Regional Trends in Inflation and Nominal Wages

Since 2020, the U.S. has experienced important economic turbulence, marked by the highest inflation rate in four decades alongside substantial nominal wage growth in booming labor markets. However, these changes haven’t affected all Americans equally, and some areas in the country have seen larger movements in prices and wages than others.

In this blog post, we study the geographic disparities in both price increases and wage growth across different metropolitan areas in the U.S., revealing a complex pattern of changes in the cost of living and real wages.

Inflation and Wage Data

Information about inflation for different U.S. regions comes from the Bureau of Labor Statistics (BLS), which publishes data on the consumer price index for all urban consumers (CPI-U) for the 21 largest metropolitan statistical areas (MSAs).As of the 2020 Census. MSAs are large cities or groups of nearby cities. In most cases, the data are available every two months, and we can distinguish between the evolution in the cost of shelter and other goods and services. We aggregate these data to the quarterly frequency.

Our measure of wage growth comes from the Quarterly Census of Employment and Wages (QCEW), also from the BLS. The QCEW has data on average weekly wages for all workers across geographic locations covered by unemployment insurance plus public sector workers. We used these data for the same set of MSAs.

Using these data on prices and wages, we look at accumulated inflation and wage growth between the first quarter of 2019 and the first quarter of 2024. That is, we look at accumulated growth over the past five years, starting in the pre-pandemic period.Our sample ends in the first quarter of 2024, as this is the last data point available. Since the data are not seasonally adjusted, we chose the same quarter in 2019 (the pre-pandemic year) to avoid seasonality affecting our results.

Shelter Inflation and the Cost of Living

Accumulated inflation since the first quarter of 2019 has differed importantly across locations in the U.S. The first figure shows accumulated inflation for all goods and services (all items) for each of the selected MSAs. Since there are important differences across cities in the cost of housing, the figure also shows the accumulated percent change in the price of shelter and all items excluding shelter for different MSAs.

Accumulated Inflation from 2019:Q1 to 2024:Q1

SOURCES: Bureau of Labor Statistics and authors’ calculations.

In general, shelter inflation and all items excluding shelter tend to move together, which means that in cities with high shelter inflation, it is likely to see also above average inflation in other goods and services. However, the correlation is far from perfect; across these cities, it is close to 0.34 during this period.

Differences in shelter inflation across cities are substantial, with shelter prices increasing by over 45% in Phoenix and Tampa in these five years, but only around 13% in San Francisco. Across these cities, the standard deviation in accumulated shelter inflation is 9 percentage points. On the other hand, the differences across cities in the five-year inflation rate of other goods and services is not as pronounced, and can vary between 16% and 23%, with a standard deviation of 2 percentage points.

One important reason for the different behavior of shelter inflation and other goods inflation is driven by how easy or difficult it is to move goods and services across cities. Goods are clearly easier to trade and ship around the country. If the price of a good increases much more in one city relative to another, it may be profitable to buy the good in the cheap location and sell it in the expensive location, which minimizes the differences in the cost of goods across cities.

Clearly, it is not possible to move some goods and many services easily. For instance, it’s very difficult or impossible to move a house to a new location, and the price of land itself greatly affects the price. In addition, there are many reasons people don’t choose to move to a city with lower housing costs, which limits the ability for markets to equalize housing prices across locations.

Nominal Wages and Inflation

In the next figure, we plot the accumulated percent change in nominal wages, as measured in the QCEW, and all-items inflation across cities from the first quarter of 2019 to the first quarter of 2024. In the scatter plot, the red line represents identical changes in these two variables. If a city lies below the line, it means that average weekly wages increased less than prices in these five years, in which case, real wages fell. Cities that lie above the line experienced wage increases in excess of price increases, representing an increase in real wages.

Nominal Wage Growth and Accumulated Inflation from 2019:Q1 to 2024:Q1

SOURCES: Bureau of Labor Statistics and authors’ calculations.

When real wages increase, individuals can afford more goods and services over time, which translates into increases in their well-being. As the figure shows, in most of the 21 cities, average wages increased more than prices, leading to increases in real wages in the five-year period. In San Francisco, the increase in real wages is notable at 19.6%, while it is more moderate in most other cities. In places like Philadelphia and Atlanta, prices increased faster than wages, and real wages declined. In the case of Atlanta, despite a moderate nominal wage increase, shelter and all items inflation were among the largest increases in our sample in this period. For Philadelphia, all items inflation was about average, but the city’s nominal wage increases were below than the national average.

In sum, we find that inflation evolved unevenly across major U.S. cities, mostly due to large differences in the evolution in the price of shelter. Coupled with disparities in nominal wage growth, real wages and individuals’ well-being across cities also had mixed evolutions.

Notes

1. As of the 2020 Census.
2. Our sample ends in the first quarter of 2024, as this is the last data point available. Since the data are not seasonally adjusted, we chose the same quarter in 2019 (the pre-pandemic year) to avoid seasonality affecting our results.