Using novel daily measures of consumption, firm sales, and employment, we show that the economy reacts within days to monetary policy shocks. Time aggregation of the measures to quarterly frequency masks these short-run dynamics.

We use the structure of FOMC meetings to measure the uncertainty policymakers face about different macroeconomic conditions, and the associated impact on policy. Uncertainty amplifies the reaction function coefficients, especially for inflation.

When central banks communicate signals on uncertainty in economic conditions, they can have their largest impact on long-run yields. This novel form of information effect explains the market reaction to the Bank of England's Inflation Report very well.

Using communication measures from machine learning, we find evidence for both the conformity and discipline effects predicted by the career concerns literature following an increase in transparency on the Federal Open Market Committee. On balance, the discipline effect appears stronger, as rookie members become more influential.

A new model of reputation for monetary policy makers predicts that all preference types become softer on inflation over time and that this evolution is more pronounced for types that put more weight on output, predictions we confirm using voting data from the Bank of England.

We adopt an automated approach to measuring the extent to which Fed statements provide forward guidance versus information on economic conditions. Overall, the former has larger macroeconomic effects, particularly on market variables.

Experts on the Bank of England's Monetary Policy Committee differ both in terms of preferences and private forecasts. A committee of internal Bank members significantly outperforms an individual because of information pooling, but both large committees and those that add externals appear to add little value.

NFL draft rules randomly allocate players of similar quality to teams of different quality. But there is no effect of initial match on long-term performance, suggesting that firms do not contribute to long-term inequality in sufficiently competitive markets.

We use a large language model trained on tens of thousands of human labels to measure remote and hybrid work adoption with unprecedented granularity. We document large heterogeneity in adoption across narrow occupation categories, cities, and firms.

We measure the demand for executive skills aross firms using a unique, large-scale corpus of job specifications. The importance of social skills is growing over time and is related to firm size, firm scope, and the information intensity of worker skills.

We combine elements of supervised machine learning and dictionary methods to uncover narrow, interpretable risk exposures from 10-K filings that account for firm-level reactions to aggregate shocks. We apply this idea to study equity returns during the COVID-19 pandemic.

We use a machine learning algorithm applied to granular CEO survey data to construct a scalar behavioral index. The index is strongly correlated with firm performance, and this relationship appears only several years after CEO appointment. Evidence suggests the correlation is due to assignment frictions, which are substantially worse in low-income countries.

When downstream firms are sufficiently risk averse, or subject to limited liability, and cannot observe their competitors' shocks, an upstream firm offers contracts that offer all input to one downstream firm.

We describe how one can interpret public sector organizations through the lens of managerial accounting, and provide anecdotal evidence that mixing strong incentives with bureaucratic administration works well in various settings.

When workers have career concerns, performance feedback increases the uncertainty of future effort but allows them to manipulate their employers' beliefs on future effort; the optimal policy only reveals intermediate performance levels.

We use comprehensive financial transactions from one of Europe's largest banks to build a large-scale consumption survey using national accounting principles. We use this to build aggregate and distributional accounts for consumption, as well as a detailed individual consumption panel.

We use a database of 1.4 billion payments transactions to track the impact of the COVID-19 pandemic and policy response across time, space, and sectors in Spain.

Plugging measures derived from AI/ML algorithms into regressions leads to biased estimation but does not distort confidence intervals. Valid inference can be performed with a bias correction or maximum likelihood estimation.

We show how observed connections among variables in multiple networks can improve estimation of the precision matrix by targeting regularization and develop a Bayesian model for inferring pairwise correlations.

Econometricians can use variation in the prior distribution to substantially improve the accuracy of existing techniques for estimating decision-makers' preferences and private signal distributions.

Review of algorithms for text analysis of economics; introduction of four core measurement problems; discussion of future challenges for text-as-data in economics.