The 2021-2022 inflation episode presented the first opportunity to examine inflation and price dispersion using U.S. scanner data in a high-inflation environment.  Data from 50,000 outlets reveals that price changes across similar goods grew more dispersed in 2022.  This paper documents how price change dispersion interacts with households' product choices to generate inflation heterogeneity.  The heterogeneity is substantial, with a 1.4 percentage point interquartile range growing to 4.0 percentage points in 2022. Households offset little of their implied budget shocks through substitution.  A model with idiosyncratic preferences rationalizes household behavior and implies that household-level inflation rates place observable bounds on their welfare losses.  Results imply that widening price dispersion seen in 2022 generated annualized losses of $600 to $1,300 for households in the 10th and 90th percentiles of the inflation distribution.

We study a model of occupational change featuring two forces: entry and exit of occupations in the labor market and shocks to relative demand/supply across occupations.  The model allows us to evaluate each force's contribution to changes in the distribution of wages and calculate wage indices that account for the growing variety of occupations in addition to these shocks. Using this framework as a lens on occupational change from 1940 to 2020, we estimate that the increasing variety of occupations reduced firms' wage cost growth by 3.3 percentage points while increasing workers' effective wage growth by 8.6 percentage points.  Occupational entry also contributed to one-fifth of the increase in wage inequality across occupations from 1970 to 2020.  The other four-fifths of that increase was driven by changes in relative labor demand, which also lowered firms' effective wage growth by 11.1 percentage points.  The model implies that workers exhibited similarly large changes in their willingness to supply labor across occupations, with rising supply to occupations with more social and less routine tasks.  Accounting for these shifts raises workers' effective wages by 13.9 percentage points from 1940 to 2020, but played no role in narrowing the interquartile range of occupational wages over the past fifty years.

Economic information contained in Big Data present a rich opportunity for improving the timeliness, quality, and efficiency of economic statistics.  A major challenge with such datasets, however, is that they are rarely representative in their coverage.  This paper tests the efficacy of several techniques that survey methodologists have advanced for using non-probability data to produce unbiased estimates of the overall population.  Simulated data featuring various forms of bias reveal two important conclusions.  First, unsurprisingly, unbiased estimation cannot be guaranteed without information from outside the Big Data source (unless it covers the entire popluation of interest).  In a more surprising conclusion, I find an opportunity for Big Data sources to highlight "mis-stratification," an arguably more problematic source of error wherein an outdated sampling frame leads traditional survey techniques to produce biased statistics.

We investigate post-recession trends in home equity extraction and how they may have impacted household spending and residential improvements. Home equity extractions—which rose and fell with house prices in the 1990s and 2000s—remained sluggish after 2010 despite low interest rates and large gains in home equity. Compared to the mid-2000s, equity extractions fell more sharply for younger households and those with lower credit scores, suggesting tighter mortgage credit supply than before the recession for portions of the population. 

Related press interview: Americans' Home Equity Surges to Over $35 Trillion (Wall Street Journal)