dynamics and Control , 25 (5),
703–719.
Jaoude, H. (2015). Labour market and employment policy in lebanon. Online:
European Training Foundation (ETF).
Merz, M. (1995). Search in the labor market and the real business cycle.
Journal of monetary Economics, 36 (2), 269–300.
Michaillat, P. (2014). A theory of countercyclical government multiplier.
American Economic Journal: Macroeconomics, 6 (1), 190–217.
Mortensen, D. T. (1982). The matching process as a noncooperative bar-
gaining game. In The economics of information and uncertainty (pp.
233–258). University of Chicago Press.
Pappa, E. (2009). The effects of fiscal shocks on employment and the real
wage. International Economic Review , 50 (1), 217–244.
34
Petrongolo, B., Pissarides, C. A. (2001). Looking into the black box:
A survey of the matching function. Journal of Economic literature,
39 (2), 390–431.
Pissarides, R., Jackman, C., Savouri, S. (1990). Labour market policies
and unemployment in the oecd. Economic policy, 5 (11), 449–490.
Quadrini, V., Trigari, A. (2007). Public employment and the business cycle.
Scandinavian Journal of Economics, 109 (4), 723–742.
Robalino, D., Sayed, H. (2012). Republic of lebanon: Good jobs needed.
Middle East.
Yaacoub, N., Badre, L. (2011). The labour market in lebanon. Statistics
in Focus(1).
35
7 Appendices
7.1 Steady State Tax Level Variation
Unconditional Job Finding Rate ConditionalJob Finding Rate Probability of Filling a Vacancy
Steady State Tax Level Steady State Tax Level Steady State Tax Level
Steady State Tax Level Steady State Tax Level Steady State Tax Level
Number of Unemployed Searching in Both Sectors Share of Unemployed Searching for a Public Sector Job Aggregate Consumption and Production
Steady State Tax Level Steady State Tax Level Steady State Tax Level
Steady State Tax Level Steady State Tax Level Steady State Tax Level
36
7.2 Steady State Public Sector Wage Premium Variation
Unconditional Job Finding Rate ConditionalJob Finding Rate Probability of Filling a Vacancy
Steady State Public Sector Wage Premium Level Steady State Public Sector Wage Premium Level Steady State Public Sector Wage Premium Level
Steady State Public Sector Wage Premium Level Steady State Public Sector Wage Premium Level Steady State Public Sector Wage Premium Level
Number of Unemployed Searching in Both Sectors Share of Unemployed Searching for a Public Sector Job Aggregate Consumption and Production
Steady State Public Sector Wage Premium Level Steady State Public Sector Wage Premium Level Steady State Public Sector Wage Premium Level
Steady State Public Sector Wage Premium Level Steady State Public Sector Wage Premium Level Steady State Public Sector Wage Premium Level
37
7.3 Steady State Public Sector Matching Efficiency Varia-
tion
Unconditional Job Finding Rate ConditionalJob Finding Rate Probability of Filling a Vacancy
Steady State Public Sector Matching Efficiency Level Steady State Public Sector Matching Efficiency Level Steady State Public Sector Matching Efficiency Level
Steady State Public Sector Matching Efficiency Level Steady State Public Sector Matching Efficiency Level Steady State Public Sector Matching Efficiency Level
Number of Unemployed Searching in Both Sectors Share of Unemployed Searching for a Public Sector Job Aggregate Consumption and Production
Steady State Public Sector Matching Efficiency Level Steady State Public Sector Matching Efficiency Level Steady State Public Sector Matching Efficiency Level
Steady State Public Sector Matching Efficiency Level Steady State Public Sector Matching Efficiency Level Steady State Public Sector Matching Efficiency Level
38
7.4 Steady State Bargaining Power Variation
Unconditional Job Finding Rate ConditionalJob Finding Rate Probability of Filling a Vacancy
Steady State Bargaining Power Level Steady State Bargaining Power Level Steady State Bargaining Power Level
Steady State Bargaining Power Level Steady State Bargaining Power Level Steady State Bargaining Power Level
Number of Unemployed Searching in Both Sectors Share of Unemployed Searching for a Public Sector Job Aggregate Consumption and Production
Steady State Bargaining Power Level Steady State Bargaining Power Level Steady State Bargaining Power Level
Steady State Bargaining Power Level Steady State Tax Level Steady State Tax Level
39
7.5 Transition Path of the Economy in Response to a Per-
manent Deterministic Shock on Technology, Public Em-
ployment and Public Wages
40
7.6 Transition Path of the Economy in Response to a Per-
manent Deterministic Shock on Public Employment and
Public Wages
41
7.7 Transition Path of the Economy in Response to a Per-
manent Deterministic Shock on Public Wages
42
7.8 Transition Path of the Economy in Response to a Per-
manent Deterministic Shock on Public Vacancies
43
7.9 Transition Path of the Economy in Response to a Per-
manent Deterministic Shock on Technology
44
7.10 Transition Path of the Economy in Response to a De-
terministic Temporary Shock on the Total Factor Pro-
ductivity
45
7.11 Transition Path of the Economy in Response to a Tem-
porary Deterministic Shock on the Total Factor Pro-
ductivity Augmented by a Permanent Deterministic
Shock on Public Vacancies and Wages
46
7.12 Transition Path of Unemployment in Response to Dif-
ferent Shocks
Transition of Unemployment Positive TFP VS Negative TFP shocks Transition of Unemployment Subject to Different Shocks
th Period Period
Period
47
7.13 Transition Path of Private Sector Employment in Re-
sponse to Different Shocks
Transition of Private Employment Positive TFP VS Negative TFP shocks Transition of Private Employment Subject to Different Shocks
th Period Period
Period
48
7.14 Transition Path of Public Sector Employment in Re-
sponse to Different Shocks
Transition of Public Employment Positive TFP VS Negative TFP shocks Transition of Public Employment Subject to Different Shocks
th Period Period
Period
49
7.15 Transition Path of Private and Public Employment as
well as Unemployment in Response to a Negative TFP
Shock
Negative TFP Shock : Employment Stimulus Plan Vs No Plan, Public Emp Negative TFP Shock : Employment Stimulus Plan Vs No Plan, Private Emp
th Period Period
Period
50