مزایای غیر مستقیم بهبود زیرساخت در مورد بازار کار ناقص

We perform a welfare analysis of transport infrastructure improvements in the presence of an imperfect labor market, allowing for endogenous wages and involuntary unemployment. Efficiency wage setting is incorporated in a spatial two-region general equilibrium model, written as a welfare program. In our model, firms set wages above the market clearing level to create an inducement for employees not to shirk. The economy-wide effect (i.e. social welfare or total welfare) consists of the direct effect in transport market and the indirect effect in other markets. We conduct welfare analyses different from the conventional cost–benefit analysis, in which we distinguish between the direct welfare effect based on consumer surplus in the transport market and the indirect welfare effect through changes in regional unemployment, which is captured in the economy-wide effect. In our model, infrastructure improvement may increase unemployment in one region, but overall unemployment levels fall. The indirect welfare effects through decreases in overall unemployment are about 10–20% of the direct welfare effects for most plausible labor market parameters. The indirect welfare effects are larger, the poorer the initial transport infrastructure and the larger the labor market imperfections.

Cost–benefit analysis (CBA) is a standard tool for the evaluation of infrastructure projects. The analysis usually starts with the effects of such projects on generalized costs in transport markets. These changes in costs lead to changes in transport demand. The pertaining welfare effects are measured by changes in consumer surplus based on the demand for transport. An important question is whether such an approach does not overlook effects of changes in transport costs on other markets. A starting point to this discussion has been given by Jara-Diaz (1986) who demonstrated that in the absence of market imperfections on all relevant markets, the change in consumer surplus on the transport market gives an adequate measurement of the change in total welfare.1 In recent years, there has been a growing interest in the question on the extent to which changes in consumer surplus measured in the transport market would underestimate (or overestimate) the actual welfare changes in the case of market imperfections. This is called the surplus equivalence problem by Jara-Diaz (1986). For example, the SACTRA (1999) report compares total welfare increases of infrastructure improvements with changes in consumer surplus and finds that the former is about 30% higher than the latter; this analysis is based on a monopolistic competition model dealing with regional trade. Newbery (1998) finds indirect welfare effects of 3–8% in the context of oligopolistic competition in a spatial price equilibrium model. Davies (1999) obtains an indirect welfare effect of about 12% in a partial oligopolistic model. Bröcker and Schneider (2002) arrive at indirect welfare effects of about 20% by a Computable General Equilibrium (CGE) model for European regions. Oosterhaven and Elhorst (2003) obtain indirect welfare effects ranging from 20% to 80% in a CGE model for Dutch regions. In a monopolistic competition model with localization economies, Venables (2007) finds increases ranging between 60% and 150%. Pilegaard and Fosgerau (2008) find the welfare gains from transport improvements in the case of labor market search imperfection typically exceed those computed by the standard CBA by 30–50%, where the magnitude depends on different assumptions and parameters. These figures indicate that there are substantial differences between the various studies depending on the nature of the imperfection and the type of model used. The policy relevance of these studies is obvious. When market imperfections prevail, conventional CBA might underestimate total welfare change, this would lead to biases in decision-making on infrastructure projects or on infrastructure pricing. More in particular, when the gap between the two welfare measures would remain smaller than (say) 10% there would not be so much reason to worry, but for larger gaps there is a real risk that promising infrastructure plans would not be executed. The studies presented thus far show that there are ‘external benefits’ from an infrastructure improvement, that is, the social benefits are different from the transport benefits due to the monopoly power on product markets and agglomeration economies. However, imperfections on labor markets have not received much attention in dealing with the surplus equivalence problem. One exception is the study by Pilegaard and Fosgerau (2008), which focuses on commuting behavior. In this paper, we focus on transport infrastructure improvements that affect the transport costs of firms combined with labor market imperfections. In this context, transport infrastructure improvement affects the firms’ demand for labor and therefore regional (un)employment levels. It is important here to distinguish between the effects on voluntarily and involuntary unemployment. Indirect welfare effects of infrastructure improvements may only arise through changes in involuntary unemployment, whereas these effects through changes in voluntary unemployment are absent. Consider an economy where unemployment is involuntary and firms set wages and employment levels. In this case, firms will ignore the effect of their decisions on involuntary unemployment levels, so, at least according to theory, they set wages too high and employment levels too low. In other words, some workers are willing to work for a wage below the wage prevailing in the market. This is relevant for welfare analysis because it implies that transport projects might bring external benefits due to reductions in the involuntary unemployment level that are not captured in the demand function for transport. 2 We find that this is the case under several numerical examples and also under simplifying assumptions considered in the following section. The reason is that transport improves productivity of firms overall, which tends to increase employment levels in our model. To understand the indirect welfare effects of transport improvements through changes in labor demand and therefore involuntary unemployment, it is important to realize that unemployment cannot be studied independently from wages. A large number of empirical studies have shown that there is a negative relationship between the (regional) real wage and the unemployment rate, usually called the wage curve (Blanchflower and Oswald, 1994). In the labor economics literature, there are nowadays two important groups of models for studying wage setting mechanism or deriving wage functions given the presence of unemployment (see, e.g. Pissarides, 1998). The first group is the efficiency wage or ‘incentive wage’ or shirking models, where the firms set the wage to induce workers not to shirk.3 The second group is the wage bargaining models (Pissarides, 1985), where the wage is determined by bargaining between firms and employees (or unions). Both models have been widely used for studying unemployment issues in different contexts in the literature.4 For example, Blanchflower and Oswald (1994) apply an efficiency wage model to explain the empirical regularity of a wage curve with an elasticity of around −0.1. The main objective of this paper is to analyze and to distinguish between the direct and indirect welfare effects of a transport infrastructure improvement with labor market imperfections in a spatial (multi-regional) general equilibrium (SGE) model, because regional general equilibrium models provide a comprehensive framework for studying the regional effects of policies (see, e.g. Jones and Whalley, 1989). For this purpose, we develop a SGE model which includes an imperfect labor market with wage setting by firms and involuntary unemployment. To be specific, we employ a shirking model, introduced by Shapiro and Stiglitz (1984) in a general equilibrium framework, as the micro foundation of the efficiency wage model based on shirking is regarded as self-evident ( Calvo, 1979). 5 The shirking model, in which firms set wages to maximize profits and workers decide on-the-job effort levels (to shirk or not to shirk) to maximize utility, is consistent with the general equilibrium framework. 6 In our model, we consider the interactions between regions through interregional trade. The interregional trade incurs iceberg type of transport costs. We describe the economy in a general equilibrium as a welfare program. The main advantage of such a model is that it is directly suitable for welfare analysis ( Ginsburgh and Keyzer, 2002). Our model can thus be used for measuring the direct welfare effects of infrastructure improvements (measured based on the demand for transport), the overall welfare effects (measured by means of compensated full-expenditure or the equivalent full expenditure for consumers) and the difference between the two, the so-called indirect effects. In our model, we focus on a short run context (e.g. less than 10 years). Hence, we presume that production factors (labor and capital) are immobile between regions, but mobile between sectors. 7 The capital market is perfect, while the labor market is imperfect, because there is (shown to be) involuntary unemployment. We presume the presence of regions with different production technologies due to differences in endowments. We determine the base equilibrium by solving a two-region model for given parameter values. The numerical simulation shows that the model is consistent with the wage curve ( Blanchflower and Oswald, 1994). The model is further applied to study the direct and indirect effects of transport infrastructure improvements through changes in transport costs. The paper is organized as follows. In Section 2, we discuss the efficiency wage setting mechanism shortly. In Section 3, we incorporate efficiency wage setting formally in a general equilibrium setting and present a SGE model incorporated with this setting. In Section 4, we solve a two-region model numerically and consider changes in transport costs. We focus on the degree of surplus equivalence and the extent to which it depends on particular model parameters. Section 5 concludes.

Previous studies showed that there was a departure between social benefits and user’s benefits due to market power in product markets and agglomeration. This paper studies the surplus equivalence problem in the context of labor market imperfections. The direct welfare effect of an infrastructure improvement is measured by the user benefits using consumer surplus, while the economy-wide welfare effect is measured by social welfare. Specifically, we develop a spatial (regional) general equilibrium model incorporated with efficiency wage theory to include an imperfect labor market. Our model is built on the basis of central planner’s welfare maximization subject to the economic constraints including wage setting and an exogenous unemployment benefit policy. Due to the imperfect labor market and the exogenous unemployment benefit policy, the model outcome (the welfare optimum or the equilibrium) reflects a second-best world. It is constrained Pareto-efficient for the welfare objective. For comparing the social welfare based on the welfare program with the direct welfare based on the consumer surplus in the transport market using the rule of half, we examine how social welfare increases with a decrease in transport costs in the welfare program. Intuitively, reduced transport costs increase interregional trade and the production quantities, which leads to a more efficient production structure among regions by reallocating labor use. In the context of our model, involuntary unemployment falls, and the benefits of this fall, which are not captured by the firm’s demand for transport when using the rule of half, must be included in the welfare calculations. Our model has been employed for analyzing the impacts of the improvement of infrastructure between regions in a numerical example, assuming tradable goods and immobile factors. We conduct the welfare analysis in the context of the surplus equivalence problem, where we distinguish between the direct welfare effect using user’s benefits in the transport market and the indirect welfare effect through changes in involuntary unemployment. The numerical simulation shows that as the transport cost rate decreases due to the infrastructure improvement, the interregional trade increases and there is a tendency of specialization for more efficient production. This leads to a welfare improvement in both regions. As a whole, the indirect welfare effects are about 10–20% of the direct welfare effects for most plausible parameter values which ensures the elasticity of the wage curve to be around −0.1. The order of magnitude of a 10–20% underestimation for welfare effects of infrastructure improvement due to labor market imperfections may be interpreted as moderate. On the one hand, the indirect effect is not so large that it would not lead to entirely different investment programs. It would only affect decisions on those alternatives where costs and benefits would be rather close. On the other hand, it is not so small that it just enters the domain of the usual uncertainties that prevail in applied policy analysis. Besides, it has a consistently positive outcome, implying that ignoring it would lead to a systematic – though limited – bias in decision-making in this field. Hence, it would be wrong to ignore it in real world applications. In addition, one should be aware that the indirect welfare effect related to labor market imperfections considered in the current paper might be complementary to indirect welfare effect related to commuting (Pilegaard and Fosgerau, 2008), as well as to indirect effects in other markets as mentioned in Section 1. The sum of these indirect welfare effects related to various markets may well reach much higher levels. We conducted an elaborated sensitivity analysis of the welfare effects to the values of some important parameters. It shows that an economy with poor infrastructure condition has more indirect welfare impacts than the economy with good infrastructure. A larger labor market imperfection has higher indirect welfare effects. This study shows that the surplus equivalence problem becomes important for assessment of transport infrastructure projects with poor initial infrastructure condition and when labor market imperfections are large.