Unequal We Stand: An Empirical Analysis of Economic Inequality in the United States, 1967-2006

Abstract

We conduct a systematic empirical study of cross-sectional inequality in the United States, integrating data from the Current Population Survey, the Panel Study of Income Dynamics, the Consumer Expenditure Survey, and the Survey of Consumer Finances. In order to understand how different dimensions of inequality are related via choices, markets, and institutions, we follow the mapping suggested by the household budget constraint from individual wages to individual earnings, to household earnings, to disposable income, and, ultimately, to consumption and wealth. We document a continuous and sizable increase in wage inequality over the sample period. Changes in the distribution of hours worked sharpen the rise in earnings inequality before 1982, but mitigate its increase thereafter. Taxes and transfers compress the level of income inequality, especially at the bottom of the distribution, but have little effect on the overall trend. Finally, access to financial markets has limited both the level and growth of consumption inequality.

Technology diffusion and its effects on social inequalities

“…situation is reversed, supporting the real data 20 (Erosa et al (2009) and Heathcote et al (2010)). This means that the skilled workers always consume more than their unskilled co-workers, as shown in Figure 3.…”

“…The investment in education (Figure 4) and the human capital accumulation ( Figure 5) tend to be proportional to the acceleration of the skill premium, which in turn means that it is proportional to the technological diffusion rate. Indeed, as soon as the technological diffusion rate decreases, and similarly the skill premium, we observe a slowdown in the 20 Heathcote et al (2010), using the households' budget constraints as a proxy for skill endowments, found a strong correlation between variance of log earnings and log hours, and an increase in the variance of the annual log hours, despite the fact that inequality in working hours for the group of white Caucasian males exhibits no obvious trend over the last forty years. Erosa et al (2009), using, just as us, the years of education as proxy for the skill endowments, found that college educated people work less before the age of 26 and more afterwards, as predicted by our model.…”

Is the Social Security Crisis Really as Bad as We Think?

“…To calibrate its distribution, I follow Zhao (2014) and assume that ln ϕ ∼ N 0, σ 2 ϕ , and then set σ 2 ϕ = 0.45, which is the variance of log male annual wages in Heathcote et al (2010). Also, I use Gaussian quadrature to transform the continuous distribution into a 5-point discrete distribution for computational convenience (see Figure 4).…”

Optimal pensions in aging economies

“…Guvenen et al (2015), for example, study the dynamics of individual labor earnings over the life cycle. In contrast to previous studies on income inequality such as Heathcote, Perri, and Violante (2010) and Guvenen (2009) that apply data from the Current Population Survey, the Panel Study of Income Dynamics or the Survey of Consumer Finance, Guvenen et al (2015) employ the more comprehensive data set from the Master Earnings File of the U.S. Social Security Administration records. They find, among others, that 1) earnings shocks display substantial deviations from lognormality in the form of an extremely high kurtosis and that 2) the statistical properties of the labor earnings process vary over the life cycle.…”