Revisiting Minimum Wage as an Anti-Poverty Policy Tool
New research casts doubt on the minimum wage's role in reducing poverty
For decades, the minimum wage has been sold to the public as a tool to lift families out of poverty. From Franklin D. Roosevelt’s Fair Labor Standards Act to Senator Sanders’s proposed Raise the Wage Act, the political rhetoric has been consistent: Raising the minimum wage helps the poor. But does the evidence back this claim?
A frequently cited study by Arindrajit Dube, published in 2019, suggested it does. Dube found that increasing the minimum wage produces long-run poverty elasticities ranging between -0.220 and -0.459 for non-elderly individuals, indicating a substantial reduction in poverty over time. His analysis shows income gains at the lower end of the distribution and has given proponents of minimum wage hikes empirical ammunition to push for legislation like the Raise the Wage Act of 2023, which proposed raising the federal minimum wage to $17 an hour.
However, a new and more rigorous analysis by economists Richard Burkhauser, Drew McNichols, and Joseph Sabia seriously challenges Dube’s findings and, by extension, the long-standing narrative that the minimum wage is an effective antipoverty tool.
In their recently accepted paper in the Review of Economics and Statistics, Burkhauser et al. revisit Dube’s work with a more robust methodological framework, extending the analysis through 2019 and deploying a battery of dynamic difference-in-difference (DiD) estimators. The authors find that the impact of minimum wage increases on poverty is statistically indistinguishable from zero.
Using data from the March Current Population Survey, which spans nearly four decades, the authors find that a 10% increase in the minimum wage is associated with a 0.17% increase in the probability of being in poverty in the long run. This effect is not only statistically insignificant, it’s in the wrong direction. At the 95% confidence level, the authors can rule out poverty elasticities below -0.129, which directly contradicts Dube’s core estimates.
The credibility of minimum wage policy hinges on its ability to reach and lift up the poor. But, as Burkhauser and his colleagues demonstrate, the data tell a different story. The minimum wage simply does not reduce poverty in any meaningful or consistent way.
The new paper’s findings are consistent across multiple estimation strategies, including traditional two-way fixed effects, stacked DiD, the Callaway and Sant’Anna approach, and synthetic DiD estimators. These estimators are designed to address common pitfalls in minimum wage research, like improper control groups, dynamic treatment effects, and the use of geographically proximate states as counterfactuals even when they fail to mirror the pretreatment poverty trends of the treatment states.
In fact, one of the most critical takeaways from the Burkhauser et al. study is that the poverty-reducing effects Dube observed are fragile. They hinge on two problematic assumptions: (1) the inclusion of macroeconomic controls like the state unemployment rate, which might be “bad controls” by capturing labor market effects of the minimum wage itself, and (2) limiting comparisons to nearby states, which can introduce bias when those nearby states are poor matches.
When these assumptions are relaxed or replaced with more empirically defensible ones, Dube’s estimated poverty reductions evaporate. Even more strikingly, when the sample is extended beyond 2012, by just one additional year, the effects become statistically insignificant. This isn’t a minor technical dispute—it represents a fundamental reassessment of the empirical case for raising the minimum wage.
The figure below compares the long-run poverty elasticities of the minimum wage in Dube’s (2019) control set (panel a) and a more empirically robust set of controls (panel b).
Why is the minimum wage ineffective at reducing poverty?
Perhaps the most obvious answer is that few workers are being paid at the federal minimum wage rate. According to 2024 BLS data, only 1% of the workforce makes the federal minimum wage or less. Most of those workers are part-time, and almost half are aged 16 to 24—not a demographic group that is typically associated with impoverished households.
More importantly, most poor families don’t have a full-time, year-round minimum wage earner. According to the March 2020 CPS, fewer than 10% of working-age individuals in poor or near-poor families earn the minimum wage. In other words, raising the wage floor simply misses most of the intended beneficiaries.
Second, the minimum wage can and often does generate unintended consequences like job loss or reduced hours for low-skilled workers, many of whom are on the economic margin. These losses can offset or even outweigh the gains among workers who retain their jobs. This redistributive effect within the low-income population does little to reduce overall rates of poverty.
Third, even among those who benefit, higher wages don’t always translate into higher family income that goes above the poverty line, particularly if other household members reduce work hours or if the increased wages disqualify the household from public assistance.
The authors conclude that if policymakers are serious about reducing poverty, they need to look elsewhere. In the end, minimum wage increases may be politically popular, but they are an ineffective, and statistically insignificant, means of reducing poverty. Policymakers would do well to pivot away from this blunt instrument and acknowledge what the evidence now clearly shows.