To Invest Or Not To Invest? The Case for Proactive Government Spending on Education in Bangladesh

It is largely expected that investment in education entails significant returns for the individuals (Appleton, 2000; Psacharopoulos and Patrions, 2002). Hence, debates over substantial allocation of budget to education remains crucial. Essentially, government investment on education in LDCs see greater and quicker returns to education than OECDs (Montenegro and Patrinos, 2013; Sianesi and Reenen, 2000). However, studies from international data persuasively argue that a “resource-based” policy of government has little effect on the education attainment of pupils (Hanushek, 2003; Blankenau and Camera, 2009). Basu and Bhattarai (2012), extended on the works of Blankenau and Simpson (2004) and Blankenau, Simpson, and Tomljanovich (2007), proposing that government spending alone cannot enhance returns to individuals and economy, rather it extends to how “proactive” the government is in the education sector.

Bangladesh remains synonymous in achieving the Millennium Development Goals related to education; however, horizontal access to education cannot alone suffice for the evaluation of collective achievements, in the form of quality, effectiveness and relevance of the learning content on capacity building abilities of individuals. In guiding and directing policies, marginal than average returns to education play an enormous role in justifying the spending of an extra Bangladeshi Taka (BDT) on education. Estimating marginal return to any policy remains a central tool in economic cost-benefit analysis. Before, we assess the international returns to education and address the Minister’s decision, let us have a brief look at the international evidence on the returns to education (henceforth, ROE).

Figures 1, 2 and 3 serve their purpose in identifying the disparity between developed and developing nations. Figure 2, highlights the global education market, majority occupied by the primary and secondary schooling. The graphs serve their purpose in linking: a) Bangladesh is a developing economy and b) investment in the education market must include special attention to primary, secondary and post-secondary schooling. To serve the purpose of this review, we will evaluate Bangladesh’s ROE to India and Pakistan. With history that dates these three nations together, after independence there has been significant changes in their economies. This country analysis of their ROE will essentially help us to understand Bangladesh’s current education scenario.

Calculation of ROE derives from the works of Jacob Mincer, in his book Schooling, Experience and Earnings (Mincer, 1974). Mincer regressed the natural logarithm of earnings as a function of years of education, and years of potential labour market experience. The most widely used model, called the Mincer’s ‘Human Capital Earnings Function’, is given below:

Where y is earnings, is earnings of an individual with no education and experience, S is years of schooling and X is years of potential labour market experience.

Kingdon (2007) evaluates India’s ROE in an international context in terms of school access and quality, role of private education and government spending. The study identifies that the wage-education curve is convex, with greater returns for secondary and post-secondary than primary. However, in 2002 secondary schools comprised of only one-fifth of primary; creating a supply-side constraint. Duraisamy (2002), estimated the ROE in India by age-cohort, gender and location, finding that greater returns are seen till secondary schooling and diminishes afterwards; there exists considerable gender and rural-urban differences. Notably, the returns were greater for technical diplomas in the rural than urban. Ethnicity, is seen to play roles in the human capital accumulation. While India possesses the caste system, parents’ social stratification is also seen to affect the ROE (Deshpande 2011). Socio-economic status of family is seen to have negative effect on the participation of woman in the labour market in Pakistan (Shah, 1986). Khan and Irfan (1985) estimated ROE on different levels of education in Pakistan, only to find that the returns for all levels are comparatively lower than other developing countries. Afzal (2011), finds results consistent with Khan and Irfan, establishing that the ROE market in Pakistan is worse than Bangladesh, India, and Sri Lanka and identifies that the performance and investment system is discouraging.

While the Mincer equation remains influential, it is seen to be consistent with data in 1980s and 1990s onwards compared to 1960s and 1970 (Lemieux, 2003). The equation leaves out variability of different returns at different schooling levels (Zhang, 2011). Considering the literature on ROE from India and Pakistan, it is evident that socio-economic characteristic of families play enormous role. In an extensive study, Pete and Fink (2014) find a 6.5% pooled return, from 61 low to middle income countries, lower in rural areas, higher returns for females and highest returns were seen from the secondary and post-secondary levels. While governments of developing countries aggressively invest in education, literature suggests the economic returns is rather weak (Mingat and Tan, 1992 and 1998; Landau, 1986; Noss, 1991 and Flug, Spilimbergo and Wachtenheim, 1998). Basu and Bhattarai (2012), model a calibrated parameter of cross-country evidence, government bias in education, depicting that greater spending of government reduces long-run per capita growth of a country. Their analysis imply that government must be proactive in the ROE, by investing sufficiently in infrastructure, vocational and professional schooling. It is further seen that 80% of education investment is towards teachers’ salaries, which is likely to be spent on consumption than investment (UNESCO, 2010).

The socio-economic factors of households limiting the ROE in India, Pakistan and Bangladesh are similar; hence it is important for Bangladesh to plan its educational investment proactively. If we were to evaluate literature on ROE of Bangladesh, Asadullah (2006) concludes the labour market returns to be 7% using data from HIES 1999-2000. A notable finding was the non-linearity in ROE in Bangladesh, where returns increased across levels of education. With primary schooling having the lowest returns, it does not essentially mean that investment in primary schooling is inefficient. Shafiq (2007), estimates ROE of boys in rural Bangladeshi households, where returns are 13.5%, 7.8%, 12.9% for primary, secondary and post-secondary levels and the resulting option value from primary levels is 5.3% indicating that non-poor rural households are likely to invest in boys’ education. Current agreement in development economics emphasizes on women’s education an investment with an enormous pay off. Like India and Pakistan, Bangladesh too faces issues of gender parities; though government continually initiates policies, changing the perspective and social traits of households remain a concern to ROE. An influential study by Kuenning and Amin (2001) highlights that rural households are more interested in investing in the marriage capital than human capital of females. If we were neglect these households, the ROE in the labour market will remain biased towards men.

For a country torn between severe gender differences and household’s illiteracy at the rural level, a proactive government spending plan is essential. The following infographic acts as a plan for the government to address the non-technical bodies. Table 1, dissects the spending allocation between primary, secondary and post-secondary.

The debate remains that developing countries do not have the necessary knowledge, policy tools and evaluative measures to enhance social welfare. Bangladesh, infamous for corruption, ideally limits equal opportunities for the disadvantaged. While GDP, RMG exporting and ICT developments enhance the lives of many, issues of female-participation, quality of education and employability remains camouflaged. While ROE is long-term, it can help households to deal with risk and increase their household wealth in terms of wages/salaries. Thus, reducing both transient and chronic poverty to an extent. While evidences on ROE suffer from endogeneity, surveys in developing countries may often leave out important variables of interest such as years of education, pupil-teacher ratio. Hence, OLS estimates may be upward or downward biased. The use of pseudo-panels to calculate ROE in developing countries may be an approach to improvement (Jamal, 2015; Himaz and Aturupane, 2016; Warunsiri and McNown, 2010). This review proposes an actor-oriented approach for public spending on education. Reducing government spending based on LFS 2010 or aggregate cross country calculations reject to take current scenario and country-heterogeneity into account.