Conceptual review and meta-analysis of school effectiveness

Production function studies in developing countries

Hanushek (1995) provides the following tabulation of the effects of resources in 69 studies in developing countries (see table 8).

Table 8: Summary of ninety-six studies on the estimated effects of resources on education in developing countries, cited from Hanushek, 1995

		Statistically significant		Statistically insignificant
Input	Number of studies	Positive	Negative
Teacher-pupil ratio	30	8	8	14
Teacher’s education	63	35	2	26
Teacher’s experience	46	16	2	28
Teacher’s salary	13	4	2	7
Expenditure per pupil	12	6	0	6
Facilities	34	22	3	9

When the number of positive significant associations are expressed in percentages the comparison depicted in table 9 with the results shown in table 5 concerning studies in industrialized countries can be made in a more straightforward way.
Table 9: Percentages of studies with positive significant associations of resource input variables and achievement for industrialized as compared to developing countries (sources: Hanushek, 1995, 1997)

Input	Industrialized countries % sign. positive associations	developing countries % sign. positive associations
Teacher/pupil ratio Teacher’s education Teacher’s experience Teacher’s salary Per pupil expenditure	15% 9% 29% 20% 27%	27% 55% 35% 30% 50%

The relevance of facilities in education in developing countries, not shown in the comparison, amounts to no less than 70 when expressed as the percentage of significant positive studies.
The larger impact of these resource input factors in developing countries can be attributed to larger variance in the independent as in the dependent variables. Both human and material resources in education in industrialized countries are distributed in a relatively homogeneous way among schools, in other words: schools do not differ that much on these variables. Regarding the outcome variables (e.g. educational achievement) Riddell (1997) has shown that schools in developing countries vary on average 40% (raw scores) and 30% (scores adjusted for intake variables). This is a considerably larger variation than is usually found in industrialized countries; where values of 10% to 15% between school variance on adjusted outcomes are more common (cf. Bosker & Scheerens, 1999).

The positive outcomes of production function studies in developing countries make intuitive sense (if basic resources and facilities are not present this will obviously be detrimental to the educational endeavor as a whole). At the same time the outcomes give rise to interesting interpretations when they are brought to bear on the theoretical principles of micro-economic theory. Jimenez & Paquea (1996), for example, present findings that support the thesis that local involvement in school finance stimulate both achievement orientation as economy in spending. Pritchett and Filmer (1997) point at the political advantages of spending on human resources (diminishing class size in particular) as compared to spending on instructional materials, despite the much larger efficiency of the latter approach, while Picciotto (1996) criticizes the narrow set of educational performance criteria that is used in most education production function research and states that “program design must be informed by assessments of overall educational performance against societal objectives; by evaluations of the relevance of the objectives themselves and by judicious design of institutions to deliver the needed services” (ibid, 5). Micro-economic theory has interesting conjectures with respect to control mechanisms in education as well; where the argument is that bureaucratic control measures are expensive and faulty and community involvement and “direct democracy” would present a better alternative. Particularly when studies are becoming more theory-driven and cost-benefit analyses are more frequently included, production function research is to be considered as a viable approach to school effectiveness studies in developing countries.