Giorgio Primicieri, from Northwestern University, will be at Nova SBE for an Economics seminar.

The presentation will be held by video conference. To attend this or other Nova SBE seminar please register here to join our mailing list.  After registration you will receive the link that allows you to be at the seminar.

Inequality and business cycles
Florin Bilbiie, Giorgio Primiceri and Andrea Tambalotti
Extended abstract
Is inequality among households important for understanding the sources and propagation of
business cycles? The rapidly growing HANK literature models household heterogeneity in detail,
with the goal of matching certain properties of the micro data, such as the high average MPC
out of transitory income shocks, or the variance and the kurtosis of individual income growth
(e.g. Kaplan, Moll and Violante, 2018, AER). But it does not address explicitly the extent to
which models that, as a result, fit the key features of household heterogeneity and inequality
change our understanding of cyclical fluctuations.
We address this question head on, in the context of a tractable estimated HANK model, which
features two classes of households: unconstrained (or standard, “Euler-equation”) households
with low MPC, and “hand-to-mouth” agents with high MPC. These households also face
idiosyncratic risk, since they occasionally switch from one class to the other. Despite this simple
modeling of heterogeneity, our framework features both cyclical income inequality and
individual income risk, both of which may amplify business cycles. Individual risk, for example,
gives rise to precautionary saving behavior, and may aggravate recessions if risk itself is
countercyclical. Similarly, countercyclical inequality—the fact that households with high MPC
(and low income) are more exposed to business cycles, leading to a widening of income
inequality in recessions—affects the propagation of fluctuations because the movements in
income of individuals whose consumption is more sensitive to shocks are larger. The question is
how strong these propagation channels are in practice.
Our model is well equipped to study this question quantitatively because it matches the
countercyclicality of labor earnings inequality, a key feature of the micro data that existing
HANK models often abstract from. In the data, both the wages and the hours of low-income
workers are more cyclical than those of middle- and high-income individuals (Heatchote, Perri
and Violante, 2020, RED). This empirical finding suggests that the cyclicality of labor earnings
inequality is unlikely to be due to a more elastic labor supply schedule for low-income workers.
Instead, it must be due to the fact that the demand for their labor is extremely cyclical. We fit
this important aspect of the data (and thus the countercyclicality of labor earnings inequality)
by assuming that the labor supplied by low-income individuals is more substitutable with
capital than that of the rest of the population, along the lines of Krusell, Ohanian, Rios-Rull and
Violante (2000, Econometrica). Consequently, firms increase their demand for this type of labor
more rapidly in expansions, since capital can only be accumulated slowly.
We estimate the model using aggregate time series and cross-sectional inequality data, both
pre and post taxes. We then use the estimated model to analyze the two-way interaction
between the macroeconomy and inequality: How income and consumption inequality respond
to a variety of macroeconomic shocks, such as monetary policy, TFP and investment shocks;
and how the presence of heterogeneity and inequality affects the nature of those fluctuations.
To this end, we compute the counterfactual evolution of our economy under an alternative
fiscal rule that redistributes income to equalize the marginal utility of consumption across
households, as in the representative-agent benchmark. We show that such a counterfactual
representative-agent economy would experience fluctuations of similar nature to those that
characterize the estimated economy. But these fluctuations would be of substantially smaller
amplitude, suggesting that inequality does play an important role in the amplification of
business cycles.