Research Article | Open Access
Optimal Educational Investment: Domestic Equity and International Competition
We construct a family of models to analyse the effect on optimal educational investment of (i) society's preferences for equity and (ii) competition between countries. The models provide insights about the impact of a variety of parameters on optimal policy. In particular, we identify a form of “overeducation” that is new to the literature and provide a counterexample to a common finding in the literature on fiscal federalism.
Economists’ interest in education often focuses upon the rate of return to schooling investments. Yet it is a characteristic of the education system in many countries that, for the most part, schooling is funded out of the public purse. While one might imagine that governments should seek to “equalise rates of return in all directions” , it is often the case that the authorities have broader objectives that inform their educational investments. For example, a government may have preferences about equity as well as efficiency. Or it may, for various reasons, be concerned to ensure that its own investment in its people’s skills does not fall behind investments made by other countries.
We examine these issues by developing, in the next section, a series of models that can aid our understanding of how, under a variety of conditions, the optimal provision of publicly funded education is determined.
2. The Model
In this section we present a family of related models of education and the tax system in order to provide insights into how governments can reach decisions about the optimal funding of education where (i) society has preferences about equity and (ii) decisions have impacts across countries. The basic structure of the model builds on the analysis of Johnes .
Suppose that the disposable income of individual is given by where is a basic income to be defined more precisely later, is a binary variable that indicates whether the th individual has undertaken schooling or not, is the proportional rate of income tax, and is the income premium associated with schooling. Both and are assumed exogenous. Tax revenues are used solely for the purpose of financing education which, we assume, takes place instantaneously. This distinguishes the model from a family of models typified by that of Bovenberg and Jacobs , where taxation also serves a redistributive purpose. In our model we keep the tax system simple in order to facilitate the extension to the international case in Section 2.2 below. In the present model, tax revenues fund education as a means of achieving the redistribution of income, but they could equally fund any investment that achieves this goal. Investments in health, for example, could equally be modelled in this way. For simplicity we consider only one public good.
Denote by the proportion of the population that undertakes education. Total tax revenue is given by
Suppose that the cost of providing schooling to each individual is an increasing function of , and is, more precisely, given by . The total cost of education is then , and this must equal the expression in (2) in order for the exchequer’s books to balance. Solving for , which must lie within the unit interval, and assuming a unique real root yields
The sum of disposable incomes is given by
To close the model, we introduce a social welfare function, maximisation of which yields solutions for the optimal tax rate and the optimal level of education. We begin with a particularly simple variant of the model in which social welfare equals and where represents a weight attached to the premium earned by higher income (educated) individuals. In this way, society expresses its preferences concerning the income distribution.
It is possible, though tedious, to derive an analytical solution for the problem of maximising (6) with respect to . We denote this solution by and note that routine substitution of this into (3) yields the optimal level of education, . Clearly
Since the analytical solutions for and are cumbersome and uninstructive, we proceed by way of numerical examples. In Table 1, we show the values of and that obtain for a variety of assumed values of . These are shown for various values of , , and . In the upper panel, we have , , and , while in the lower panel we have , , and . The lower panel therefore represents a state in which returns to education are lower, and costs of education are higher, than in the upper panel.
|Note. Throughout it is assumed that .|
It is readily observed that investment in education and, consequently, also tax rates are lower in the lower panel than in the upper panel. This follows directly from the fact that returns to education are lower in the lower panel—with both the earnings premium to educated workers being lower and the cost of education being higher. It is also clear that investment in education, and tax rates, fall as society places more weight on equity. This result would, of course, need to be qualified in the case of a model where taxes also serve a redistribution role. Raising educational investment offers greater return in a society where the incomes of the educated workers carry more weight.
2.2. International Issues
The second variant of the model that we examine is chosen to provide insights into international issues. In order to build in some interaction between the two countries, we assume a decreasing function of global education levels. This is to reflect the labour market impact on one country of the educational investments made by another country, through changes, for example, in comparative advantage. Some references in the received literature (e.g., ) suggest that, owing to complementarity effects, the impact of education in one country on the rate of return to education in another should be positive. Here we are adopting the alternative view that rising skill levels in one economy worsens the competitive position of workers in another, following, for example, Freeman . Our model assumes no migration across countries. Hence assume that where and , respectively, denote the proportion of the population in each country that undertakes education, and where is a constant. For simplicity we assume that , , and are identical across countries.
Noting that the balanced budget constraint implies we may solve a pair of simultaneous cubic equations to establish the levels of and as
Define social welfare within each country, in analogous fashion to (5), as
To keep matters simple, suppose . The maximisation of with respect to can proceed either with the two countries competing with one another, following Nash , or with them playing cooperatively. In neither case is there a straightforward analytical solution, so we proceed by way of example. Results are shown for a variety of parameter assumptions in Table 2. By symmetry, each country has the same optimal tax rate and education level as the other in both the Nash and the cooperative case.
|Notes. Throughout it is assumed that , , , and . The value of welfare is reported as a per capita measure.|
The results indicate that the Nash solution implies higher tax and education levels than the cooperative solution. The intuition behind this result is straightforward. Starting from a cooperative position, each country, taking the other’s behaviour as given, has an incentive to raise its own investment in education. Consequently, Nash behaviour leads to a type of “overeducation” that is new to the literature. In contrast with the overeducation identified by authors such as Daly et al.  and Dolton and Vignoles , where some graduates fail to find work commensurate with their qualifications, the overinvestment in education that we observe in the present model represents a shortfall in welfare due to competition between countries.
The results reported here provide a striking contrast to a finding that is common in the fiscal federalism literature—namely that competition between tax jurisdictions leads to lower tax rates . When, as in this model, the tax is spent on activity that is welfare enhancing, competition can have the opposite effect.
In the above analysis a positive value has been assumed for . It is conceivable, however, that , indicating complementarity of human capital stocks across countries. This could lead to undereducation as each country fails to take into account the positive externalities associated with its education investments.
2.3. Equity in the International Model
Extension of the model of the previous section to include values of is straightforward, requiring no change to (8) through (13), and only a minor change in the programming. Results for a variety of parameter assumptions appear in Table 3. These results follow the patterns identified in the earlier sections of the paper and hence do not require extensive discussion here. As the returns to education, here measured by , increase, so does the optimal level of educational investment in either the Nash or the cooperative model, other things being equal. Likewise as rises, indicating weaker preferences for equity, so the optimal level of educational investment increases. It is worth noting, moreover, that as rises, the gap between the Nash and cooperative equilibria tends to widen.
|Note. See notes to Table 2.|
The notion that competition between countries leads to the setting of tax rates that differ from those that would obtain in the absence of such competition is a familiar one. In this paper, we have extended this to examine the international competition in tax and government expenditure, where the expenditure takes the form of educational investments that in themselves yield gains in the form of enhanced income. We have also examined the operation of the model in the context of alternative societal preferences for equity.
The type of overeducation identified in this paper is new to the literature, and it is not at all clear how extensive this effect might be in practice, where not all of the restrictions of the present theoretical model apply. An interesting avenue for future research might therefore be to evaluate this effect. It is suggested that multicountry computable general equilibrium models could prove to be a useful tool in this endeavour.
Without implication, the author thanks Jaideep Roy and anonymous referees for helpful comments on an earlier draft.
- M. Blaug, L. Richard, and W. Maureen, “1969,” in The Causes of Graduate Unemployment in India, Penguin, Harmondsworth, UK.
- G. Johnes, “The evaluation of welfare under alternative models of higher education finance,” in Markets in Higher Education: Rhetoric or Reality? P. Teixeira, B. Jongbloed, D. Dill, and A. Amaral, Eds., Kluwer, London, UK, 2004.
- A. L. Bovenberg and B. Jacobs, “Redistribution and education subsidies are Siamese twins,” Journal of Public Economics, vol. 89, no. 11-12, pp. 2005–2035, 2005.
- D. Acemoglu and J. Angrist, “How large are human capital externalities? Evidence from compulsory schooling laws,” in NBER Macroeconomics Annual 2000, B. Bernanke and K. Rogoff, Eds., pp. 9–59, MIT Press, Cambridge, Mass, USA, 2001.
- R. B. Freeman, “Are your wages set in Beijing?” Journal of Economic Perspectives, vol. 9, no. 3, pp. 15–32, 1995.
- J. F. Nash, “Non-cooperative games,” Annals of Mathematics, vol. 54, pp. 286–295, 1951.
- M. C. Daly, F. Büchel, and G. J. Duncan, “Premiums and penalties for surplus and deficit education: evidence from the United States and Germany,” Economics of Education Review, vol. 19, no. 2, pp. 169–178, 2000.
- P. Dolton and A. Vignoles, “The incidence and effects of overeducation in the UK graduate labour market,” Economics of Education Review, vol. 19, no. 2, pp. 179–198, 2000.
- J. Edwards and M. Keen, “Tax competition and Leviathan,” European Economic Review, vol. 40, no. 1, pp. 113–134, 1996.
Copyright © 2012 Geraint Johnes. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.