Endogenous Growth Mechanism as a Source of Medium Term Fluctuations in the Labor Market: Application to the US Economy

National Bank of Poland Working Paper No. 57

131 Pages Posted: 2 Feb 2011

Date Written: April 1, 2009

Abstract

A considerable part of the economic literature focuses on the sources and mechanisms of economic cycles. The bulk of this literature, including Real Business Cycle theory (see e.g. papers of Kydland and Prescott 1982, Prescott 1986, King, Plosser, and Rebelo 1988) or New-Keynesian theory (see e.g. Woodford 2003, Smets and Wouters 2003, Christiano, Eichenbaum, and Evans 2005), aim at explaining business cycles. Business cycles are usually defined, following a seminal contribution of Burns and Mitchell (1946), as fluctuations with periodicity between 1 and 8 years. But in the recent years, there has been a growing recognition of the importance of economic cycles that last more than 8 years - the so called medium term cycles. Blanchard (1997) and Solow (2000) were among the first, who stressed the importance of research on this issue and the need to develop models accounting for medium term fluctuations.

The most apparent empirical evidence on the importance of medium term cycle is the behavior of unemployment rate in the US economy. Unemployment was relatively low in the 1950s and 1960s of the last century, then increased on average for roughly next 20 years and later, in the 1990s, went back to a lower level. These fluctuations occur with periodicity far greater than a decade. Also the divergence of unemployment experience in US and large continental European countries in the 1970s and 1980s (for a discussion, see e.g. Blanchard 2006) is an indirect evidence of the importance of medium term fluctuations. The literature on changes in productivity growth trend in US (see e.g. Basu, Fernald, and Shapiro 2001) documents another important aspect of this issue.

An important paper of Comin and Gertler (2006) documents, in a rigorous way (using band-pass filters of Christiano and Fitzgerald 1999), various facts on medium term fluctuations in goods and capital markets. The paper also defines medium term cycles as fluctuations of periodicity up to 50 years. Comin and Gertler proposed a theoretical framework well suited for analyzing medium term cycles – they introduced concepts from the endogenous growth theory into the RBC model. Their approach follows a seminal paper by Romer (1990), with modifications accounting for the Jones’ critique of Romer’s model(see Jones 1995). Within their framework, short term shocks affect the profitability of production activity and influence the incentives to innovate and develop new products. Ultimately, it induces fluctuations in the number of available products, resulting in medium term fluctuations of the whole economy. One of the main findings of Comin and Gertler is that medium term fluctuations can be explained by the same factors as business cycle fluctuations1. What is important from our perspective, Comin & Gertler focused on capital and goods markets, leaving the analysis of labor market for further research. This study aims at filling this gap.

The empirical evidence on medium term fluctuations in the labor market is presented e.g. in the papers of Hall (2005d) and Hall (2005c). He stressed the importance of fluctuations in medium term frequencies in many macro variables. Additionally, he hypothesized that medium term cycles can result from adjustments that take place in an asymmetric information environment. Alternative explanations of the lower frequency variation in the labor market variables focus more closely on factors specific to the labor market itself. One of them is the hysteresis effect (see e.g. Blanchard and Summers 1986, Blanchard and Summers 1987), as predicted by the insider-outsider theory. Another branch of the literature highlights the role of demography in generating low-frequency labor market volatility, e.g. the prolonged effects of baby-boom generations (see e.g. Flaim 1990) or the changes in participation rates (see e.g. Juhn and Potter 2006) .

In this study we focus on the question if medium term fluctuations in the labor market may be explained by the prolonged effects of short lived shocks coming from the goods market. We focus simultaneously on the short term component and the medium term component of medium term cycle in a unified way. As the data suggest that the medium term volatility is present in various markets of the economy, we do not explain lower frequency variation in the labor market with factors specific only to labor market, but instead we look for a common source of volatility in various markets of the economy.

So, in other words, this study aims at answering the question, whether the shocks, believed to be the source of traditional business cycles, are able to generate substantial medium term fluctuations in the labor market.

The main theses of this study can be stated as follows: 1. Variation of economic activity in medium term frequencies is substantial and comparable to the variation in business cycle frequencies. 2. A large part of medium term fluctuations in both labor and goods markets may be explained by the same sources. 3. Endogenous growth mechanism is able to explain a large part of variation in medium term fluctuations. We construct a theoretical model (with explicitly specified micro foundations), that belongs to a class of Dynamic Stochastic General Equilibrium models, and then we calibrate (and partially estimate) it and verify its predictions against the data. As our analysis requires longer time series, we decided to focus on the US economy. Additionally, there have been many empirical papers analyzing US economy, which simplifies the calibration of the model. Following Romer (1990), we use the endogenous growth framework augmented with the search-matching description of the labor market. It follows closely the Diamond-Mortensen-Pissarides framework (the notion of this framework originates form a seminal contributions by Diamond 1982, Mortensen 1982, Pissarides 1985). The search theory introduces an inherent friction into the functioning of the decentralized labor market and allows to model unemployment as an equilibrium phenomenon.

As a source of volatility we use the technology shock, as it is commonly used in the Real Business Cycle literature and, as Hall (2005c) noticed, could also be the main driving force of the medium term labor market fluctuations. The literature acknowledges the fact that the standard search-matching models underestimate the volatility of unemployment, as observed in the data (see e.g Costain and Reiter 2003, Shimer 2005, Hall 2005b). Thus, we will analyze two extensions of the model that address this issue: shocks to matching technology and real wage rigidity.

Our framework focuses on the consequences of the changes in the developments of the goods market for the labor market. Thus, we do not model explicitly the labor supply decisions and treat them as exogenously given. We admit, that labor supply shifts could be an important source of economic fluctuations, also in the medium term, but to simplify the analysis we are leaving it outside the model. It allows us to see how important the main mechanisms of our model are in explaining the patterns in the data.

Introduction of the endogenous labor supply could only improve the model performance. This theory has at least two implications. First, in order to account for economic variation in medium term frequencies, it seems that there is no need for a new generation of models, but it is enough to augment the current generation of DSGE models with elements of the endogenous growth theory. Second, if our theory is true, the effects of economic policies are more persistent than it is usually implied by the standard DSGE framework. The last issue may be especially important for the monetary policy, but we will leave this for further research.

The study is organized as follows. First, we briefly discuss the existing literature in the context of the issues that are important from the perspective of the study. Then, we describe the US data features, concentrating on the medium term characteristics and derive a set of ”stylized facts” that will be useful from the modeling perspective. Next, we present the details and derivations of the theoretical model, which includes both the endogenous growth component and the search-matching mechanism on the labor market. This section also discusses the steady state properties and restrictions imposed on the model structure by the balanced growth path assumption. Next, we discuss our calibration strategy, along with the data and information sources used for this purpose. As the stochastic parameters of the model, together with stochastic shocks, are estimated from the US data, this section also addresses the estimation issues. The last section presents the predictions of the estimated model and verifies them against the US data. As the basic version of the model understates the extent of labor market volatility, we also extend our analysis in two distinct dimensions. Firstly, we introduce shocks to matching technology and secondly - real wage rigidities, both extensions aimed at resolving the volatility issue. In each case, we check the model predictions and verify its properties. The last part of this section compares the predictions of the model extended for wage rigidities with the predictions of the benchmark model - basic RBC model with search-matching and wage rigidities - and presents some evidence on the importance of the endogenous growth component. The last section concludes and discuss some implications for our results for policy and the economic modeling.

Suggested Citation

Gradzewicz, Michal, Endogenous Growth Mechanism as a Source of Medium Term Fluctuations in the Labor Market: Application to the US Economy (April 1, 2009). National Bank of Poland Working Paper No. 57, Available at SSRN: https://ssrn.com/abstract=1752228 or http://dx.doi.org/10.2139/ssrn.1752228

Michal Gradzewicz (Contact Author)

National Bank of Poland ( email )

00-919 Warsaw
Poland

Do you have negative results from your research you’d like to share?

Paper statistics

Downloads
57
Abstract Views
639
Rank
659,215
PlumX Metrics