تجزیه و تحلیل اقتصادی رفاه از سیاست های بازار کار در مدل هریس تودارو

This paper presents a welfare economic analysis of the benefits of various labor market policies in the Harris–Todaro labor market model. The policies considered are a policy of modern sector job creation, which I call modern sector enlargement (MSENL); a policy of rural development, which I call traditional sector enrichment (TSENR); and a policy of wage limitation in the urban economy, which I call modern sector wage restraint (MSWR). First, I analyze the inequality effects of these policies. I then perform two welfare economic analyses, the first based on summary measures of labor market conditions (total labor earnings, unemployment, inequality of labor incomes, and poverty rates) and the second based on dominance analysis in the labor market, in both cases assuming that the costs are borne elsewhere. The results of the welfare analyses are compared, and it is shown that TSENR unambiguously increases welfare in the labor market using both approaches, the other policies yield ambiguous results, and no policy is unambiguously welfare-decreasing.

Since its introduction in 1970, the Harris–Todaro (HT) model has become the workhorse for analyzing labor market policies in dualistic labor markets. In the intervening years, many aspects of the model have been studied including unemployment, development policies, tax and transfer policies, and many others; see Todaro and Smith (2003) for a review. However, one aspect of the model has not yet received thorough attention, and that is the welfare economics of labor market policies in an HT economy. The purpose of this study is to help fill that gap. The paper proceeds as follows. In Section 2, I present the original Harris–Todaro model and review prior analyses of the effects of various labor market policies in that model. Three labor market policies were considered by Harris and Todaro themselves. The first was a policy of modern sector job creation, which I call modern sector enlargement (MSENL). MSENL could come about by a tripartite agreement, as in Harris and Todaro; by a government-sponsored employment creation scheme, as many countries have done; or by technical change and/or capital accumulation in the modern sector, which can (but need not) shift the demand curve for modern sector labor outward. The second policy considered by Harris and Todaro was rural development; focusing on the labor market effects of such a policy, I label it traditional sector enrichment (TSENR). A third policy considered by Harris and Todaro was a policy of wage limitation in the urban economy; I call this modern sector wage restraint (MSWR).1 The review of past literature in Section 2 demonstrates that prior work has provided valuable lessons but is not yet complete in analyzing the benefits of the various policies. The rest of the paper develops new results. Section 3 completes the inequality analysis. Section 4 then performs a welfare economic analysis of labor market conditions based on an “abbreviated social welfare function” (“abbreviated,” because they are functions of variables which themselves are summary statistics of different aspects of the labor market).2 The labor market indicators used in the abbreviated social welfare function here are total labor earnings, unemployment, inequality of labor incomes, and poverty rates. Section 5 turns to dominance analysis in the labor market (Hadar and Russell, 1969, Saposnik, 1981 and Foster and Sen, 1997). Section 6 compares the welfare economic results for labor market changes using the different approaches. The main conclusions are highlighted in Section 7. Before proceeding, let me add a word about what the results of this paper imply and do not imply about policy. If a government is choosing between MSENL, TSENR, and MSWR policies, it needs to consider not only the consequences of putting the policy into effect, which is what is analyzed here, but also the costs of putting the policy into effect. MSENL and TSENR both require expenditures of resources to create more modern sector jobs or to achieve rural development, respectively (unless, that is, an outside donor provides the money as an outright grant to the country). On the other hand, MSWR is costless (economically if not politically); actually, it saves the government money if the government is itself a modern sector employer. The social costs of these policies need to be taken account of along with the social benefits before a policy recommendation can be offered.

This paper has analyzed the welfare economics of labor market policies in the Harris–Todaro model. Three policies considered by Harris and Todaro have been evaluated: MSENL, TSENR, and MSWR. The novelty of this paper has been to widen the terms of the analysis. Harris and Todaro themselves, and many others who followed them (myself included), analyzed policies in terms of their unemployment effects. More recent analysis been conducted in terms of inequality effects, and this paper has added to that strand of work. In addition, I have gone beyond these variables to also include poverty and total labor earnings within the abbreviated social welfare function framework. I have also evaluated the various labor market policies using welfare dominance methods applied to the labor market. When the labor market analysis is broadened beyond unemployment alone, Harris and Todaro's justifiably famous policy recommendations—avoid modern sector employment creation and instead seek rural development—are partially but not fully supported. Rural development would indeed produce better labor market outcomes on all of the dimensions considered (if, as in HT, the costs of rural development can be ignored). Modern sector employment creation is not unambiguously a bad thing, though. The increase in unemployment and inequality of labor earnings (putative “bads”) should be balanced against the increased number of high wage jobs and the consequent reduction in the poverty headcount ratio (putative “goods”). Modern sector wage restraint was also favored by Harris and Todaro. However, the results here show that such a policy does not unambiguously improve labor market outcomes. This is because the lowering of wages itself receives negative welfare weight and because inequality can rise if the demand for labor is sufficiently elastic. I have shown in this paper that even when the underlying labor market model is quite straightforward (which is my reason for working with the simplified version of the original HT model), the welfare evaluation of various policy alternatives depends critically on which welfare economic approach is adopted. Qualitatively different policy judgments are obtained depending on whether the welfare judgment is a function of unemployment alone, inequality alone, an abbreviated social welfare function, or first order welfare dominance. These different welfare economic approaches have produced qualitatively different results. Thus, in the HT model, it is not enough just to be concerned about distribution. How distributional concerns are brought into the policy analysis can and does make an important difference.