Real Wages are Not Real: Reassessing CPI-Deflated Wages as a Measure of Living Standards
Abstract
Deflating aggregate income measures such as hourly wages by the CPI is commonly usedto estimate real purchasing power for the average consumer and, by extension, living
standards. Yet, even when accounting for income inequality, changes in real wages show
little correlation with alternative indicators of economic well-being in the U.S., such as food
insecurity, housing instability, or consumer debt defaults. I argue this is because CPIdeflated aggregate wages are conceptually misleading, as the exercise of aggregation
obscures information on the relationship between incomes and costs of living on the level
of the individual. I demonstrate this empirically using microdata from the Consumer
Expenditures Survey and a large retail scanner dataset that tracks household-level
consumer purchases. Accounting for the wide dispersion in the experiences of
households, I show that, contrary to standard real wage data, various components of living
costs have risen faster than incomes since 1994 as well as during the COVID-19 inflation. I
propose a more comprehensive measurement of real incomes that correlates more closely
with alternative indicators of economic well-being.