Skip to main content
U-M Poverty Solutions Logo U-M Poverty Solutions Logo


Back to publications

How Much Did Child Poverty Fall Between 1993 And 2019? How Does That Compare To What Happened Between 2019-2021?

Download full PDF essay

Download data (excel spreadsheet)


In a recent story in the New York Times, Jason DeParle wrote about a new report by the research center Child Trends, which concludes that between 1993 and 2019, child poverty in America fell by nearly 60%. By all available measures, child poverty did decline over this period. Yet based on a close look at the data, it’s not clear that the decline is as large as the Child Trends analysis suggests, or, what we should learn from it.

Poverty measurement is largely a function of accounting – most approaches set a certain income threshold that we think is required for households to avoid material hardship, and then account for all the resources coming into a home. If the resources clear that threshold, that household is not in poverty.

Based on where one sets thresholds and how one accounts for resources, the number of people deemed impoverished can vary dramatically. For example, according to the official poverty measure, which sets a threshold based on living standards in the 1960s and only counts cash income as resources, 15.3% of children – more than 11 million children – were poor in 2021. According to the supplemental poverty measure (SPM), which sets a threshold based on modern household budgets and includes tax-credits and in-kind safety-net benefits in the resource column, just 5.2% of children – or roughly 3.8 million children – were poor in 2021. Depending on which methodology you use, that’s a difference of more than 7 million children – not exactly a rounding error.

The Child Trends analysis, which reports large declines in child poverty between 1993 and 2019, uses the SPM, but “anchors” the income threshold in 2012 living standards, only adjusting for inflation in prior and future years, rather than adjusting the threshold itself to account for changes in living costs. This is not the Census Bureau’s SPM, but rather a variation of it that some scholars prefer. The choice to “anchor” makes a big difference. This methodology put roughly 3 million more children in poverty in 1993 than would have been found under the Census SPM methodology, and moves nearly 800,000 children out of poverty compared to the Census SPM measure in 2019. Thus, over the study period, roughly 3.8 million children are lifted out of poverty simply by redrawing the poverty lines in 1993 and 2019.

Accurately measuring progress against child poverty matters if we are to understand the state of child well-being in America, and what “works” in fighting child poverty. But if one’s findings are based largely on the assumptions made in building a measure, how can we have confidence in them?

Download full PDF essay

Download data (excel spreadsheet)