When Uncertainty Strikes, It Does Not Affect Everyone in the Same Way
When economic uncertainty rises, the initial reaction is almost intuitive: businesses postpone investments and hiring, households become more cautious, and financial markets pull back. Production and consumption slow down, and unemployment tends to rise. But behind these aggregate indicators lies a more difficult question: who really pays the price for uncertainty?
The answer is not necessarily the same for a low-income worker, a middle-class family, and those at the top of the income distribution. And it can change radically over time.
This is the issue addressed by Massimiliano Marcellino of the Department of Economics at Bocconi University and director of the Baffi Center on Economics, Finance, and Regulation, together with Florian Huber of the University of Salzburg and Tommaso Tornese of the Catholic University of the Sacred Heart. Their study, The Distributional Effects of Economic Uncertainty, published in the International Economic Review, analyzes how shocks of economic uncertainty are transmitted throughout the entire distribution of labor income and consumption in the United States.
Many previous studies measure inequality using a single indicator (such as the Gini index) or examine specific percentiles of the distribution. Huber, Marcellino, and Tornese, however, choose to study the entire shape of the distribution, tracking its shifts following a macroeconomic shock. To do so, they develop a Functional Structural Vector Autoregression (F-SVAR) model that links macroeconomic variables to the distributions of income and consumption.
This distinction is important because, as the authors write, “the effect of uncertainty shocks on inequality depends on the horizon and the specific distribution one focuses on.” In other words, simply asking whether uncertainty “increases or decreases inequality” risks being too broad a question. One must specify which inequality is being measured and when it is measured.
The first effect: workers at the lower end of the distribution disappear
The study identifies a two-stage dynamic. Following an unexpected rise in uncertainty, real GDP, consumption, investment, and stock prices decline, while unemployment rises. So far, nothing particularly surprising. The novel finding emerges when examining the income distribution.
In the first few years following the shock, among those who remain employed, the share of workers earning less than per capita GDP declines, while the relative weight of workers with higher incomes increases. At the same time, however, the share of households with low levels of consumption rises.
The result may seem counterintuitive: how can the share of low-income workers decrease just as the economy worsens?
The key lies in the denominator. The income distribution analyzed in the paper pertains to the employed population, not the entire population. The authors’ interpretation is that, in the initial phase, less-skilled, low-income workers are more vulnerable to job loss. When they leave the workforce, they also drop out of the observed wage distribution. Inequality among those who retain their jobs may thus decrease not because the poorest are becoming richer, but because a portion of them has become unemployed.
On the consumption front, however, something very concrete happens: a portion of households slides from the middle of the distribution toward the lower end. The deterioration in expected income and the risk of unemployment thus seem to quickly translate into greater caution in spending. The authors summarize the mechanism in a single sentence: “the propagation of uncertainty shocks happens in two phases.” And it is the second phase that changes the picture once again.
A few years later, inequality begins to rise again
As unemployment gradually begins to be absorbed, the share of low-income workers grows again, while the share of workers with intermediate incomes shrinks. Meanwhile, real wages and labor productivity weaken: in the study, they reach their lowest point about 5–6 years after the shock.
The interpretation put forward by the authors centers on investment. During periods of uncertainty, firms invest less; over time, lower investment can lead to weaker productivity and, consequently, lower wages, especially for less-skilled workers. This is an explanation proposed by the authors, not definitive proof of a single causal mechanism.
The result is that income inequality from labor, after initially declining among the employed, begins to rise again. The increase in the Gini index during the second phase is fully reversed only after about 8 years.
Consumption patterns, however, tell a different story. Overall consumption inequality rises immediately after the shock, but the effect is much less persistent: “total consumption inequality rises on impact, but the effect dissipates within 2 years of the shock.”
Two years, then, compared to effects on income distribution that can last much longer.
A refrigerator can wait; daily groceries cannot
To better understand this apparent return to normal consumption patterns, the authors separate durable goods from non-durable goods and services. This reveals another significant difference: the decline in consumption of durable goods is highly persistent and does not subside until about 6 years after the shock. For non-durable goods and services, however, the recovery is faster.
One possible explanation lies in access to credit. Households with lower incomes or those emerging from periods of unemployment may continue to face difficulties in financing major purchases, even as they resume spending on daily necessities. The authors also cite the possible role of improved confidence and fiscal transfers, but note that there is no single explanation based on the available data.
Economic uncertainty does not simply create “more” or “less” inequality. It alters its geography over time. First, it hits the labor market, pushing certain categories of workers out to a greater extent; at the same time, it curtails consumption among the most vulnerable households. Then, as employment recovers, more persistent effects may emerge in the form of missed investments, weak productivity, and tighter credit conditions.
The result therefore depends on the snapshot we choose to look at. Taken immediately after the shock, it may show lower wage inequality among the employed. A few years later, the same economy may show exactly the opposite. This is also why focusing solely on the average—or even on a single inequality index—can obscure a crucial part of the story.
Finally, an important limitation should be noted: the analysis of incomes concerns only the employed, and, due to the way survey data are constructed, the study does not describe in detail the wealthiest end of the income and consumption distributions. But by combining labor market dynamics with spending behavior, the paper demonstrates why shocks of economic uncertainty cannot be understood by looking solely at GDP: even during the same recession, different households may follow very different trajectories.
