میکروبنیادهایی برای افزایش بازده در انباشت سرمایه انسانی و تله تحت مشارکت

This paper considers educational investment, wages and hours of market work in an imperfectly competitive labour market with heterogeneous workers and home production. It investigates the degree to which there might be both underemployment in the labour market and underinvestment in education. A central insight is that the ex post participation decision of workers endogeneously generates increasing marginal returns to education. Although equilibrium implies underinvestment in education, optimal policy is not to subsidise education. Instead it is to subsidise labour market participation which we argue might be efficiently targeted as state-provided childcare support.

This paper considers educational investment, wages and hours of market work in an imperfectly competitive labour market with heterogeneous workers and home production. It investigates the degree to which there might be both underemployment in the labour market and underinvestment in education. A central insight is that the ex post participation decision of workers endogeneously generates increasing marginal returns to education. This non-convexity can result in a large discontinuity in educational choice and labour market participation across workers. The paper shows that for some workers, a competitive labour market would imply they invest significantly in education and participate with a high probability in the labour market. But wages below marginal product (in a non-competitive labour market) and increasing returns to education together imply a non-marginal switch to low educational investment and home production. These large substitution effects yield large welfare losses and so corrective taxation plays an important role. Although there is underinvestment in education, optimal policy is not to subsidise education. Instead it is to subsidise labour market participation, which we argue might be efficiently targeted as state-provided childcare support. The paper considers a hold-up problem where in the first phase of their lives, youngsters increase their future workplace ability by investing in general skills. Those investments are made prior to becoming employed in the workplace. Clearly some skill investments, such as primary school education in literacy and numeracy skills, are invaluable both at home and in the workplace. But the focus here is on educational choice past the compulsory school level, by which time literacy and numeracy skills have presumably been well honed. Instead students might further invest in a university degree in mathematics or a qualification in information technology, imbuing them with expertise that is valuable in the workplace but is unlikely to increase their skills in the home. A central feature of the model is that there are increasing marginal returns to education. We stress that these increasing returns do not arise because we assume a Mincer wage equation with increasing returns. Indeed the arguments are consistent with a Mincerian wage rate w=a+ew=a+e, where the wage rate ww depends on endowed ability a and is linearly increasing in education ee. But such a wage equation does not describe the marginal return to education. For example, the person who intends to specialise entirely in child rearing and home-making has a zero financial return to investing in workplace skills, regardless of the size of the Mincer wage effect. The marginal return to education depends both on the Mincer wage effect and expected labour supply, where increased labour supply implies human capital investments are “used” more intensively in the workplace. We shall show there are three reasons for increasing returns to education. First, there are increasing marginal returns to education because of a participation effect. More highly skilled workers earn higher wages in the workplace and so are more likely to participate in the workplace, thereby raising the ex ante expected returns to human capital investment. Second, increasing returns arise through an increasing labour supply effect, where more educated workers may find it worthwhile to work longer hours. But with a frictional labour market there is a third reason for increasing marginal returns to education—an increasing wage competitiveness effect. We show that firms bid more competitively for the worker's services as the value of employment increases. As wage compression decreases at higher productivity levels, the marginal return to education increases as education increases. A second important feature of the paper is that it assumes workers have different productivities both at home and in the workplace. We introduce this assumption not only because it is realistic, although that is clearly an advantage.1 But more importantly, it allows us to demonstrate how expected home productivity affects optimal educational choice and labour supply, where home and workplace productivities vary across individuals. Specifically we show that the deadweight losses that arise through an imperfectly competitive labour market are not equally spread across all workers. Increasing returns to education coupled with an imperfectly competitive labour market generates an “under-participation trap”. If the labour market were competitive, then workers in that trap would choose a high level of education and high expected labour supply in the workplace. But because the labour market is not competitive and so wages paid are below marginal product, they substitute instead to home production. The increasing returns to education, however, imply the substitution effect is non-marginal for workers in this “trap”. Instead they make very low skill investments ex ante, and participate with low probability in the labour market ex post. This large substitution effect implies a correspondingly large deadweight loss. The next section describes the model and Section 3 determines equilibrium remuneration and participation rates of workers by productivity type. Section 4 examines the worker's optimal investment decision and Section 5 develops the implications for optimal childcare policies. We establish that a participation subsidy, paid to the worker, not only corrects the ex post under-participation problem, but also corrects the ex ante under-education problem.

It is surprising that the increasing returns argument presented here has received no attention in the profession. Possibly it has been missed as there are decreasing marginal returns to labour supply and, given labour supply, there are also decreasing returns to education. Of course this does not imply a concave programming problem as there are joint increasing returns. When decisions are sequential, as in the hold-up problem considered here, these joint increasing returns generate increasing marginal returns to education in the first period. We have shown that, in an imperfectly competitive labour market, increasing returns to education generate an under-participation trap. Optimal corrective policy is an employment subsidy, which we argue might be efficiently targeted as a public childcare program. A popular alternative model of an imperfectly competitive labour market assumes instead search frictions and that wages are determined by Nash bargaining. In particular given (α,b)(α,b) and free entry of firms, the axiomatic Nash bargaining approach would imply the firm negotiates profit ππ and labour supply l as View the MathML sourcemaxπ,l[π]1-γ[αl-π+bx(1-l)-bx(1)]γ, Turn MathJax on where γ∈[0,1]γ∈[0,1] is the worker's bargaining power, bx(1)bx(1) is the worker's threatpoint [i.e. the value of home production] and the firm's threatpoint is zero in a free entry equilibrium. By definition of V in the text, this reduces to View the MathML sourcemaxπ[π]1-γ[V-π]γ Turn MathJax on and Nash bargaining implies worker remuneration y*y* satisfies dy*/dV=γdy*/dV=γ. As in Claim 1, this implies equilibrium wage compression and so one would anticipate the same effects on education and participation as discussed here. But there are two main advantages to the Bhaskar and To (1999) approach. One is that it rules out search externalities, such as thick market and congestion externalities, which would otherwise complicate the policy discussion. It also does not require solving for the steady-state distribution of job seeker productivities which, in equilibrium, affects the vacancy creation decision of firms. The simpler approach shows clearly that under-participation and wage compression generate mutually reinforcing distortions on human capital investment: Wage compression implies workers tend to underinvest in workplace skills, and lower skills imply a lower participation probability which further reduces the expected return to human capital accumulation. In currently ongoing work we examine how increasing returns to education interact with various other market distortions such as (i) endogenous household formation with matching frictions in the marriage market (Booth and Coles, 2005); and (ii) government tax policy, where increasing returns to education causes large substitution effects, and hence large deadweight losses, around the non-participant margin (Booth and Coles, 2006). In a third paper, we show that increasing returns to education arise even in a perfectly competitive labour market (Booth et al., 2006) and identify these effects empirically using individual-level data.