Abstract
A growing number of cities and counties have recently raised their minimum wages. How employers respond to these mandates provides insight into the impact such policies might have on workers and local labor market. Drawing on two survey waves tracking initial responses to Seattle’s $15 Minimum Wage Ordinance by 439 employers with low-wage workers, we show how employers adjusted to higher wages. Most commonly, firms raised prices (56% reported this); smaller percentages reduced employee headcount or hours, limited internal wage progression, or took other measures. Single-site Seattle employers responded similarly to those with multiple sites. Food and accommodation sector employers were more likely to raise prices than firms in other sectors. Relative to other ownership structures, franchises disproportionately reported reducing their workforces. Very few employers reported withdrawing from Seattle. Overall, initial employer responses to this city-level minimum wage law align with predictions from the literature, findings that highlight trade-offs that policy makers must consider in future local wage regulation.
Since 2012, more than 30 cities or counties have enacted local minimum wage requirements that are higher than the federal minimum of $7.25 per hour, compared with just five with minimum wage rates higher than the federal minimum prior to 2012 (UC Berkley Labor Center 2016; U.S. Department of Labor 2017). Higher local minimum wages have been enacted across a wide array of metropolitan areas, including large urban centers such as Los Angeles, New York City, and Chicago, and smaller cities such as Las Cruces, New Mexico, and Tacoma, Washington (Economic Policy Institute 2017). Proliferation of minimum wage regulations is occurring at the same time municipalities nationwide are experimenting with other labor market regulations, including fair scheduling rules and paid sick leave (A Better Balance 2017; National Women’s Law Center 2007).
The impact of higher minimum wage laws rests, in large part, on the anticipated responses of employers. Wage laws require employers to offer workers at the lower end of the wage distribution higher wages, but there are numerous pathways through which employers may adjust business practices to accommodate higher labor costs. For example, it may be that employers respond to higher labor costs by simply redistributing a share of profits from owners to workers’ wages. Economic theory, however, suggests employers are most likely to respond with some reductions to employee hours and headcount. Although debates about methods and data persist, this literature generally finds small negative effects on earnings and employment in the aggregate (Aaronson and French 2006; Belman and Wolfson 2014; Neumark 2014). The modest size of these employment effects support claims that higher minimum wages do not lead employers to dramatically shift employment arrangements or business practice (von Wilpert 2017). Yet, concerns persist that higher minimum wages might lead businesses to raise prices, relocate, or even close (e.g., Employment Policies Institute 2015). Apart from economic impact, there is the possibility that if employer responses to higher minimums create a tipping point in the health of a community’s business climate (Saltsman 2017), then local elected leaders may pay a political cost if higher minimums are perceived by voters to have reduced employment options or increased the local cost of living. Because wage laws hinge so closely to the decisions of employers, we argue it is increasingly important for scholars of urban politics to assess how employers adjust practice and strategy to such wage regulation.
Despite the importance of understanding employer responses to higher minimum wage, conventional data sources are limited in the information gathered about employer location, characteristics, or channels of adjustment. Administrative records collected by state Unemployment Insurance (UI) agencies are the most common source of data for tracking employment and earnings—the outcomes most commonly measured in studies of the minimum wage (e.g., Jardim et al. 2017). Such data, however, cannot geographically locate employment for employers with more than one location who do not report employment disaggregated by location, a limitation of the data that draws criticism (Zipperer and Schmitt 2017). Typical sources of employment data contain limited information about channels of adjustment other than through payroll. Research based on UI data contains no employer characteristics, ruling out analysis on how mandates might affect potentially economically vulnerable employers, such as women- and minority-owned operations or small local businesses. As a result, there are many blind spots in the empirical assessments of the impact of local minimum wage laws.
In this article, we focus on the initial response of employers in the City of Seattle to the first step ups in the local minimum wage rates starting in April 2015, when the minimum wage for workers in the city went from $9.47 to $11 an hour for many employers. To this point, most of the literature has focused on the impact of small increases in statewide or national minimum wage rates on aggregate employment and earnings. We hypothesize that employers may respond to higher wage mandates through many different possible strategies or channels of adjustment (see Hirsch, Kaufman, and Zelenska 2015; Lester 1960). Most obviously, firms adjust internal wage ladders or staffing. Employers also may respond in a number of other ways. For example, employers may raise prices or service fees to bring in additional revenue to cover higher labor costs. We might expect employers to reduce nonwage compensation and benefits. At the extreme, employers facing a city- or county-level wage ordinance also may choose to relocate to other lower cost jurisdictions. Moreover, it seems likely that employers may deploy different strategies over time, or simultaneously pursue responses across multiple channels of adjustment, which is not well-captured in studies focusing on a single channel of adjustment such as employment. The particular response of any given employer, however, is likely shaped by considerations about internal wage ladders, pricing sensitivities, geography, and characteristics of that firm’s particular sector.
Drawing on a unique survey of firms and nonprofit organizations subject to the City of Seattle Minimum Wage Ordinance, this article examines how employers within one city adjust to local minimum wage laws in the short run. In doing so, this work makes several important contributions to the literature on urban politics and labor markets. First, it examines employer self-reported responses to the initial wage step ups of one of the largest increases in a local minimum wage in the United States to date; Seattle was the first major city to pass a wage floor that will reach $15 per hour. Second, the survey data captures adjustment strategies including employment, nonwage compensation, pricing, and location. As such, this evidence expands on the conventional literature surrounding minimum wage laws, which focuses largely on adjustment to wages and employment rather than considering the larger, interconnected set of employer decisions. Finally, we examine whether and how employer responses vary across geographic reach (single- or multisite employers, located within the city or in multiple jurisdictions), sector, and ownership characteristics (franchise status; nonprofit status; and women-, minority-, and immigrant-ownership). Robust local economies include diverse sets of employers who may respond differently to wage interventions, but these characteristics are not available in the administrative data often used to evaluate wage regimes. We conclude with early lessons learned and implications for future research.
Background
Theories about the effects of raising the minimum wage on employment and other factors fall into three general camps that, as a set, yield competing hypotheses. Conventional economic theory holds that the market for labor acts like any other market, with supply rising and demand falling as prices increase. Following this logic, we should expect that raising the minimum wage to a level above the current market-clearing rate constitutes an increase in the price of labor; hence, demand for labor (employment) will fall. One alternative to the classical supply and demand model holds that employers act as a monopsony, a single buyer of labor, relative to the many potential workers (Bhaskar and To 1999; Katz and Krueger 1992). Monopsony power allows employers to keep wages and employment lower than it would be under perfect competition. This theory predicts that raising the minimum wage to a level above the monopsonistic rate should increase employment while redistributing gains from employers to workers. A third body of thought focuses on institutions and social ties, rather than markets, as key determinants of firm behavior, suggesting that changes will be moderated through the interpersonal practices and technologies of production (Kaufman 2010; Lester 1960). Like the monopsony model, the institutional model offers the possibility that a minimum wage increase could increase worker productivity enough to partially or more than offset additional labor costs. Ex ante theoretical models on how the minimum wage might affect prices produce similarly ambiguous competing hypotheses (Lemos 2008). Rising labor costs or shrinking aggregate supply might push prices up; or increased productivity, labor force participation, or lower aggregate demand might lower prices.
Beyond these broad economic theories, the minimum wage literature suggests a set of possible mechanisms or channels of adjustment through which firms could accommodate or respond to higher mandated wages (Hirsch, Kaufman, and Zelenska 2015; Schmitt 2013). Employers often are expected to pursue one or more common strategies: increase prices, reduce hours or employment, or simply live with lower profit margins. Firms also might choose to alter hiring practices, reduce training or nonwage benefits, change wages of nonminimum wage workers, upgrade the skill level required for jobs, or rely more on technology, among other options. Most of the current minimum wage literature, however, focuses on federal or state shifts in hourly minimums. Certain channels of adjustment, such as relocating to a lower cost area, may be more likely or plausible responses to a city-level wage mandate than to a statewide or national increase.
The process through which an employer weighs the advantages and disadvantages of any given response or set of response strategies should vary across firms and industries (Lemos 2008). Raising prices or fees may be more feasible for businesses with relatively inelastic demand for products and services. Other firms may be constrained by elastic demand, pricing schedules across sites or franchises intended to reduce variation, or preexisting contracts or negotiated fee schedules. Employers competing primarily with local businesses may respond differently than employers whose competitors are outside the local area and not subject to higher wage laws. Similarly, some employers may be more mobile or have more options for mobility than others have, which may allow them to exit a local market while other employers stay put. Some business models rest on providing in-person goods or services to customers who work or live in a city. Such location-bound employers might increase prices or change staffing models, rather than take on the risk of relocation. Employers with extant operations in multiple jurisdictions may be more able to shift operations outside of a city that increases wages relative to firms doing business only in the affected area. Other firm or organizational needs also may shape location choice. For example, employers hiring positions requiring a specific type of skill or expertise may locate in places in which qualified workers are more common, regardless of the minimum wage rate. Entrepreneur-owned small firms may intentionally locate close to the home of the owner to reduce daily commute time and to be nearby to respond to an urgent matter. Location may also be particularly important for minority- or immigrant-owned firms, who are thought to rely on enclaves of minority or immigrant-group members to source workers, attract customers, and access capital (Wilson and Portes 1980). One might expect immigrant or minority-owned enterprises may be more likely than employers without those ownership characteristics to remain in a specific city location.
Despite the central role of employer decisions in how local minimum wage laws play out, there is relatively little research on firm responses outside of employment outcomes and very little on employer responses to local wage ordinances in particular. A vigorous body of scholarship examines the impact of federal and state minimum wages on employment, earnings, and prices (see Belman and Wolfson 2014 for a summary of this literature). In contrast, the peer-reviewed evidence on responses to local (city or county) minimum wage increases consists, as far as we are aware, of two studies. Dube, Naidu, and Reich (2007) use data from a survey of restaurants before and after the 2004 San Francisco minimum wage increase and find no statistically significant effect on employment. Allegretto and Reich (2018) use aggregated UI data and Internet restaurant menus to estimate that the 2013 minimum wage increase in San Jose caused a modest price increase of 1.45% while leaving restaurant employment largely unchanged. Other recent research on the effects of local minimum wages focuses largely on employment outcomes and exists as working papers or reports not formally published in peer-reviewed journals (e.g., Jardim et al. 2017; Schmitt and Rosnick 2011; The Seattle Minimum Wage Study Team 2016; Yelowitz 2012).
Our current inquiry advances insight into firm responses to local wage ordinances by focusing on multiple channels of adjustment across multiple industries and types of employers. We specifically examine employer responses to the Seattle Minimum Wage Ordinance (hereafter, the Ordinance) enacted by the Seattle City Council in June 2014 after a robust public discussion and an election in which the successful mayoral candidate and a council candidate endorsed a $15 wage (Balk 2014). As summarized in Table 1, the Ordinance initially set out two different phase-in schedules, one for smaller employers (those with 500 or fewer workers across all business locations worldwide) and one for larger employers. Within each category, employers can count the value of employer-provided health insurance toward the required hourly wage, and smaller employers can count tips. The first wage step up—from the state minimum of $9.47 to $11.00—took effect as of April 1, 2015, and the second step up, to $12.00 for smaller employers and $13.00 for larger employers, took effect January 1, 2016. Despite the complexity of this four-pronged schedule, interviews with employers and workers suggest that the highest wage at each step up prevailed (The Seattle Minimum Wage Study Team 2016).
Seattle’s Minimum Wage Ordinance Phase-In Schedule.
Note. After the minimum wage reaches $15.00/hour, it will be adjusted each year on January 1, based on the Consumer Price Index for the Seattle Tacoma Bremerton Area.
Small employers are defined as those employing 500 or fewer employees worldwide; large employers are those with more than 500 employees worldwide.
The initial years of Ordinance phase-in coincided with a period of robust local economic growth. Over the two periods covered here, 2014–2016, King County, in which Seattle is located, added 78,000 jobs, about a 6.3% increase from baseline. Average annual wages over that period grew 8.8% to $76,828 (WA Employment Security Department 2017). Much of this growth was concentrated in high-earning jobs and jobs in the City of Seattle. In 2016, Amazon.com alone listed 19,766 job openings in Seattle, which—if filled—would account for a nontrivial share of total growth (Fung 2017). This influx of high-paying professional jobs likely increased demand for some goods, such as restaurant meals and personal services. More broadly, we should be cautious in evaluating impacts of minimum wage laws based on the experience of any one place. Findings from a single case may be affected by spurious trends and hence should not necessarily be generalized (Dube, Lester, and Reich 2010).
In this article, we consider the impact of the first and second wage step ups, which lifted the minimum wage from $9.47 to between $10.50 and $13 depending on where employers fell in the phase-in matrix. Our main hypothesis is that (hypothesis 1) the Ordinance mandated a significant enough change that employers reacted through one or more channels. Although we cannot test the relationship between the growth in the city’s economy and observed changes, we believe the Seattle case is one in which price changes are likely due to the likely increase in demand. In addition, we test a set of secondary subhypotheses about how adjustment varies by firm characteristics including geographic reach, sector, and ownership. Specifically, multisite employers with sites outside Seattle were more likely to consider relocation and less likely to consider price and labor adjustments as compared with firms solely located within the city (hypothesis 2a); that employers in the food and accommodation sector are more likely to raise prices and less likely to pursue other channels of adjustment relative to firms in other sectors (hypothesis 2b); and that locally owned and immigrant-owned employers are less likely to relocate and more likely to adjust staffing or prices relative to firms without these characteristics (hypothesis 2c).
Data
To examine these hypotheses, we fielded the Survey of Seattle Employers (SSE), a survey of Seattle area businesses aimed at understanding how businesses and the nonprofit sector are responding to regional wage increases. The SSE survey sample was drawn from the universe of 90,481 City of Seattle business license holders registered in December 2014, a publicly available data set. Employers listed as sole proprietors were excluded since such employers are unlikely to have employees subject to the Ordinance. This design decision reduced the number of records to 52,643. Businesses with more than one branch appear in business license data multiple times. We removed these duplications so each “business legal name” contained one record as well as a flag for having had multiple locations within the city. The final population of Seattle businesses used for sampling numbered 48,962. We drew a random sample of firms from each of eight strata formed from four industry sectors and two size groups. North American Industry Classification System (NAICS) codes were used to define the four sectors: retail/trade, manufacturing, and food/accommodation, and “all other”—which included firms from all other NAICS codes, including agriculture, construction, real estate, finance, entertainment, and waste management. We attempted to survey more large employers by oversampling employers with more than one branch within the City, the only proxy for size available in the data. The final random sample of employers we selected to survey totaled 3,780 employers.
The Survey Research Division of the University of Washington’s Social Development Research Group completed an initial baseline survey of eligible employers between March and May 2015. Survey staff recruited business owners or senior managers knowledgeable of the employer’s business practices. In the screening phase, the survey research team successfully reached 80% of the sample (3,011 of 3,780 employers) by phone. Screening calls focused on identifying a respondent and determining whether the firm or organization employed workers earning less than $15 an hour at the time of the interview. In some cases, the person answering the phone did not know the needed wage information; these cases were screened-in for further follow-up. Of the employers contacted, 1,119 reported that they had or possibly had workers in Seattle earning $15 per hour or less as of the time of the call and were considered eligible for the survey. In the survey phase, the data collection team then attempted to contact all eligible or possibly eligible employers for a phone or web survey that collected detailed information about business practices and responses to the local wage ordinance. The team successfully contacted 693 employers (62% of 1,119) in the survey phase, and were able to exclude 127 employers who definitively reported that they did not have low-wage workers. Excluding these employers yielded 566 baseline respondents.
The same survey group conducted a second survey one year later between June and August 2016 with respondents to the initial baseline survey. The follow-up survey was web-based and gathered responses from 443 employers. An additional three employers stated they were no longer in business and one employer was a duplicate. This lead to a wave 2 retention rate of 78% and final analytical sample of 439. Multivariate analysis of respondents versus nonrespondents to wave 2 shows that response was unrelated to most tracked business characteristics and nonresponse did not obviously reflect attrition due to firm closure. 1
The SSE collected detailed information about employers in Seattle. We categorized employers into industry types according to NAICS codes accompanying the business license data, but also through confirmation of those codes with respondents. We asked employers about the number of Seattle and national employees, which helps determine which minimum wage rate employers are required to pay during phase-in. We also asked respondents if they operated in a single-site location (SS), at multiple sites within Seattle (multisite or MSS), or in multiple sites across municipal jurisdictions in the metro area or elsewhere in the United States (multijurisdiction or MJ). Providing additional information about the spatial location of business, employers also indicated whether they provide goods and services to customers located inside of Seattle, outside the city, or a mix of both. We also asked for-profit firms if the business was a franchise, family-owned, woman-owned, minority-owned, and/or immigrant-owned, using a rule of 51% or more of ownership, not official certification.
Table 2 summarizes the characteristics of SSE employers included in this analysis. Nearly half of respondents were single-site employers (46%), slightly less than one-quarter are multisite Seattle employers (22%), and just under one-third reported being multijurisdiction (31%); this distribution suggests we successfully oversampled larger employers. 2 Overall two-thirds of the sample reported at least half of their employees earned less than $15 per hour at baseline. Employers in the accommodation and food sector comprised almost 40% of all employers interviewed, followed by retail and trade (25%), and manufacturing (7%). Just over two-thirds reported that over half of their Seattle employees were paid less than $15 in 2015. 3 About 11% of employers interviewed self-identified as a charitable nonprofit organization. The median employer reported 15 employees in the City of Seattle, with the median single-site employer having nine employees and the median multijurisdictional employer having 31 employees. Nearly two-thirds of for-profit companies interviewed—63%—were family-owned firms and just under a quarter identified as women-owned (25%) or minority-owned (22%). Nearly one in eight employers interviewed—12%—reported being a franchise.
Descriptive Statistics for Sample, by Employer Geographic Reach.
Note. Sample includes N = 439 eligible respondents participating in wave 1 and wave 2. Values are percentages unless otherwise indicated.
The baseline survey was fielded in advance of the city’s writing of the Minimum Wage Ordinance regulations. Thus we ask about national employees, which was consistent with the understanding of the law in the months leading up to its enforcement. The current language of the law has been amended to include employees worldwide.
Firm demographic characteristics included only for for-profit firms (total n = 391).
Next, we asked survey respondents about their familiarity with the minimum wage ordinance: As of April 1, 2015, the City of Seattle requires firms with workers operating in the city to provide a higher minimum wage to employees than was previously mandated. Is {Business Name} aware of the City of Seattle minimum wage requirement?
As we show in Table 3, a strong majority of employers—94%—were aware of the minimum wage mandate at the time of the 2015 survey.
Reported Responses to MWO in 2015 and 2016 by SS, MSS, and MJ Employers.
Note. MWO = minimum wage ordinance.
Plan to do, do not plan to do, do not know, refuse, logical skip, and does not apply are included as not done.
This question was only asked in 2015. N = 430.
n = 429.
Increase hourly earnings for employees earning between $13 and $15 per hour.
Increase hourly earnings for employees making more than $15/hour. Only asked in 2016.
Raise hourly wage rate for employees working outside the City of Seattle who are paid the state minimum wage—only applies to MJ firms.
Limit raises or decrease wages for employees earning more than minimum wage.
Limited to firms who offered any benefit in 2015 (n = 353).
p < .10. *p < .05. **p < .01.
Particularly important for our purposes, each wave of the survey asked respondents a series of questions about how they were adapting business practice or strategy to the Ordinance: “Have you made or do you intend to make any of the following changes to accommodate this new policy?” Our analysis focuses on answers to these questions in wave 2, after the second step up in the wage ordinance. Employers provided self-reports about action taken or planned across a wide variety of potential strategic responses. Respondents indicated if they had or planned to raise wages of one or more Seattle employees. They also reported on whether they were raising wages for workers not bound by the minimum wage, a phenomenon observed as “spillover” in the larger literature (Belman and Wolfson 2014). Spillover questions use the survey language of “increase hourly earnings for employees earning . . .” with separate questions for just above the 2016 minimum (“between $13 and $15 per hour”) and higher amounts (“more than $15 per hour”). We asked multijurisdiction employers if they had raised the hourly wage rate for employees working outside the City of Seattle who are paid the state minimum wage. We also asked employers to report if they had taken steps to limit wage progression using the survey language “limit raises or decrease wages for employees for earning more than the minimum wage.” Employers indicated whether they had reduced hours for minimum wage employees who work inside the City of Seattle or had reduced headcount and cut back the number of employees in Seattle. Respondents indicated if they had withdrawn business sales or services from the City of Seattle. Survey items also captured whether employers were contracting out for work currently or previously provided in-house, had raised prices on goods or services, or added fees or service charges specifically meant to offset the wage mandates. Finally, employers were asked if they had taken steps to eliminate benefits (e.g., health insurance coverage, paid vacation) for some employees.
Analysis begins with descriptive statistics about the prevalence of changes in response to the ordinance. Using paired tests of equal proportions, we test for differences between employers with a single Seattle site, those with multiple sites in Seattle only, and those with sites in Seattle and other jurisdictions. Because some important employer characteristics are correlated—for instance, employers with only one location are also often smaller—we use multivariate analyses to better tease out the contribution of key employer and business model characteristics to employer decisions. We estimate a series of logistic regressions to examine three possible summary outcomes: whether an employer pursued any of the eight channels of adjustment, whether an employer raised prices or fees, and whether an employer reduced hours or headcount. Estimates for individual outcomes are provided in the appendix. Covariates include a number of firm and business model characteristics that may shape responses to local wage ordinances, including indicators for sector, whether the firm is part of a franchise, total number of employees, whether more than half of the employees made below $15 per hour at baseline and whether firms report serving customers in person at their Seattle sites. We also include dummy variables to reflect whether an employer was a multisite firm in Seattle only or a multisite firm with operations inside and outside of Seattle, and captures the unobserved heterogeneity in employer response.
Findings
Table 3 reports the share of all employers indicating that they had “done” or pursued a given strategic response to the higher minimum wage. Using tests of equal proportion, we tested the hypothesis that the proportion of employers reporting having taken each strategy was equivalent across single-site employers, employers with multiple sites in Seattle only, and multijurisdiction employers. About half of respondents, 52%, reported having raised wages in response to the Ordinance at the time of the 2015 survey, which was conducted in the months leading up to and spanning the implementation date. Multisite Seattle employers are more likely than single-site employers and multijurisdiction employers to have reported having already raised wages at the time of the baseline survey (64% versus 48% and 50%, respectively).
After the second increase took effect in 2016, eight in 10 employers reported having raised wages. In wave 2, however, the proportion of employers who raised wages does not vary significantly across single-site, multisite, or multijurisdiction employers. Half of employers, however, reported efforts to decompress wages by raising the hourly rate for workers earning between $13 and $15 an hour and 30% indicated raising wages for workers making more than $15 per hour. Combined, these efforts constitute positive spillover of the policy to higher wage levels. Multisite employers were marginally more likely than single-site Seattle employers to have increased wages above $15 per hour. Among the 138 multijurisdiction employers, 20% reported having raised the wages of employees outside of Seattle at wave 2 in response to the Ordinance.
The bottom panel of Table 3 reports the frequency of other possible channels of adjustment. Respondents most often reported raising prices (56%). Other possible strategic responses or channels of adjustment were reported less frequently. For example, about one in five respondents indicated they had reduced labor costs through cutting employee hours or reducing headcount. Roughly one in eight employers reported limiting wage progression for employees earning just above the new minimum wage levels. A far smaller share of businesses reported having contracted out work (6% of all employers). Among the 353 employers who reported offering benefits at baseline, 6.2% reported eliminating a benefit; 11% of multijurisdiction employers reported dropping benefits, a rate that was significantly higher than that of Seattle-only firms. Fewer than 4% of employers reported having withdrawn from the City of Seattle.
The bottom two rows of Table 3 show the proportion of employers who reported making any of the eight changes or more than one. Across all firms, about two-thirds of employers (65%) reported adjusting at least one element of their business practice in response to the minimum wage ordinance. About half of all employers (32%) reported two or more responses. Overall, we find little evidence that single-site, multisite, or multijurisdictional employers reported different responses to the higher minimum wage.
Finally, we assess the relationship between employer characteristics and self-reported responses through several different logistic regression models. Table 4 displays coefficients and odds ratios from multivariate models estimated using logistic regression. Models estimated with penalized maximum likelihood (Firth’s correction) for low-incidence outcomes produced similar results. For discussion in the text, we translated key results to predicted probabilities (using the marginal standardization default approach in Stata).
Coefficients and ORs from Logistic Models, 2016 Reported Channels of Adjustment on Firm Characteristics.
Note. Logistic models also estimated with Penalized Maximum Likelihood Estimation to account for low-incidence events produced similar results. OR = odds ratio.
“Any changes” includes limit wage progression, reduce hours, reduce headcount, withdraw from city, contract out, raise prices, add fees, and eliminate a benefit.
The baseline survey was fielded in advance of the city’s writing of the Minimum Wage Ordinance regulations. Thus we ask about national employees, which was consistent with the understanding of the law in the months leading up to its enforcement. The current language of the law has been amended to include employees worldwide.
p < .10. *p < .05. **p < .01. ***p < .001.
Sector and some firm ownership characteristics are significantly associated with making any changes, changing prices, or changing employment. For employers in the food and accommodation sector, the odds of reporting having made any changes is 1.8 times higher than employers in the omitted category (“all other”). This odds ratio translates into a 71% predicted probability of food and accommodation sector employers reporting any change compared with a 50% predicted probability of omitted employers, after controlling for all other predictors. Model 2 results show that these employers are responding via the price channel. A predicted 69% of food and accommodation employers report raising prices or adding service fees compared with 46% of “all other” employers. Additional analyses not reported here also show that employers in the food and accommodation sector are significantly more likely to report raising prices or adding fees than employers in the retail or manufacturing sectors.
Net of the industry effects, a predicted 71% of family-owned employers report making any changes, compared with a predicted probability of 58% for nonfamily-owned employers (odds ratio of 1.8). Family-owned employers have a 62% predicted probability of reporting raising prices or adding fees, compared with a predicted 48% of nonfamily-owned employers. A predicted 48% of immigrant-owned employers report making any changes compared with a predicted 67% of employers with native-born ownership. Similarly, a predicted 36% of immigrant-owned firms report raising prices or adding fees, compared with a predicted 59% of employers with native-born citizenship. Employers that reported being part of a franchise have a 41% predicted probability of reporting reducing employee hours or headcount, while nonfranchised employers have a 21% predicted probability.
Although we anticipated that adjustment strategies might vary by geographic reach, this did not appear to be the case in the multivariate (Table 4) results. Single-site Seattle employers, employers with multiple sites in Seattle, and employers with multiple locations including some outside of Seattle did not report significantly different overall responsiveness nor price or employment responses. Models of the eight individual adjustment strategies separately (in the appendix) do support the one significant difference and a slight trend visible in the descriptive (Table 3) results. Specifically, multisite Seattle and multijurisdiction employers are both slightly more likely than single-site firms to report withdrawing business sales or services from the city, although these effects are marginal (Appendix model 6, p < .10). Controlling for other characteristics, multijurisdiction employers are more likely than single-site Seattle employers to report eliminating a benefit as well. These adjustments may represent real changes for the small number of firms that make them, but overall they are rare events.
Limitations
Before proceeding to a discussion of our research findings it is important to note limitations of our analysis due to the data source and sample size. First, survey data based on self-reports may be inaccurate due to random error, socially acceptable response bias, or politically motivated strategic response. In particular, because the minimum wage issue was controversial within Seattle, it may be that employers viewed our survey as a way to express to the City their displeasure about the wage mandate and thus overstated the extent to which they were taking steps to modify employment arrangements, prices, or other aspects of business operations. Although we cannot directly rule out error due to strategic response, patterns within our data and parallels to other data reassure us that this is not a large concern. We do not see inconsistencies in employers self-reported responses over time, which one might expect if self-reports were not tied to business practice. In-depth interviews with workers, nonprofits leaders, and for-profit employers produced for other parts of the Seattle Minimum Wage Study corroborate survey findings reported here (The Seattle Minimum Wage Study Team 2016, 2017b). Our findings also align with analyses of UI administrative data to assess the impact of the minimum wage ordinance in Seattle (Jardim et al. 2017; Jardim and van Inwegen 2017).
Second, despite a large initial sample, strong survey recruitment effort (the team made more than 48,000 call attempts across the first two waves), and robust sample retention, our sample size of 439 coupled with item-specific nonresponse limits the types of subgroup analysis possible. In addition, differences across employer subgroup characteristics may, theoretically, be modest necessitating larger sample sizes to detect small effects. Therefore, it is likely our results are at increased likelihood for Type II errors as evidenced by power simulations, given our analytical sample size, and the standard errors associated with the binary outcomes.
Moreover, our analysis is limited by the survey items themselves. By tracking binary responses to a set of possible responses, this work can describe trends broadly but lacks detail about the intensity of responses. For instance, while we know that over half of the employers sampled raised prices, this information does not tell us much about how the price increases will affect consumers. We do not know if employers raised prices across-the-board or targeted price increases to a few products that might appeal to less price-sensitive customers. Similarly, when employers reported reducing employee headcount or hours, we do not know the magnitude of these reductions nor whether some types of employees were more likely to face decreased work opportunities. In some cases, research using other data sources can illuminate some of these magnitude questions; we discuss the cases of employment reductions and price increases below. Survey questions also did not address all possible channels of adjustment. For instance, we do not know if employers invested more in training or substituted higher paid and higher skilled workers for now relatively more expensive minimum wage employees.
Discussion
This article examines how employers with low-wage workers in Seattle report responding to initial step ups in local minimum wage rates between 2014 and 2016. Our findings are based on a unique survey of a random sample of employers of firms most likely to be exposed to higher labor costs due to the ordinance. Apart from details about firm characteristics and activity along a number of different channels of adjustment, our survey data follow firms over time to see how responses evolve or shift. We capture firms across a range of industry sectors, with different ownership arrangements, and with differing exposure to the law according to the share of low-wage workers. Survey respondents represent multiestablishment and single-site firms, which allows us to compare employers located solely in the City of Seattle to those with sites inside and outside the city.
Overall, our findings compare predictably with prior research on the minimum wage—although focus on firms in the context of a city-level intervention often means there are no easy parallels among extant work. Perhaps not surprisingly, we find that Seattle employers knew about and complied with the minimum wage ordinance by raising wages. We find evidence that firms with more than one location within Seattle initially raised wages earlier than employers with only one Seattle location or than firms with locations inside and outside the city. By the second year of ordinance phase-in, there were no differences in reported wage increases between firms with different geographic footprints. Apart from raising wages of employees below the new mandated minimums, we find evidence that many firms (almost half) raised wages of employees above the newly set minimum wage rates, the positive spillover sometimes observed in the larger minimum wage literature (Aaronson, Agarwal, and French 2012; Luttmer 2007). These actions reflect efforts to avoid wage compression and are consistent with findings elsewhere that employers seek to maintain internal wage ladders in the face of higher minimum wage rates (Knudsen 2017).
As expected, our findings indicate that most Seattle employers surveyed have adjusted business practices in response to the minimum wage ordinance. Nearly two-thirds of employers reported making some change in business practice to accommodate the higher minimum wage, and three in 10 made two or more changes. Our survey data cannot convey the magnitude of employer business practice responses to higher minimum wages, but we see a sizeable percentage of firms that reported pursuing more than one of eight possible channels of adjustment presented in our survey. Contrary to our expectations, however, patterns of adjustment did not vary significantly by employers’ geographic reach on channels tracked with the exception of benefit reductions. Single-site Seattle employers, those with multiple locations in Seattle only, and those with locations in Seattle and other jurisdictions reported similar patterns across the types of adjustments tracked.
About half of all employers reported raising prices to offset increased labor costs. We find that price increases were far more likely to occur in the food and accommodation sector than other types of industry sectors. There also is evidence that immigrant-owned employers were less likely to increase prices, perhaps a reflection of business models reliant on serving coethnics who may be more price sensitive. Although our data cannot speak to the magnitude of these price increases, a concurrent study of menu prices found that full-service restaurant prices increased 7.7% over the first year (The Seattle Minimum Wage Study Team 2016). Consistent with our findings that retail firms did not increase prices, another study found no impact of the minimum wage on grocery store prices (Otten et al. 2017). Taken together, we posit that price increases probably fell most heavily on the relatively higher income consumers who ended up paying more for food in full-service venues.
Fewer than one in four employers reported reducing their workforces through cuts in hours or headcount. Franchises were more likely than nonfranchises to report workforce reductions, perhaps because they have less price-setting ability than independent businesses. The overall findings about workforce reduction are consistent with early analyses based on UI data that find reductions in overall employment (Jardim et al. 2017; The Seattle Minimum Wage Study Team 2016). It should be noted, however, that our binary survey measures do not capture the magnitude of possible employment effects. Table 4 shows that employers in the food and accommodation sector were significantly more likely to raise prices than employers in other sectors, but not particularly likely to reduce hours or headcount is consistent with studies finding no reductions in employment among the hospitality sector (Reich, Allegretto, and Godoey 2017). Also important given the methodological debates about whether to include multiestablishment firms in UI data analyses (see Jardim et al. 2017), we do not find that single-site firms differ from multiestablishment firms in their self-reported responses to the higher minimum wage. Critics noted that excluding multisite firms from analyses of UI data may overstate the extent to which the minimum wage ordinance caused reductions in employment (Zipperer and Schmitt 2017). If anything, our evidence weakly suggests the opposite might be equally or more likely. Relative to single-site Seattle employers, employers whose operations spanned multiple jurisdictions reported reducing hours or headcount at a greater rate (21% of Seattle only versus 28% of multijurisdiction, p = .143).
Together, price and employment changes account for a large proportion of the reported adjustments. Fewer than one in 20 employers reported withdrawing business sales or services from the city, a response primarily concentrated among multisite employers. We did not find widespread evidence that employers were eliminating benefits to reduce total compensation, but again such adjustments were found to be mostly concentrated among employers with multiple locations.
Our findings should be read as reactions to the initial step ups in Seattle’s minimum wage that occurred in particular labor market context. The period covered by our data has the Seattle minimum wage stepping from the 2014 state minimum of $9.47 to at most $13. The 2017 step up to $15 for large employers will create greater payroll costs, and we expect more employers will use the strategies we track in response to the 2017 increase. As noted in the introduction, the initial phase-in of higher minimum wages occurred in a rapidly growing Seattle economy, where historic rates of high-skilled job growth and population growth have corresponded with rising rent and housing costs (Balk 2016). The influx of high-paid professionals into Seattle—many younger adults employed in the technology sector—has provided businesses with a growing pool of customers with significant disposable income. Such economic trends may have buffered many businesses from the higher labor costs.
These findings are consistent with expectations of the current literature and other research findings produced by the Seattle Minimum Wage Study. Yet, we believe our findings are of great scholarly value. First, our findings show how prevalent certain types of responses to higher minimum wage laws are across a local economy. But, our comparisons of firms across industrial sectors and across multiple channels of adjustment also show that certain types of firms are more likely to pursue certain strategies. Moreover, the wave of local minimum wage laws being passed across the country represent a shift from past federal or statewide increases. Our results here advance understandings of how employers respond to local minimum wage ordinances that go much higher on a faster timetable than previous federal or statewide minimum wage increases commonly studied. Often times the advocates who support these local minimum wage laws dismiss that substantial changes to local wage structures for the lowest-wage portion of the workforce will dramatically affect business practice or the consumer experience. We think these are empirical questions that should be explored through a variety of high-quality methodological approaches.
Second, economic theory and a large portion of the research literature suggest we should expect price increases and reductions in jobs or work hours. Although we feel it important for researchers to focus on outcomes for workers and extrapolate employer responses from worker-level data, we feel there is value in directly exploring how internal business practice or strategy may have changed in the wake of higher local minimum wage laws. We believe that the manner in which firms respond matters to workers and the surrounding community. Certain firms may have many channels of adjustment to choose from, others may have relatively few. Reductions in hours and jobs affect low-wage workers directly, but may not be salient to consumers. Increases in prices and fees, however, are more easily discerned by consumers. Although we do not find many firms relocate in the immediate wake of the higher minimum wage, we should expect that mobility varies by industry or business model. Firms that serve local customers via face-to-face interactions, such as is common in the hospitality industry may face less pressure to relocate than manufacturing firms whose primary competitors are located outside local jurisdictions enacting higher minimums. Our data contain several hundred firms split across industries and observed over two years; observing more firms over a longer time horizon may better reveal relocation patterns.
Our findings suggest several future avenues for urban researchers concerned with labor markets and workplace regulation. First, it will be important to compare the initial experiences in Seattle alongside comparable early phase-in windows in other locales. Our study highlights the importance of commissioning high-quality evaluation research prior to the implementation of higher minimum wages to add to the body of knowledge around firm responses to this new generation of workplace regulation. Other cities may have different experiences given the local economic circumstances and the nature of nonlabor costs, and we hope similar efforts track these other experiments. Second, while the current analysis focuses only on experiences within the city that raised wages, future research should track employer behavior in parallel across neighboring places. Such work can serve two important purposes: showing how effects spillover into affected contiguous areas, and helping isolate impact through the experiences of nonaffected comparisons areas. Third, our survey data only roughly capture business decisions to raise prices. Future scholarship should explore study designs that can more accurately capture the magnitude of price changes across a wide range of products, services, and settings. In addition, because minimum wage laws may be one part of a multipronged effort by local communities to create more just or equitable workplaces, great care should be taken to explore how minimum wage laws might interact with other workplace regulations, such as secure scheduling or paid sick leave initiatives.
Finally, we believe there are takeaways from this work relevant to communities weighing the adoption of higher local minimum wage laws. Although claims of widespread job loss or firm exit largely represent political rhetoric, it is not realistic to assume that firms can implement higher wages without affecting prices, employment, or other factors important to citizens. In particular, policy makers should attend to firms and organizations that may be most acutely affected by higher local minimums or that may have more limited responses. For example, our work suggests that immigrant-owned businesses were less likely to raise prices in response to initial step ups in Seattle’s minimum wage, which may complicate their ability to remain viable businesses as future step ups occur. Other work by the team points to the unique challenges faced by nonprofit human service organizations, whose revenue streams often are set by contracts and fee arrangements at the federal or state-level that are not responsive to local wage laws (The Seattle Minimum Wage Study Team 2017a). Policy makers should consider how to support workers or new entrants to the labor market in particularly affected industrial or economic sectors. Such efforts would help to minimize any negative consequences for vulnerable workers. Finally, in the interest of balancing workers’ and employers’ needs policy makers should consider how higher minimum wage laws may interact with other workplace regulations, such as compounding costs upon employers or limiting the channels of adjustment available to firms.
Footnotes
Appendix
Coefficients and ORs from Penalized Maximum Likelihood Logistic Models, 2016 Reported Channels of Adjustment on Firm Characteristics.
| Firm Characteristics | Model 1: Raise Prices |
Model 2: Add Fees |
Model 3: Reduce Hours |
Model 4: Reduce Headcount |
||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Coefficient | SE | OR | Coefficient | SE | OR | Coefficient | SE | OR | Coefficient | SE | OR | |
| Sector (base: all other) | ||||||||||||
| Food/accommodations | 0.974** | 0.312 | 2.648 | 0.252 | 0.458 | 1.287 | −0.186 | 0.402 | 0.831 | −0.060 | 0.388 | 0.942 |
| Retail/trade | 0.034 | 0.302 | 1.034 | 0.224 | 0.469 | 1.252 | 0.236 | 0.391 | 1.267 | 0.247 | 0.379 | 1.280 |
| Manufacturing | 0.378 | 0.443 | 1.460 | −0.372 | 0.771 | 0.689 | −0.930 | 0.746 | 0.395 | −0.316 | 0.606 | 0.729 |
| Nonprofit | 0.014 | 0.406 | 1.014 | −0.474 | 0.609 | 0.623 | −0.975 | 0.668 | 0.377 | −0.333 | 0.622 | 0.716 |
| Family-owned | 0.632** | 0.244 | 1.881 | −0.027 | 0.344 | 0.974 | 0.078 | 0.294 | 1.081 | 0.466 | 0.305 | 1.594 |
| Woman-owned | −0.402 | 0.272 | 0.669 | −0.827 † | 0.473 | 0.438 | −0.339 | 0.351 | 0.713 | 0.131 | 0.320 | 1.140 |
| Minority-owned | 0.305 | 0.316 | 1.357 | 0.839* | 0.415 | 2.314 | −0.724 † | 0.404 | 0.485 | −0.275 | 0.369 | 0.759 |
| Immigrant-owned | −1.119** | 0.373 | 0.327 | −0.560 | 0.552 | 0.571 | 0.697 † | 0.415 | 2.008 | 0.384 | 0.411 | 1.468 |
| Franchise | 0.495 | 0.372 | 1.641 | 0.426 | 0.467 | 1.532 | 1.213*** | 0.359 | 3.365 | 0.817* | 0.361 | 2.264 |
| Site number and location (base: single site) | ||||||||||||
| Multisite in Seattle | −0.013 | 0.291 | 0.987 | 0.038 | 0.407 | 1.038 | −0.103 | 0.364 | 0.902 | −0.261 | 0.362 | 0.771 |
| Multijurisdictional | −0.066 | 0.288 | 0.937 | 0.018 | 0.410 | 1.018 | 0.111 | 0.345 | 1.117 | 0.235 | 0.334 | 1.265 |
| Provide goods/services to customers in Seattle | −0.093 | 0.211 | 0.911 | 0.020 | 0.305 | 1.020 | 0.021 | 0.263 | 1.021 | 0.482 † | 0.252 | 1.619 |
| Over 50% employees paid less than $15 in 2015 | 0.510* | 0.240 | 1.666 | −0.392 | 0.360 | 0.676 | 0.867* | 0.347 | 2.380 | 0.641* | 0.324 | 1.899 |
| Log of number of Seattle employees | 0.215* | 0.094 | 1.240 | 0.312* | 0.136 | 1.366 | 0.126 | 0.116 | 1.135 | 0.091 | 0.113 | 1.096 |
| 500 or more national employees a | −0.706 † | 0.401 | 0.493 | −1.248 † | 0.670 | 0.287 | −0.642 | 0.508 | 0.526 | −0.331 | 0.500 | 0.718 |
| Constant | −1.233** | 0.430 | 0.291 | −2.632*** | 0.641 | 0.072 | −2.366*** | 0.573 | 0.094 | −2.744*** | 0.561 | 0.064 |
| Pseudo R2f | .138 | .045 | .098 | .069 | ||||||||
| Total N | 439 | 439 | 439 | 439 | ||||||||
| Firm Characteristics | Model 5: Limit Wage Progression |
Model 6: Withdraw from City |
Model 7: Contract Out |
Model 8: Eliminate a Benefit
c
|
||||||||
| Coefficient | SE | OR | Coefficient | SE | OR | Coefficient | SE | OR | Coefficient | SE | OR | |
| Sector (base: all other) | ||||||||||||
| Food/accommodations | 0.111 | 0.443 | 1.117 | −0.171 | 0.744 | 0.843 | 0.889 | 0.767 | 2.434 | −0.066 | 0.655 | 0.936 |
| Retail/trade | 0.075 | 0.450 | 1.078 | −0.382 | 0.700 | 0.682 | 1.154 | 0.714 | 3.169 | 0.244 | 0.621 | 1.276 |
| Manufacturing | −0.297 | 0.763 | 0.743 | −0.071 | 1.041 | 0.931 | 1.923* | 0.802 | 6.843 | −1.037 | 1.535 | 0.355 |
| Nonprofit | −0.056 | 0.615 | 0.946 | 0.630 | 0.982 | 1.878 | −0.297 | 1.027 | 0.743 | 0.207 | 0.928 | 1.230 |
| Family-owned | −0.070 | 0.331 | 0.932 | 0.520 | 0.666 | 1.682 | −0.045 | 0.473 | 0.956 | 0.356 | 0.582 | 1.428 |
| Woman-owned | −0.304 | 0.408 | 0.738 | −0.173 | 0.636 | 0.841 | 0.571 | 0.519 | 1.770 | −0.190 | 0.660 | 0.827 |
| Minority-owned | −0.299 | 0.434 | 0.742 | −0.212 | 0.779 | 0.809 | −0.161 | 0.725 | 0.851 | −0.064 | 0.716 | 0.938 |
| Immigrant-owned | 0.614 | 0.456 | 1.847 | −0.120 | 0.912 | 0.887 | −2.184 | 1.545 | 0.113 | −0.191 | 0.831 | 0.826 |
| Franchise | 0.140 | 0.434 | 1.150 | 0.496 | 0.703 | 1.642 | 0.992 | 0.678 | 2.697 | 0.760 | 0.596 | 2.139 |
| Geographic reach (base: single site) | ||||||||||||
| Multisite in Seattle | 0.155 | 0.399 | 1.167 | 1.445 † | 0.739 | 4.241 | −0.016 | 0.630 | 0.984 | 0.090 | 0.725 | 1.094 |
| Multijurisdictional | 0.367 | 0.390 | 1.443 | 1.250 † | 0.688 | 3.492 | −0.654 | 0.662 | 0.520 | 1.219* | 0.581 | 3.382 |
| Provide goods/services to customers in Seattle | −0.060 | 0.294 | 0.942 | 1.118* | 0.524 | 3.059 | −0.050 | 0.428 | 0.952 | 0.888* | 0.450 | 2.429 |
| Over 50% employee paid less than $15 in 2015 | 0.840* | 0.382 | 2.317 | −0.452 | 0.574 | 0.636 | −1.163* | 0.468 | 0.313 | 0.822 | 0.542 | 2.275 |
| Log of number of Seattle employees | 0.129 | 0.127 | 1.138 | −0.477 † | 0.252 | 0.620 | −0.108 | 0.203 | 0.898 | 0.102 | 0.197 | 1.108 |
| 500 or more national employees a | 0.058 | 0.503 | 1.059 | 0.230 | 1.004 | 1.259 | 0.679 | 0.840 | 1.972 | −0.799 | 0.842 | 0.450 |
| Constant | −2.929*** | 0.646 | 0.053 | −2.972** | 1.073 | 0.051 | −2.440** | 0.936 | 0.087 | −4.556*** | 1.077 | 0.010 |
| Pseudo R2b | .036 | .056 | .099 | .109 | ||||||||
| Total n | 439 | 439 | 439 | 353 c | ||||||||
Note. Logistic models estimated with penalized maximum likelihood estimation (Firth’s correction) to account for low-incidence events. OR = odds ratio.
The baseline survey was fielded in advance of the city’s writing of the Minimum Wage Ordinance regulations. Thus we ask about national employees, which was consistent with the understanding of the law in the months leading up to its enforcement. The current language of the law has been amended to include employees worldwide.
Tjur’s Pseudo R2.
Limited to firms who offered any benefit in 2015 (n = 353).
p < .10. *p < .05. **p < .01. ***p < .001.
Acknowledgements
The authors thank the staff and supervisors at the University of Washington Social Development Research Group for collecting the data, hundreds of unnamed Seattle area business owners and managers for sharing their experiences on our survey, and our colleagues on the Seattle Minimum Wage Study (Heather Hill, Ekaterina Jardim, Mark Long, Jennifer Otten, Robert Plotnick, Emma van Inwegen, and Jacob Vigdor) for their comments and suggestions. Mistakes, omissions, and interpretations are our own.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Partial support for this research came from the Laura and John Arnold Foundation; the City of Seattle; and the Eunice Kennedy Shriver National Institute of Child Health and Human Development research infrastructure grant, R24 HD042828, to the Center for Studies in Demography & Ecology at the University of Washington.
