Blog: NFIB's Flawed Minimum Wage Study Doesn't Tell Us Much

Authors: 
Mark Price
Publication Date: 
March 26, 2014

Mark PriceCross posted at Third and State

How do you know that momentum is building in the campaign to raise the minimum wage? Opponents arrive with inflated claims of job losses if we raise the wage.

The latest is from the NFIB Research Foundation in a report claiming that three minimum wage bills before the Pennsylvania Legislature could potentially cost 28,000 to 119,000 jobs as a result of increased labor costs.

To put that wide range in some context, the high watermark of predicted job loss is 11 times higher than the amount of “lost jobs” predicted in a 2005 report from the Employment Policies Institute, a well-known front group for the National Restaurant Association, that was co-released in Pennsylvania with the Commonwealth Foundation.

The 2005 report estimated that the then-proposed 39% increase in the minimum wage (from $5.15 to $7.15 per hour) would reduce employment in Pennsylvania by 10,027 jobs. A prediction like this has never been confirmed in a rigorous examination of actual employment data. 

In fact, after the Pennsylvania Legislature increased the minimum wage in two steps in 2007, job growth remained steady in the sector that relies most heavily on minimum wage workers and would be most impacted by a raise — food service and drinking establishments. It was not until the recession’s impact began to be felt in Pennsylvania in late 2008 that this sector began to lose jobs.
Following the last round of increases in the minimum wage in PA, the sector that relies the most heavily on minimum wage workers, food service and drinking places, saw steady employment growth up until the Great Recession sharply reduced employment in all sectors.
The more recent NFIB study argues that raising Pennsylvania’s minimum wage from $7.15 to $10.10 an hour, also a 39% increase, is projected to cost 11 times as many jobs as the 39% increase proposed in Pennsylvania nearly a decade ago.

The key assumption in all analysis of the potential jobs impact of a minimum wage increase is the percentage change in employment for each percentage point change in the wages of affected workers (technically called an elasticity). Researchers for the past two decades have been using state-level variation in minimum wage laws to observe little or no impact on employment from modest minimum wage increases.

In fact, as John Schmitt at the Center on Economic and Policy Research has argued, the most interesting aspect of the recent report by the Congressional Budget Office (CBO) is that for the first time the CBO acknowledges that the lower range of a potential jobs impact from increasing the minimum wage is zero, a clear sign that careful research has shifted the debate away from assuming large job losses from minimum wage increases.

Table 1 below presents the assumptions in the 2005 report (by David A. Macpherson of Trinity University) and the recent CBO report of the percent change in employment associated with a 10% increase in earnings. You will note that Macpherson assumed that each 10% increase in wages reduced employment by 2.2%. This figure is well above even the high range in the CBO report, meaning that Macpherson was assuming a relatively large employment impact when he predicted the loss of 10,000 jobs in Pennsylvania.

The NFIB study doesn’t disclose the precise relationship that it assumes between wages and employment so we cannot compare its assumptions to the research literature. We do know that job losses are baked into the model NFIB is using in that it assumes no other effects that are known to offset higher hourly wage costs (such as increased productivity or reduced turnover).

So the model the NFIB uses will always give you the same answer: higher wages for those at or near the minimum wage will always lead to job losses, even though there is a large body of research focused on actual state-level minimum wage increases that fails to find employment losses when the wage rises by a few dollars.  

In the end, this study doesn’t include any new information. Its failure to disclose details that would facilitate comparisons with other simulations of a minimum wage increase's employment impact signals pretty strongly that the NFIB would rather you not look too deeply into how the sausage is made. Instead, just repeat after us, hundreds of thousands of jobs lost.

Or as Dr. Peter Venkman might say: ”Human sacrifice! Dogs and cats, living together! Mass hysteria!”
Table 1. For each 10% increase in hourly earnings, what is the percentage change in employment