Comments by Michael Reich on the Congressional Budget Office report, The Effects of a Minimum-Wage Increase on Employment and Family Income
February 18, 2014
The headline findings in this report are that a minimum wage increase to $10.10 would raise wages for 16.5 million Americans, adding up to a total of $19 billion in pay increases, and lift 900,000 out of poverty. These findings are well explained in the report and its appendices.
Other effects, on employment, employee turnover, public spending, prices, and profitability, are less well-explained, opening the CBO report to important questions.
1. The methods used by CBO to estimate employment losses of 5000,000 cited in the report are never clearly explained. According to the Appendix, they appear to come almost entirely from CBO's "synthesis" of the research literature-- which they conclude implies an elasticity of teen employment of -0.075 percent. This estimate is a bit below the low end of the Neumark-Wascher and Sabia-Burkhauser estimates, but higher than those in various papers by Dube, Reich et al. No details are provided concerning how CBO arrived at this number, except that they discount some older studies slightly because of publication bias.
2. CBO uses an adult minimum wage employment elasticity of about -0.025 percent, even though Neumark and Wascher do not report statistically significant employment effects for adults.
3. As is explained in detail in the Allegretto et al. (2013) paper that the CBO cites, the Neumark-Wascher and other studies that find negative employment effects suffer from methodological flaws that bias their estimates. Our best employment estimates, correcting for the biasing pre-trends in the older studies, are closer to zero. Allegretto et al. also refute Neumark and Wascher's attempts to critique our previous studies.
4. It is disappointing that no mention of the problems in the previous literature is mentioned in the CBO's Appendix, suggesting that the CBO included flawed studies in its estimated "synthesis." CBO thus appears to have counted the studies equally, without making any adjustments for their quality. This is especially surprising since the economic research literature has been divided on how to measure employment effects.
5. CBO arbitrarily increases its estimates of negative employment effects by providing a theoretical argument that longer term effects will be greater than short-term effects. However, no evidence for longer-term effects is cited. In Allegretto et al. we find no such lagged effects, even four years after a minimum wage increase, enough time for businesses to add more capital equipment.
Savings on reduced employee turnover
The report cites the Dube, Lester and Reich paper that finds major reductions in employee turnover. But it is not clear that these savings are included in the CBO study. To do so would require measures of recruitment and retention costs per hire, which are available, but are not mentioned in the report. The savings for businesses provide an important explanation for why minimum wages do not affect employment in high-turnover industries.
The report includes an excellent exposition of why minimum wage increases would have little or nor effect on EITC spending. This conclusion is sensible because EITC benefits increase at very low incomes and then decrease, but do not phase out entirely until at about a $48,000 family income level.
Surprisingly, the CBO report is silent on the effects of a minimum wage increase on SNAP spending. SNAP benefits decline by 30 cents for every additional dollar of income and phase out entirely for a family of three at about $25,000. SNAP spending is thus much more likely to fall because of a minimum wage increase. My forthcoming causal study suggests that the savings to government because of reduced spending on SNAP would be quite substantial.
The CBO report states that minimum wages will increase consumer prices. However, the assumed price effects are not spelled out anywhere in the report, nor is there any reasoned explanation of their effects on consumer demand.
The report cites, but does not at all discuss, research by Aaronson et al. (2008) on price effects. Aaronson's data, which are quite old, indicate that a 39 percent minimum wage increase to $10.10 would increase prices in restaurants on a one-time basis by about 3 percent. Restaurants comprise only about one-tenth of consumer spending. The effects in retail and in accommodations, the other major users of minimum wage workers, are likely to be much less than 1 percent. The report unfortunately does not explore such details.
The report suggests profitability will decline but offers no evidence. The minimum wage research literature contains only one study of profitability effects, based on nursing homes in Britain. We do not know that profits will fall, since cost increases are largely passed passed forward as price increases to customers.
In summary, while the headline numbers in the CBO report are well estimated, a more careful CBO report could have reported that employment effects are much closer to zero, that the effects on public spending are negative,saving taxpayer dollars, and that effects on prices are very modest.