ریسک درآمد، تغییرات اقتصاد کلان و جمعیتی، و نابرابری های اقتصادی در ژاپن

Using an OLG model with heterogeneous households, we investigate the relationship among income risk, macroeconomic and demographic changes, and economic inequality between 1980 and 2000 in Japan. By decomposing the primary factors in earnings and consumption inequality into macroeconomic variables and the demographic variable, we find that our model replicates the evolution of economic inequality in Japan. By performing counterfactual simulations, we demonstrate that two factors—changes in time-varying macroeconomic factors and the unexpected decline in the total factor productivity growth rate—played important roles in the increase in earnings and consumption inequality in the 1990s.

According to the life cycle permanent income hypothesis (LC-PIH), households base their consumption levels, savings rates, and labor supply on their expectation of future shocks, such as idiosyncratic labor productivity risks, as well as their anticipation of aggregate economic conditions that may affect their lifetime income. LC-PIH has been the focus of much research into the relationship between income risks and consumption inequality. It is, likewise, a focus of this study, which aim to explain two empirical findings regarding wage risk, income inequality, and consumption inequality in Japan between 1980 and 2000. First, income and consumption inequality, as measured by the variance of logarithms, increased between 1980 and 2000 (Kohara and Ohtake, 2006). Second, contrary to this upward trend in income and consumption inequality, the level of cross-sectional inequality over the life cycle remained relatively stable over this period (Abe and Yamada, 2009). Between 1980 and 2000, the Japanese economy experienced a series of the so-called “economic bubbles”—economic booms—followed by their inevitable bursting, and subsequently, economic recessions. At the same time, the Japanese population was rapidly aging, leading to significant demographic changes. As would be expected with a changing economy, time-varying macroeconomic variables such as total factor productivity (TFP) growth rates, capital share, and government expenditures, changed significantly between 1980 and 2000.1 As shown in Table 1, TFP growth rate, for example, was extremely low in 1990s. Facing various idiosyncratic risks regarding household income, such as promotions, bonuses, and unemployment, as well as changes in macroeconomic factors such as changes in technology and factor prices, households likely adjusted their decision-making in accordance with the new reality. By doing so, they likely impacted the income and consumption inequality.The Japanese economy between 1980 and 2000, thus, provides a good case study with which to consider the relationship between macroeconomic and demographic changes and economic inequality. Among the researchers who have done so, Hayashi and Prescott (2002) demonstrated that a standard neoclassical growth model with calibrated parameters could be used to identify the factors in major recessions in Japan, including those that occurred during the so-called “Lost Decade” of the 1990s.2 Many subsequent researchers have investigated Japanese economy using a dynamic general equilibrium framework, including Chen et al., 2006 and Chen et al., 2007 and Braun et al. (2007), who constructed a general equilibrium model for identifying the macroeconomic and demographic changes that contributed to the fluctuations in the Japanese savings rates. If macroeconomic factors do indeed explain changes in the savings rate, they may also explain changes in earnings and consumption inequality, as changes in the savings rate affect households' capacity to share risks, as well as their labor supply decisions through wealth and substitution effects. Other empirical researchers including Ohtake and Saito (1998), largely attributed the economic inequality of the Japanese economy in the 1980s to the aging of the population. As shown in Section 2, earnings and consumption inequality, as measured by the variance of logarithms, do indeed sharply increase with age. Therefore, if the proportion of middle-aged and older households increased because of aging, then economy-wide inequality would be expected to increase. Even when considering earnings and consumption inequality within a dynamic general equilibrium model, the impact of population aging cannot be ignored. Recognizing this fact, we construct an incomplete market overlapping generations (OLG) model calibrated to the Japanese economy focusing on the transition path between 1980 and 2000. Similar models have been extensively used for research into social security and public finance, including models developed by Conesa and Krueger (1999) and Nishiyama and Smetters (2005). Using these models, we decompose the primary factors in economic inequality into macroeconomic factors and the demographic factor of an aging population, and examine the extent of their impact on earnings and consumption inequality. To account for market incompleteness, our model assumes that households face idiosyncratic labor productivity risks that are ex post heterogeneous within cohorts and the existence of earnings and consumption inequality. Our primary consideration is determining whether changes in time-varying macroeconomic variables, such as TFP growth, and the demographic changes resulting from an aging population lead to the upward trend in earnings and consumption inequality and the increase in cross-sectional inequality with age observed in Japan between 1980 and 2000. According to the LC-PIH, perfectly foreseeable changes in macroeconomic factors should not affect decision making regarding consumption, savings, and labor supply. However, as households are heterogeneous in not only their idiosyncratic labor productivity but also their experienced history of macroeconomic variables, foreseeable changes do indeed affect decision-making. As our model quantitatively demonstrates, different cohorts with different histories experience fluctuations in their earnings and consumption paths that increase inequality, generating an upward trend in inequality paths. Our analysis is at the cross roads of two streams of economic literature. The first is quantitative macroeconomics research into the Japanese economy in accordance with that of Hayashi and Prescott (2002), who used a standard neoclassical growth model to demonstrate that time variations in TFP growth and decreases in workweek length greatly contributed to the sustained recession experienced during the Lost Decade. Our research also draws on Chen et al., 2006 and Chen et al., 2007 and Braun et al.'s (2007) extension of Hayashi and Prescott's (2002) results to explain fluctuations in Japanese saving rates over time. We, therefore, consider our research to be an extension of this stream of literature because it applies a model similar to those used in previous studies. However, unlike previous studies, which generally used a general equilibrium model to investigate the so-called first moments of the Japanese economy—macroeconomic variables such as GDP and savings rate—we use our model to examine the so-called second moments—earnings and consumption inequality as measured in variances of logarithm. The second strand of literature upon which our study draws is quantitative research into incomplete asset markets based on the Bewley–Aiyagari–Huggett setup.3 In particular, our approach is similar to that of Heathcote et al. (2010), who attribute changes in cross-sectional inequality in the US to an increasing college premium, a gender gap, and wage volatilities. Many other researchers have investigated the relationship between income (wage) risks and consumption inequality in the US (e.g., Storesletten et al., 2004a), focusing on the apparent sharp increase in income inequality throughout the 1980s. However, according to the Consumer Expenditure Survey, consumption inequality in the US did not increase significantly over that decade. If an income shock is transitory, income inequality appears to increase, although the corresponding consumption inequality does not necessarily change from the LC-PIH. On the other hand, increasing consumption inequality could be explained by an increase in permanent shocks experienced by households. Based on the same observations in the US, Blundell et al. (2008) attribute income dispersion and lack of dispersion in consumption inequality to the partial insurance of permanent shocks. Using a dynamic general equilibrium model with limited commitment, Krueger and Perri (2005) find that even if the income risk disperses within the model, consumption inequality does not necessarily do so. A significant implication of these findings is the importance of gaining awareness of the link between income/wage risks and consumption inequality in controlling business cycle fluctuations. However, these studies neglected consideration of the importance of macroeconomic factors themselves. To address this research gap, we examine whether macroeconomic factors such as the TFP growth rate generate earnings and consumption inequality between 1980 and 2000 in Japan, assuming the existence of idiosyncratic labor productivity risks over this period. Our examination results in three major findings. First, we find that we can explain the time path of earnings inequality between 1980 and 2000 using an OLG model with calibrated parameters to explain earnings inequality, although we find it difficult to explain consumption inequality using the model, particularly that which is used in the 1980s. Second, the results of our counterfactual simulations reveal that not only the demographic factor of an aging population but also changes in macroeconomic factors generate an upward trend in earnings and consumption inequality. In particular, the sharp increase in income earnings inequality over the 1990s cannot be explained without considering the impact of time-varying macroeconomic factors. Last, we find that the unexpected decline in the TFP growth rate in the 1990s might have been a factor in the increase in consumption inequality over that decade. In our benchmark case, we assume that all households have perfect foresight regarding the macroeconomic factors that will impact their decisions. In such a case, the model generate consumption inequality paths that appear to be U-shaped or flat. However, if households expect an increase in the TFP growth rate in 1990s, although such an expectation would be proven incorrect, consumption inequality would rise over that decade. Therefore, changes in macroeconomic factors, the demographic factor of an aging population, and households' erroneous predications regarding the TFP growth rate are keys to explaining the upward trend in earnings and consumption inequality between 1980 and 2000 in Japan. The remainder of this paper is organized as follows. In Section 2, we provide the background for the construction of our model by reviewing the empirical data regarding income and consumption inequality in Japan and defining the income risks examined in our model. In Section 3, we construct an OLG model with heterogeneous households that face idiosyncratic labor productivity risks. In Section 4, we calibrate the parameters for the Japanese economy, and then, based on the numerical results thereof, discuss income and consumption inequality in Japan in Section 5. After we confirm the sensitivity of our analysis in Section 6, we present our conclusions in Section 7.

In this study, we investigated the time path of earnings and consumption inequality in Japan between 1980 and 2000 using a dynamic general equilibrium framework that allowed us to analyze the quantitative impact of changing macroeconomic factors and the demographic factor of population aging. By using such a framework, our investigation extended Hayashi and Prescott's (2002) research into the factors contributing to the economic inequality during the 1990s in Japan. We found that by using an incomplete market OLG model with heterogeneous household, we could explain not only changes in macroeconomic variables, such as interest and saving rates, but at least partially explain time variations in the variance of log earnings and consumption. Our model is therefore an appropriate model to understanding of transitional dynamics within the Japanese economy. By performing counterfactual simulations to determine why economic inequality in Japan increased between 1980 and 2000, we could identify that time variations in the TFP growth rate and a reduction in workweek length impacted not only macroeconomic variables but also economic inequality. In particular, we found that even if the population had not been aging and had faced constant idiosyncratic wage risks, earnings inequality would still have increased due to the impact of changing macroeconomic factors. We also identified that households' erroneous prediction that the TFP growth rate would be higher than the actual rate may have contributed to the increase in consumption inequality in the 1990s. We could not explain several aspects of the time path of earnings and, in particular, consumption inequality. This may be partly attributed to the fact that we did not consider several factors that may affect earnings and consumption inequality, such as unemployment, female labor supply, and progressive taxation. In addition, changes in fixed effect, persistent, and the transitory shocks may also have contributed to the upward trend in economic inequality, although cross-sectional income and consumption inequality over the life cycle do not appear to have changed significantly between 1980 and 2000. Examining these variables in future research will prove fruitful in resolving the debate regarding the causes of economic inequality and stagnation in Japan, and in determining whether the upward trend in cross-sectional economic inequality has truly decreased household welfare. Making such a determination regarding welfare requires consideration of the factors contributing to increases in earnings and consumption inequality, such as skill-biased technological change or changes in human capital. As Kawamoto (2005) has argued, the decrease in the TFP growth rate in the 1990s may not be due to the low level of productivity of the Japanese economy but rather inefficient resource allocation between its productive and unproductive sectors. To determine the veracity of such an argument, we must consider the impact of problems in sector-specific allocation on economic inequality within the Japanese economy. These are important topics for future research.