جهت گیری مهارت فنی به عنوان پاسخ شرکتها به مشکل بیکاری: یک مدل مطابق همراه با رتبه بندی متقاضی و الزامات مهارتی درونی
|کد مقاله||سال انتشار||مقاله انگلیسی||ترجمه فارسی||تعداد کلمات|
|13594||2009||7 صفحه PDF||سفارش دهید||محاسبه نشده|
Publisher : Elsevier - Science Direct (الزویر - ساینس دایرکت)
Journal : Labour Economics, Volume 16, Issue 3, June 2009, Pages 304–310
This paper considers an economy with heterogeneous workers where identical firms optimally decide on the degree of complexity of jobs. Meetings are depicted by an urn-ball process where firms rank their applicants and pick the best one. We show that a general rise in unemployment induces an increase in the employment shares of high-skilled workers which, in turn, makes firms choose more complex jobs, leading then to a decrease in the output of low-skilled workers. The technical skill bias is therefore related to the usual explanations of unemployment. Next, we state that a decentralized equilibrium is efficient in terms of job complexity but inefficient in terms of job creation when firms internalize the usual congestion effect. We then extend the analysis to a dynamic model.
One salient feature of the past several decades is the increase in wage inequality. This phenomenon is particularly marked in the United States where the college premium (the wages of college graduates relative to the wages of high-school graduates) increased by over 25% between 1979 and 1995 (Acemoglu, 2002). The change in the wage structure was associated with a general rise in unemployment as well as a decrease in the employment share of low-skilled workers (Autor et al., 1998). According to many authors, the (relative) poverty of unskilled workers stems from an exogenous bias in technology, see for instance Nickell (2004). A new technological revolution would have made high (low) skilled workers become more (less) productive in their jobs. The purpose of this paper is to show that this technical skill bias can be viewed as an optimizing response of firms to unemployment. Low wages and low employment shares of low-skilled workers as well as more complex jobs would directly result from a general rise in unemployment. Our analysis is based on a matching model of the labor market with endogenous skill requirements and applicant ranking. This model has two main features. First, although workers are vertically differentiated, there is a single labor market and the hiring process is described by an extension of the urn-ball model (Petrongolo and Pissarides, 2001) where firms rank their applicants and pick the best one. As a consequence, unemployment rates depend on the workers' ability. The higher the ability of a worker, the lower the unemployment rate is. Second, a job can be more or less complex; its degree of complexity being a firms' choice variable. The output of a filled job then depends on two variables which are the ability of the worker and the degree of complexity. Intuitively, we assume that an increase in the degree of complexity of jobs raises the output of high-skilled workers and lowers that of low-skilled ones. In the first part of the paper, we take the tightness of the labor market (hence the average rate of unemployment) as a given variable, and we study its effects on the dispersion of employment shares and on the degree of complexity of jobs. We state that a decrease in market tightness (which is associated with an increase in general unemployment) lowers the employment shares of low-skilled workers and raises the employment shares of high-skilled ones. As the number of applicants per vacant job increases, firms can recruit higher-ability workers. This gives them an incentive to offer more complex jobs, i.e. jobs which better suit high-skilled workers. In addition, the increase in job complexity lowers the productivity of low-skilled workers, leading then to a decrease in their relative wages. Conversely, when the market becomes tighter, firms have to switch to lower-ability workers, thus choosing less complex jobs. In other words, the bias in the technical change can be seen as a rational response of firms to a general rise in unemployment. In the second part of the paper, market tightness is made endogenous by assuming free entry and we study the welfare properties of a decentralized equilibrium. In terms of job complexity, a decentralized equilibrium appears to be always efficient. The reason for this is that firms maximize the expected output of jobs. This is exactly what a social planner should do. We also show that job creation tends to be too high. This result does not depend on firms' technological choices. It comes from the ranking of applicants which creates a new externality: an increase in job creation makes firms hire lower-ability workers, lowering then the average output. Concerning the average unemployment rate, the comparative statics of the model are the same as the basic matching model (Pissarides, 2000). So, in fine, job complexity is related to the usual explanations of unemployment. Recently, other search models with ex ante heterogeneous jobs and workers have been proposed (see Teulings and Gautier, 2004 for a brief presentation of this literature). In those models, the complexity of the jobs (or their specialization when workers are horizontally differentiated as in Marimon and Zilibotti (1999) or Gavrel and Lebon (2008)) is exogenous. In addition, contrary to ours, models with vertical differentiation often assume that the market is segmented in as many sub-markets as skill levels, see Mortensen and Pissarides (1999) or the literature about optimal taxation with equilibrium unemployment ( Hungerbühler et al., 2006). On the other hand, in Acemoglu (1999), Shimer and Smith (2000), Albrecht and Vroman (2002), Gautier (2002) and Blázquez and Jansen (2008), the labor market is not segmented and heterogeneous workers form a common pool of applicants; but, contrary to our setting, firms' search is random. In our opinion, the assumptions of segmentation or random search are too strong as they tend to ignore a very intuitive reason for the dispersion of unemployment rates: the simple fact that firms prefer to hire the best workers. Our contribution is closer in spirit to Acemoglu (1999). In this matching model, search is also random but skill requirements are endogenous. Firms decide on job complexity, hence the output of the workers in their jobs.1 Like Acemoglu (2002), we understand the behavior of technical change “by recognizing that the developments and use of technology is, at least in part, a response to profit incentives.” According to Acemoglu, the strengthening of firms' skill requirements results from the change in skill labor supply; this adaptation also explains the increase in the unemployment rates of all skill groups of the past decades. With skilled workers, firms choose complex technology. With unskilled workers, technical change tends to be “deskilling” as in the nineteenth century. In our contribution, the technical skill bias directly derives from a general rise in unemployment and understanding this phenomenon is equivalent to explaining unemployment. However from a formal point of view, our analysis is similar to Acemoglu (1999).2 When firms rank their applicants and pick the best candidate, an increase in general unemployment has the same effect as an increase in the supply of skilled workers when search is random. The paper is organized as follows: The (static) model is exposed in Section 2. Section 3 studies the effects of a market-tightness increase on employment shares and job complexity. Market efficiency and comparative statics are the topic of Section 4. In Section 5, we prove that our main results still hold in a dynamic setting. Section 6 is the conclusion.
نتیجه گیری انگلیسی
To sum up, these results suggest that low wages and employment shares of low-skilled workers may come from a general rise in unemployment which would also explain the increase in job complexity. So, the technical skill bias would not be the cause of high unemployment but its consequence. Our analysis also showed that allowing firms to rank applicants highlights a new externality of job creation. The model we proposed is based on some restrictive assumptions. In particular, the analysis should be extended to the case where the destruction probability of a job depends on the worker's ability. Introducing alternative wage-setting mechanisms is likely to affect the efficiency results. In this respect, we surmise that with posted wages (Moen, 1997), market efficiency would be restored. It would be particularly interesting to introduce a mandatory minimum wage. This labor market institution could explain why the increase in wage inequality is not as sharp in many European countries as in the United States and the United Kingdom (Katz and Autor (1997)). As Acemoglu (2002) points out, a binding minimum wage makes the firm be the full residual claimant of an increase in the output of unskilled workers (who are on the minimum). Firms would therefore have a greater incentive to raise the productivity of unskilled workers by reducing job complexity. Consequently, the wage of skilled workers as well as inequality would be lower in the presence of a minimum wage. To conclude, we would like to note that market uniqueness is as unrealistic an assumption as perfectly segmented sub-markets. A middle term should be found. In this respect, one preliminary question must be addressed which is how the number of markets is determined.