Transactional-based: the school is a buyer or seller. This may include promoting
the school, recruiting pupils, or buying suplles and services from external providers. It includes the traditional elements or marketing: identiflying and providing the curriculum demanded by the schools customers (parents, pupils, govemment, society). Relationship-bassed: the school works in partnership with those beyond the school boundaries parents of pupils, community organizations (churches, police etc.), special services providers (for example, education psychologists). Here the relationship is based on mutual profesional trust rather than competiton, buying and selling. Public accountability: the statutory management of accountability processes such as inspection, the pubclication of examination results and accounting for the resources allocated to the school. The priority of these processes has meant thaa school leaders must be increasingly fpcused on the external environment as well as the internal world of the school. Furthermore, understanding the interations between the three sorts of external relations and between internal and external processes is an important task in the planning and strategic dimensions of leadership roles. Scanning the external enviroment to identify necessary and desirable respoinses within the school is a key task, but so is the monitoring of internal processes and the prioritization of internal marketing activities to ensure that the perceptions of the school by all external stakeholders are favourable. Evidence of the prioritization of external relations management has emerged from a number of studles. In the context of England and Wales Gewlrtz, Balls and Brow (1995) show how senior managers in a wide range fo schools in London have bugen to consider the external relations dimensions of the tasks they lead. Foskett (1998) has shown, too, how a cultural shift has accured to some schools in which external rerations management has a higher status and roler; a view echoed in the studies og Thrupp (1999) and of Lauder and Hughes (1999) in New Zealand. Such shift are not confined to the highly marketized educational systems. In Pakistan (Memon et at., 2000) show how the management of relations with parents is a significant challenge for headteachers, despite the mainntenance of a strongly centralized system, and in Tanzania (Babyegyege, 2000) a key funtion for headteachers is the management of relations with village communities and parents to encourage support in terms of resources and also in the enrolment of pupils within the school. Finally, we need to conside the impact of marketization on learning and teaching and on pupil perfomence, which is one of the key raisons detre of the prioritization of markes and choice. How far has market policy changed leaoership, such that there is a measurable impact on pupils rather than a simple cosmetic change to the external appearance of schools? Meansuring the effects of specific policy actions in education is notoriously difficult, for three reasons. First, there needs to be the idenfication of appropriate indicators and measure, and agreement on their value and their application: What outcomes are we actually looking to identify as the intended effect of policy change? Second, system complexity means that linking specific ourtcomes to particular actions in policy or practice is rarely possible, for the of other factors cannot be isolated in the analysis. Third, the impact of change may take many years to emerge. Individual pupils spend 10-15 years in the education system., and the evidence of the impact of change can be only whwn cohorts of puplis have passed though the wholesystem under the new set of conditions. While earlier evidence may emerge. It is clear from out understanding of the ways in which innovations are taken up and implemented thet patterns of change in the early phases of cycles of development may not indacate the overall impact, patterns or outcomes (Roger, 1993). The time requirement for empirical evidence meant that much or the substantial literture on the impact of marketization on pupil performance in the first of the 1990s was entirely theoretical in nature, arguing the likely outcomes on the basis of assumed oprational models of distinct ideological standpoints. However, evidence is now beginning to emerge, and Caldwell (2000) has identified three generations of research studies examining the relationship between marketization, self- management in schools and pupil performence. The first generarion is the reasearch undertaken pricipally in the context of the USA in the early 1900s, and summarized by Summers and Johnson (1996). Drawing on a large number of studies, they suggest that three is no demonstrable link between marketization and the enhancement of pupil performance in schools. They emphasize, however, than the nature of the data available therough such studies does not make any testing or the markets-pupil achievement link easy. Furthermore, they stress that in the context of thr USA there is little explicit infent in the development of self-managing schools to emphasize the enhancement of pupil achievement, and that most of the experiments are rooted in aims that are primanly related to libertarian ideologies and school/teacher empowerment. The second generation studies are those undertaken principally in the mid-1990s in the context of significant shift towardsbself-managing schools in England and Wales, Australia and New Zealand. While such studies demonstrate increases incost efficiency in such schoold (e.g. Levacic, 1995, in the context of local management of schools in England). They are unable to show any significant impact of marketization on pupil standards. Drawing from range of studies in differing national contexts. Whitty, Power and Hlpin (1998) suggest that there is still insufficient evidence to show that schools operating in a regime of self-management enhance pupil achievement. Third generation studies are those that have emerged since 1998, and are founded in a number of important characteristics. First, they draw on evidance from system that have been in operation for periods of up to a decade and so have had time to begin to demonstrate impacts at a systemlevel. Second, the operation of perfomence measurement and accountability system such as pucblic examination performence and inspection reports has provided a significationt set or data to enable analysis ti be undertaken. Third, the development of sultable teachniques of data analysis has analysis in relation to the complex mulufactorial processes at work in such educational markets. The increased sophistication that such studies provide, though, indicates that although the relationship between self-management and enhanced pupil performance may exist, it does so only in some circumstances. At best, such markets provide the circumstances for pupils across the social and achievement spectrum to enhance performance. At worst, marketization generates the heat of competition and the accrued losses from declining collegiality while simply exacerbating contrasts between schools and enhancing polarization on the basis of socio-ecoriomic status.