Research

The Economic Impact of  Low- and High- Frequency Temperature Changes

with Nikolay Gospodinov and Serena Ng

May 2025

Abstract: Temperature data have low- and high-frequency variations that may have distinct impacts on economic outcomes. Analyzing data from a panel of 48 states in the U.S., and a panel of 50 countries, we find slowly evolving, low-frequency components with periodicity greater than 32 years. These components have a common factor that trended up around the same time that economic growth slowed. Panel regressions using U.S. data fail to find a statistically significant impact of low-frequency temperature changes on growth, though the impact of high-frequency temperature changes is marginally significant. However, using the international panel (which includes several European countries), we find that a 1°C increase in the low-frequency component is estimated to reduce economic growth by about one percent in the long run. Though the first-order effect of high frequency changes is not statistically significant in this data, a smaller non-linear effect is detected. Our estimation and inference procedures control for common, business cycle variations in output growth that are not adequately controlled for by an additive fixed effect specification. These findings are corroborated by time series estimation using data at the unit and national levels.

A Jackknife Variance Estimator for_Panel Regressions

with Richard K. Crump and Nikolay Gospodinov

FRBNY Staff Reports – October 2024

Abstract: We introduce a new jackknife variance estimator for panel-data regressions. Our variance estimator can be motivated as the conventional leave-one-out jackknife variance estimator on a transformed space of the regressors and residuals using orthonormal trigonometric basis functions. We prove the asymptotic validity of our variance estimator and demonstrate desirable finite-sample properties in a series of simulation experiments. We also illustrate how our method can be used for jackknife bias-correction in a variety of time-series settings.

A Simple Diagnostic for Time Series and Panel-Data Regressions

with Richard K. Crump and Nikolay Gospodinov

FRBNY Staff Reports – October 2024

Abstract: We introduce a new regression diagnostic, tailored to time-series and panel-data regressions, which characterizes the sensitivity of the OLS estimate to distinct time-series variation at different frequencies. The diagnostic is built on the novel result that the eigenvectors of a random walk asymptotically orthogonalize a wide variety of time-series processes. Our diagnostic is based on leave-one-out OLS estimation on transformed variables using these eigenvectors. We illustrate how our diagnostic allows applied researchers to scrutinize regression results and probe for underlying fragility of the sample OLS estimate. We demonstrate the utility of our approach using a variety of empirical applications.

The Nonlinear Case Against Leaning Against the Wind

with Nina Boyarchenko, Richard K. Crump, Keshav Dogra, and Leonardo Elias

FRBNY Staff Reports – May 2024

Abstract: We re-examine the relationship between monetary policy and financial stability in a setting that allows for nonlinear, time-varying relationships between monetary policy, financial stability, and macroeconomic outcomes. Using novel machine-learning techniques, we estimate a flexible “nonlinear VAR” for the stance of monetary policy, real activity, inflation, and financial conditions, and evaluate counterfactual evolutions of downside risk to real activity under alternative monetary policy paths. We find that a tighter path of monetary policy in 2003-05 would have increased the risk of adverse real outcomes three to four years ahead, especially if the tightening had been large or rapid. This suggests that there is limited evidence to support “leaning against the wind” even once one allows for rich nonlinearities, intertemporal dependence, and crisis predictability.

Analyzing Recent Price Anomalies in Argentina: Global Influences and Domestic Distortions

with Emiliano Basco, Emilio Blanco, and Luis Libonatti

Latin American Journal of Central Banking – September 2025

Abstract: This study investigates Argentina’s unusually high and persistent goods inflation in the aftermath of the COVID-19 pandemic, using a cross-country monthly panel and a model that decomposes inflation into observable marginal costs, global price pass-through, and changing markups. We find that, at their peak in January 2024, goods price markups were approximately 40% higher than in November 2011. Further analysis indicates that these elevated markups were primarily driven by distortionary policy interventions—particularly foreign exchange controls, non-tariff barriers, and complex import regulations—which disrupted market pricing mechanisms and significantly amplified inflation, positioning Argentina as a clear outlier in the global inflation cycle.

Look Out for Outlook-at-Risk

with Nina Boyarchenko, Richard K. Crump, and Leonardo Elias

Liberty Street Economics – May 2023

Introduction: The timely characterization of risks to the economic outlook plays an important role in both economic policy and private sector decisions. In a February 2023 Liberty Street Economics post, we introduced the concept of “Outlook-at-Risk”—that is, the downside risk to real activity and two-sided risks to inflation. Today we are launching Outlook-at-Risk as a regularly updated data product, with new readings for the conditional distributions of real GDP growth, the unemployment rate, and inflation to be published each month. In this post, we use the data on conditional distributions to investigate how two-sided risks to inflation and downside risks to real activity have evolved over the current and previous five monetary policy tightening cycles.

What is "Outlook-at-Risk"?

with Nina Boyarchenko, Richard K. Crump, and Leonardo Elias

Liberty Street Economics – February 2023

Introduction: The Federal Open Market Committee (FOMC) has increased the target range for the federal funds rate by 4.50 percentage points since March 16, 2022. In tightening the stance of monetary policy, the FOMC balances the risk of inflation remaining persistently high if the economy continues to run “hot” against the risk of unemployment rising as the economy cools. In this post, we review a quantitative approach to measuring the evolution of risks to real GDP growth, the unemployment rate, and inflation that is inspired by our previous work on “Vulnerable Growth.” We find that, in February, downside risks to real GDP growth and upside risks to unemployment moderated slightly, and upside risks to inflation continued to decline.

"Zoned out of Opportunity": Causal Effect of Land Use Regulations on Adult Incomes

Senior Honors Thesis

Abstract: This paper uses a novel natural language processing technique – denominated concept recovery – to extract latent variables from the text of a large corpus of municipal regulations, in particular three 'concepts' relating to land use: 1) restrictions on intensive and extensive development, 2) minimum consumption requirements, and 3) 'regulatory taxes' (i.e. regulations that do not interdict development, but make it more expensive). With these estimands, we derive estimates of the causal effect of land-use regulations on the incomes in adulthood of children born in 2000 that grow up in urbanized areas. We find that land-use regulations have clear distributional consequences, advantaging children born to parents in the 75th percentile of the income distribution, while disadvantaging children born to parents in the 25th percentile of the income distribution.

Remote Learning and Parental Labor Market Outcomes

Junior Honors Term Paper

Abstract: This paper estimates the effects of remote learning on parental labor market outcomes. Using mobility indices, it captures school closures as they occur in real time, obviating the need for possibly erroneous or outdated survey data. The results suggest that if a parent having one child learning remotely corresponds to a 13% reduction in the odds of being employed, ceteris paribus. Disaggregating this effect by parental sex, it finds that school closures have an ambiguous impact on fathers, whereas mothers are affected at the intensive and extensive margin. These results are robust to alternative specifications, including to the use of a time-invariant state-level instrument and a Bartik instrument (as in Goldsmith-Pinkham et al. [2020]).