Comments on Professor Richard Sander's Study of Santa Monica's Minimum Wage Ordinance

By Dr. Jared Bernstein and Dr. Carol Zabin

This document is intended to offer critical reflections regarding the new report by Professor Richard Sander et al. that analyzes the costs and distributional effects of the proposed Santa Monica Wage Ordinance.

We are economists with experience evaluating the impact of such wage mandates. We are thus motivated to comment on some aspects of the Sander report because they are unconvincing and potentially misleading to this important debate. We have also read the study by Professor Robert Pollin and Professor Mark Brenner on the same issue, and are among the group of 120 economists who have reviewed it favorably.

We wish to address two major areas where we think the Sander study is misleading. First, we question his predictions of the overall economic impact on area businesses and employment. Second, we disagree with his claim that the wage increases will go to households who, by dint of their income levels, do not need the added income. In fact, based our reading of the two studies, we believe this policy will in fact provide a needed lift to the living standards of the covered workers.

Economic Impact

First, the Sander study predicts that the ordinance will create economic problems far greater than other studies have shown. The studies of the actual impact of living wage ordinances -- as opposed to predictions such as Sander's -- show very minor costs for businesses and very low job loss.

Sander, on the other hand, predicts "fairly large-scale job losses," the closure of two of the three affected department stores, staff reductions at most of the affected restaurants, profitability declines of over 50 percent, and a decline in local property values. The authors argue that the ordinance "may be fatal to the overall viability of Santa Monica Place," a shopping area within the coverage zone.

The study does not make a convincing case for such dire predictions. Their predictions appear to be driven by estimates of the cost of the ordinance that are far higher than the study by Pollin and Brenner. These discrepancies are based on a number of differences in both data sources and interpretation.

First, the Sander study estimates that the number of firms covered by the ordinance is 100, compared to 47 firms estimated by the Pollin and Brenner study. Based on this difference, Sander's predicted number of affected workers was more than twice the number of affected workers in the Pollin/Brenner study, and his estimates of direct costs due to those wage increases were 1/3 higher ($33 million per year) than those of Pollin's ($21 million).

We think the estimations by Pollin/Brenner are more reliable because they used actual City of Santa Monica records from businesses that filed for business licenses, rather than the commercial database used by Sander.

Second, we believe that Sander greatly exaggerates the so-called "ripple" or indirect effects of the wage mandate. These effects are due to the desire of employers to maintain a wage hierarchy among their employees. Thus, both studies predict that employers will increase wages of those employees who were earning slightly more than the wage mandate so that they will still be earning more than fellow employees below them in the hierarchy.

The difference, again, is in how much. Pollin/Brenner conduct a sophisticated analysis based on a body of empirical research on the effects of minimum wage increases on the wages of near-minimum wage workers. They conclude that the ripple effect would add only about $1.5 million to total costs. Sander, using a made-up formula that has no basis in the literature, estimates a whopping $15 million dollars in extra wages that employers voluntarily give to workers who are already earning more than the mandated wage increase.

Sander et al also assume that firms are much more responsive to cost increases than is conventionally the case in the literature on how employers adjust to such increases in labor costs. The driving factors here are the amount by which labor costs rise, and the employers' ability to absorb, offset, or avoid the higher costs. In addition to inflating the amount by which costs will rise, the report is too dismissive of the various absorption mechanisms.

First, the affected hotels and restaurants cannot easily escape the increase by relocating elsewhere, unless they're willing to give up the prime location, which is presumably one of their draws. They are thus much more likely to try to absorb the increases through either price increases, profit declines, staff reduction, and productivity gains.

The extensive literature on this issue of how firms respond to wage mandates show that the increase is most commonly absorbed through small price increases, slightly lower profit margins, and the productivity gains associated with higher wages (e.g., through less turnover, lower recruitment costs and less absenteeism). Immobile employers have a long history of absorbing wage increases, whether through mandate or through changes in labor market conditions, without closing down shop.

Distributional Impacts

Here the Sander study is particularly on shaky ground. He makes the wild claim that for every dollar in the cost of the ordinance, less than a penny will benefit low-income Santa Monicans. This argument is twofold: many who will benefit from the ordinance do not reside in Santa Monica, and many live in families that don't really need the extra income.

The first argument is simply irrelevant. While we cannot evaluate the reliability of their survey of where these workers reside, it is certainly believable that some, perhaps many, will live outside of the city of Santa Monica. But it is not clear to us why this is a drawback to the policy.

The goal of the ordinance appears to be to raise the earnings and thus the living standards of those who work in the city more so than those who live there. The Sander study does point out that this distinction is "blurred" in the language of the ordinance, and so we do not claim to be speaking for the coalition behind the ordinance on this point. We simply point out that the fact that the benefits will flow to some who live outside the area does not make a case against the policy.

His assertion that most of the benefits will flow to families who do not need the wage increase is also erroneous, we believe. The literature on the actual distribution of the gains from wage mandates consistently shows that most, though not all, of the benefits flow to low-wage workers in low-income families. Again, the Pollin/Brenner study uses a more realistic analysis to assess to whom the benefits of the living wage will accrue. There are two major elements that lead to the differences in their conclusions.

The first issue concerns what is the appropriate income threshold (for various household sizes) that distinguishes poor from non-poor families, so that we can assess whether the wage increase is benefiting a poor or a non-poor family. Pollin/Brenner use a threshold based on the calculation of basic needs developed by the California Budget Project, a non-partisan policy research institute in California.

Sander, in contrast, uses the federal poverty level, which is widely agreed to be an inadequate measure of economic need. It also makes no adjustment for the approximately 25 percent higher cost of living in Los Angeles. Sander also asserts that 6 out of 7 of the wage earners are not the primary bread winners of their households, but presents no data to back this up.

The second issue concerns how many tipped workers will get the wage increase. Pollin/Brenner, again using City data, estimate that only one restaurant will be affected because they meet the 5 million dollar threshold, while Sander estimates that 10 restaurants will be affected. Sander, using data he obtained from the hotels, also asserts that 900 out of 2200 (41 percent) hotel workers are tipped, but other studies and government data show that on average, only about 10 percent of hotel workers receive significant tips, as defined by equaling at least one-half of their total wage.

Sander asserts that tipped workers in the coastal zone earn $20 per hour in tips, in addition to their base wage. While this might be true for high-end bartenders, waiters and waitresses, it surely isn't true for workers who bus tables or clean hotel rooms, whom Sander seems to include in his estimates of the number of tipped workers. Sander thus concludes that tipped workers will account for $13 million of the costs, and that this extra income will go to workers who make $20 an hour in tips and are thus not poor. This conclusion seems quite exaggerated.

Finally, the authors raise the argument that if the ordinance is passed, employers will shed their current employees and hire workers who have higher skills and are from "middle class backgrounds" and "are more likely to speak English." There is, again, no evidence of this type of substitution of one labor force for another in the empirical research that has measured the actual impact of wage mandates. In fact, in this range of wages, there is not much observable difference in the qualities of workers, i.e., the characteristics-demographics, education, experience-of workers earning $7 per hour are not that different from those earning $12 or $13.

In addition, even in low-wage industries, there are fixed costs associated with hires that lead firms to want to hold on to workers, even when they become "more expensive." In fact, in some labor market environments, the fact that workers "cost more" leads employers to invest more in their quality, say through worker training, in order to achieve some of the efficiency gains noted above, which serve to offset the cost increase.

In sum, the Sander et al. study does not, in our view, offer a very useful policy analysis of the proposed ordinance. The estimated costs appear too high, the reactions of firms too "elastic" (i.e., employers and markets are assumed to be much more responsive to the increase than has been the historical case), and the argument that most of the benefits will go to families who are not low-income is unconvincing.