Uncertainty is often regarded as a major driver of economic downturns. Since Bloom (2009), the prevailing view in macroeconomics is that heightened uncertainty invariably leads to economic contractions. The conventional wisdom holds that when uncertainty rises, firms postpone investment, households reduce consumption, and financial markets experience heightened volatility, ultimately leading to slower economic growth (Bernanke 1983, Baker et al. 2016). However, recent research challenges this assumption, demonstrating that uncertainty episodes do not uniformly translate into recessions. Empirical studies have shown that uncertainty often coexists with positive economic growth, and its effects can depend on the type of uncertainty at play (Mumtaz and Zanetti 2013, Jurado et al. 2015, Ludvigson et al. 2021, Fernandez-Villaverde et al. 2025).
In a recent paper (Gambetti et al. 2025), we introduce two novel concepts of uncertainty — agreed and disagreed uncertainty — to explain the differential macroeconomic effects of uncertainty shocks. Our findings show that when uncertainty is accompanied by widespread consumer disagreement about economic conditions (disagreed uncertainty), the economy remains stable, despite the high level of uncertainty. In contrast, when uncertainty coincides with low consumer disagreement (agreed uncertainty), economic activity falls in response to high uncertainty, as originally shown in Bloom (2009). These insights have important implications for policymakers and economic forecasters, particularly as they navigate periods of economic stress such as the 2008 financial crisis and the COVID-19 pandemic (Bloom et al. 2018, Cascaldi-Garcia and Galvao 2021).
Uncertainty and economic activity: Revisiting the evidence
While the prevailing view in macroeconomics links heightened uncertainty to recessions, our work shows that this relationship is not always borne out in the data. Figure 1 presents indicators of macroeconomic and financial uncertainty from Jurado et al. (2015) alongside industrial production (IP) growth. It reveals that while several episodes of elevated uncertainty coincide with economic downturns (grey-shaded areas), there are also many periods where uncertainty rises but the economy continues to grow (horizontally hatched areas). This challenges the assumption that uncertainty is always contractionary.
Figure 1 Uncertainty indicators and the growth rate of industrial production
Note: Annual growth rate of industrial production (IP, grey-dashed line) and the index of macroeconomic (top panel) and financial (bottom panel) uncertainty (dark-solid line) from Jurado et al. (2015) normalised to have zero mean and unitary variance for the period 1979.M1-2019.M12. Patched areas show periods when uncertainty was above its sample mean, differentiated between negative IP growth (grey-shaded area) and positive IP growth (horizontally hatched area). Unshaded/unhatched areas show periods with uncertainty below the sample mean.
To understand these differences, we focus on consumer disagreement — a key but often overlooked dimension of uncertainty. Using survey data from the University of Michigan, we construct a new index of consumer disagreement, capturing the dispersion in expectations about economic conditions. Figure 2 shows the disagreement calculated using the cross-sectional dispersion of views from the Michigan Survey of Consumers. Our analysis shows that disagreement is procyclical, increasing in economic expansions and decreasing during downturns. This suggests that periods of high uncertainty often occur alongside varying degrees of consumer consensus or disagreement.
Figure 2 Time series of consumer disagreement
Note: Shaded areas are NBER recession dates.
Figure 3 further illustrates this relationship by plotting the scatterplot of uncertainty against the monthly growth rate of industrial production (IP). This figure helps differentiate between periods of agreed and disagreed uncertainty. In particular, we define disagreed uncertainty as periods where both uncertainty and consumer disagreement are above their respective sample medians. The left panel of Figure 3 shows data points corresponding to high uncertainty and high disagreement, while the right panel includes observations with high uncertainty but low disagreement or low uncertainty overall. The regression line and confidence intervals in each panel demonstrate that during periods of disagreed uncertainty, the correlation between uncertainty and industrial production growth is not significantly different from zero. In contrast, during periods of agreed uncertainty, the correlation is negative and statistically significant, supporting our hypothesis that agreed uncertainty is contractionary while disagreed uncertainty is not.
Figure 3 Uncertainty and the growth rate of industrial production
Note: Scatterplot of uncertainty versus the growth rate of industrial production (IP), overlaid with a regression line (solid) and 95% confidence intervals (dashed). The left panel highlights periods where both uncertainty and disagreement are above their sample medians. The right panel includes periods with high uncertainty but low disagreement, as well as periods with low uncertainty regardless of disagreement levels.
Distinguishing between agreed and disagreed uncertainty
We propose a simple model where agents receive imperfect and dispersed information about economic fundamentals. In this framework, uncertainty is linked to the forecast errors that agents make when trying to predict future economic conditions. Disagreement, in turn, arises when individuals interpret economic signals differently due to noisy information. The key predictions of the model are that agreed uncertainty (where agents' expectations align) leads to strong negative effects on economic activity, whereas disagreed uncertainty (where agents hold divergent views) does not result in contractionary outcomes.
The key insight from the simple model is that uncertainty can take two forms:
- Agreed uncertainty: When uncertainty rises, but consumer disagreement remains low, economic contractions follow. This is because agents react similarly to shocks, leading to a synchronised downturn in demand and investment.
- Disagreed uncertainty: When uncertainty is high, but consumer disagreement is also high, the economic impact is muted. In these cases, a portion of consumers remains optimistic, sustaining economic activity despite rising uncertainty.
Empirical evidence: The effect of uncertainty depends on disagreement
Using a Bayesian vector autoregression (VAR) model with sign restrictions informed by our theoretical model, we empirically identify agreed and disagreed uncertainty shocks in US data (1979-2019). Figure 4 reveals striking differences: Agreed uncertainty shocks lead to significant and prolonged declines in industrial production and employment. Disagreed uncertainty shocks, despite increasing overall uncertainty, do not cause economic contractions. In fact, we show that in 45% of heightened uncertainty episodes (i.e. high uncertainty and high disagreement), the economy did not contract — suggesting that a large share of uncertainty shocks may be less harmful than previously thought.
Figure 4 Figure 4: Agreed (left) versus disagreed (right) uncertainty shocks
Note: The figure shows impulse responses from a eight-variable VAR system on JLN 12-month ahead uncertainty indicator (JLN12), disagreement index (DISAG), industrial production (IP), private consumption (CONS), Consumer price inflation (INFL), employment (EMPL), S&P 500 index (SP500), Federal funds rate (FEDFUNDS). The shaded gray areas are the 16% and 84% posterior bands generated from the posterior distribution of VAR parameters.
Evidence from the New York Fed Survey of Consumer Expectations
To further validate our findings, we examine survey data from the New York Fed Survey of Consumer Expectations (SCE). We leverage the 2022 Russian invasion of Ukraine as a natural experiment, distinguishing between consumers who responded to the survey before and after the Federal Open Market Committee (FOMC) statement on 16 March 2022. The FOMC statement provided information about future economic conditions, reducing consumer disagreement among respondents surveyed afterward. Our analysis shows that spending plans contracted more significantly for consumers with reduced disagreement after the FOMC statement, reinforcing our main finding that agreed uncertainty has stronger recessionary effects than disagreed uncertainty.
Policy implications
Our findings suggest that policymakers should monitor and distinguish between different types of uncertainty before responding with stabilisation measures. If heightened uncertainty is of the agreed type — marked by widespread consumer consensus and low disagreement about a downturn — then policy interventions, such as monetary easing or fiscal stimulus, may be warranted. However, if uncertainty is of the disagreed type, intervention may not be necessary, as the economy may be more resilient than expected.
Furthermore, central banks and policymakers should closely monitor consumer disagreement indicators, as they provide a valuable early signal of whether uncertainty is likely to be contractionary or benign. This is particularly relevant during crises, when uncertainty is elevated but its economic impact is unclear.
Conclusion
Our work shows that uncertainty is not always recessionary. By distinguishing between agreed and disagreed uncertainty, we show that consumer disagreement plays a crucial role in shaping the macroeconomic effects of uncertainty shocks. Our findings challenge conventional narratives and offer a new perspective on how uncertainty interacts with economic activity. Future research should explore how these dynamics play out in different contexts, such as financial crises and policy-induced uncertainty episodes.
References
Baker, S, N Bloom and S Davis (2016), “Measuring economic policy uncertainty”, Quarterly Journal of Economics 131(4): 1593–1636.
Bernanke, B S (1983), “Irreversibility, uncertainty, and cyclical investment”, The Quarterly Journal of Economics 98(1): 85–106.
Bloom, N (2009), “The impact of uncertainty shocks”, Econometrica 77(3): 623–685.
Bloom, N, M Floetotto, N Jaimovich, I Saporta-Eksten and S Terry (2018), “Really uncertain business cycles”, Econometrica 86(3): 1031–1065.
Cascaldi-Garcia, D and A B Galvao (2021), “News and uncertainty shocks”, Journal of Money, Credit and Banking 53(4): 779–811.
Fernández-Villaverde, J, Y Yang and F Zanetti (2025), “Technological Synergies, Heterogeneous Firms, and Uncertainty Shocks”, NBER WP 32247.
Gambetti, L, D Korobilis, J D Tsoukalas and F Zanetti (2025), “Agreed and Disagreed Uncertainty”, CEPR Discussion Paper No. DP19946.
Jurado, K, S C Ludvigson and S Ng (2015), “Measuring uncertainty”, American Economic Review 105(3): 1177–1216.
Ludvigson, S C, S Ma and S Ng (2021), “Uncertainty and Business Cycles: Exogenous Impulse or Endogenous Response?”, American Economic Journal: Macroeconomics 13(4): 369–410.
Mumtaz, H and F Zanetti (2013), “The Impact of the Volatility of Monetary Policy Shocks”, Journal of Money, Credit and Banking 45(4): 535-558.
