1Mexican dictator Porfirio Díaz famously remarked that Mexico’s curse was to be “So far from God, and so close to the United States.” This saying is a good metaphor for jobs in Mexico’s retail sector, which, as in the United States, are among the worst jobs. Retail trade constitutes one of the largest employment sectors in high income countries, as well as middle-income ones like Mexico. Modern retail chains generate large quantities of entry-level jobs, many of which are low-wage. The sector’s employment patterns and practices make up an important strand among processes that generate and reproduce economy-wide inequality.
2Retail trade jobs are also emblematic of service work as a whole. Soft skills are needed to perform adequately even in entry-level positions but these soft skills, while “hired for,” are expected to be acquired outside the market. The sector (in the United States, Mexico, and elsewhere) hires a high proportion of women for customer contact positions, relying upon socialization processes and gender role expectations to access these skills. In so doing, it also does its part in maintaining gender-based hierarchies and occupational gender segregation. Also, long hours of operation are common, another feature of in-person service work. The ways in which modern retailers deploy labor in order to provide long opening hours have significant consequences for workers and for inequality. Weekend and night work are common. Increasingly the competitive model has had the twin, oftentimes conflicting, goals of cutting labor costs through thin staffing, short and variable schedules, and low wages, while also seeking to improve customer service and product quality and variety in order to retain or expand market share. For all of these reasons, retail work is an ideal window for understanding bad jobs—and what it is that leads to variation in the quality and characteristics of these jobs.
3The processes which shape retail job quality are a complex mix of corporate strategy—as it evolves in response to threat and opportunity—institutional frames for labor and product markets, aspects of the reproductive sphere of society, and worker responses. In this paper, we explore the role of work-related and non-work related societal institutions in directly and indirectly affecting retail job quality and its consequences for workers. As discussed immediately below, the role of societal institutions and norms is best studied with cross-national comparative studies. We compare the processes that shape retail jobs in Mexico and the United States relying upon sectoral field work in both countries conducted over the past nine years. Three dimensions of job quality are used as focal points of comparison: hours of work; compensation structure; and job mobility.
4The next section addresses the framework for thinking about the role of societal institutions in a particular sector, why Mexican and US retail job quality can be fruitfully compared, and the questions we seek to answer. The following sections describe data and methods and review key institutional differences between Mexico and the US. The main part of the paper reports findings on three dimensions of job quality. A final section draws conclusions.
- 1 The following draws upon Askenazy et al. 2012.
5We begin with an interest in job quality outcomes and seek to understand the role of societal institutions and norms in shaping business practices (affecting their decision terms) regarding the deployment and compensation of labor. Of course, jobs are affected by product market characteristics and corporate strategy—and this is the primary reason for looking within a particular sector, retail, and broadly similar firms within it–here chain stores (big-box and otherwise) in food and consumer electronics. Still, as is well documented, even in a globalized economy nation-level societal institutions and norms continue to impact job quality outcomes. The literature on varieties of capitalism has documented this variation at the macroeconomic level (Gallie 2007, Hall and Soskice 2001, Katz and Darbishire 2000, Whitley 1999). To fully understand the interrelationships of markets, technology, work organization, and societal institutions, a sector-based comparative approach yields the richest information (Bank Muñoz 2008, Coiling and Clark 2002, Royle 2006). Such an approach starts from observed differences in average job quality and analyzes the contribution of societal institutions to these differences using sector-specific cross-national comparisons (see Maurice, Sellier and Silvestre 1986 for a standard-setting study).1 Starting from a micro-level analysis focusing on selected job characteristics, authors move up to locate jobs within systems of employment and the latter within a constellation of interacting social and economic spheres. For example, Baret, Lehndorff and Sparks (2000) examine working time in retail in three European countries and Japan and Bosch and Lehndorff (2005) analyze multiple service industries (including retail) in ten European countries. Bosch and Lehndorff (2005) summarize four main spheres of influence on work (Gadrey 2000 offers a slightly different formulation with similar implications):
Product market and consumers, including parameters of competition and product market regulation
Management strategies, including corporate governance and technology
The labor market and associated institutions and norms, public policy, customers and industrial relations
The welfare state and gender relations including family structures and taxes and benefits
6Existing retail sector studies have primarily focused on comparisons across rich countries (Askenazy et al. 2012, Baret 2000, Carré, et al. 2010, Jany-Catrice and Lehndorff 2005, as well as those above) and there is much to be learned from comparing business practices across a broader range of countries. Middle income countries, such as Mexico, have their national modern retail chains and discounters as well as, increasingly, foreign-owned chains. Large, multinational chains—primarily those based in Europe, the United States, and Japan—act on business practices in middle and low income countries both through direct investment and through the diffusion of their competitive model. Modern retail chains in Mexico, as in the US, seek ways to cut labor costs, while seeking to maintain, or even improve, service levels and the quality and variety of goods that they provide. Hence, it is important to extend cross-national studies of modern retail to countries with more dissimilar institutions.
7We focus here on retail job quality in Mexican and US chains in the modern sector of retail in both countries. Such a sectoral comparison of retail job quality can include average measures but, in order to explain how institutions and market forces relate, detailed case-study work yields the intermediate processes that come into play in such interactions.
8The comparison of US and Mexican retail jobs juxtaposes two very distinct sets of societal institutions and social norms, on one hand, with modes of retail service that are quite similar in the two countries, on the other hand—chain supermarkets and convenience stores in food retail, and consumer electronics chains. The first set of relevant institutions for comparison is those with direct impact on the workplace; to wit, employment regulations and their enforcement, and the labor relations system. In Mexico, weak enforcement of a wide range of labor regulations (Bensusán 2009, Cook 2007) must be taken into account because it mitigates regulatory effects. Second, there are institutions with direct labor supply impacts as well as indirect impact on the workplace. Buttressing the latter are social norms related to gender roles, family and leisure time.
9Prior work has examined the activities of specific multinationals (particularly Wal-Mart) in Mexico (Chavez 2002, Schwentesius and Gómez 2002, Tilly 2005, 2006) and has summarized the characteristics of Mexican retail jobs (Álvarez and Tilly 2006, Tilly and Álvarez 2006). The present study takes a view that is broader than the case studies (examining the full range of chain stores in food and electronics), deeper than the job quality summaries (capturing a wider range of retailer practices and why they vary), and is comparative. This richer perspective enables us to bring Mexico into the discussion of the variegated ways in which national institutions and norms matter in job characteristics.
10In comparing the relative roles of institutions, norms, and markets, as well as resulting business practices, within retail, it is essential to take into account the context of national labor markets and economies that differ across the two countries. The majority of Mexican workers face a far thinner set of work alternatives and income supports than US workers (de la Garza and Salas 2006). While movement to informal retail is always a possibility, informal retail employment (as opposed to retail entrepreneurship) yields low incomes relative to formal work.
11We address the following questions:
To what extent can we explain national differences in retail jobs as resulting from institutional differences?
Which institutions seem to be most important in explaining which differences in job characteristics?
Within the parameters set by institutional and market differences, how much discretion is left for companies to adopt different policies and practices?
12Guided by these questions, we explore in depth three main dimensions of job quality: schedules; compensation structure; and mobility. They display similarities and differences which we summarize as follows. Regarding hours of work, the most striking contrast is the heavy use of part-time in the US and its very thin use by Mexican retailers. Regarding compensation, retailers in both countries experiment with varied forms of incentive pay, but display divergent practices regarding pay ladders, with Mexican retailers showing more pay inequality than US ones. Regarding mobility patterns, we find higher average turnover rates in Mexican than in US retailers. Moreover, more traditional gender roles in Mexico translate into women being more likely to drop out of the labor market to care for children for extended periods, and less likely to be able to advance into management. (Gender differences are highly relevant because retail employment is disproportionately female in both countries: a retail worker was 1.09 times as likely to be a woman as the average worker in the US in 2007, 1.20 times as likely in Mexico in 2009 [US Bureau of Labor Statistics 2012b, INEGI 2010, 2012]) However, retailers in both countries share a common tendency toward diminishing reliance on internal promotion. Within these average patterns, retailers do not behave uniformly and there is room for individual corporations to diverge from the main trend.
- 2 With varied degrees of comprehensiveness, US retail chains in the sample provided HR indicators su (...)
13In the United States, we conducted case studies at 16 grocery and consumer electronics chains (regional and national) during 2005–2007, interviewing 195 executives, managers, and workers and gathering quantitative human resources indicators.2 These quantitative indicators provide a context for information gathered in individual interviews and document broad patterns such as pay scales across job categories. In Mexico, we conducted 133 interviews at 20 chains and 12 small retail firms (in food, general merchandise, clothing, and consumer electronics) located across Mexico in 2003–2004; we also conducted a three-period longitudinal survey of 91 food retail workers in the state of Tlaxcala in 2006–2008. In both countries, about one-third of interviews fell into each occupational category (US: 30% executives, 30% managers, 40% workers; Mexico: 31% executives, 42% managers, 27% workers). Interviews were tape recorded, transcribed, and coded for qualitative analysis. In all three studies we asked about hours of work, compensation, and work history and job mobility. In the US and earlier Mexican study we also asked about company operations and the nature of tasks and skills involved in jobs. Mexican chains were reluctant to participate, so at 6 of the Mexican chains we only succeeded in getting access at the single store level. Some Mexican executives and managers also declined to give certain details, especially about salaries.
14The US chains follow formats that will be familiar to most readers: Large supermarkets, big boxes (food and broad-line, electronics-only, office equipment and supplies), smaller format electronics specialists, and two smaller format food chains (a compact warehouse store and a large convenience store selling a range of merchandise in addition to food).
- 3 The study also includes a chain of mid-sized clothing stores, family-owned food and clothing store (...)
15The two Mexican studies include a broader range of retail formats than the US study but, for this analysis, we limit our attention to the Mexican food and electronics chains most comparable to the US sample. The 2003-04 Mexican sample includes all the same formats as in the US (except the last two more idiosyncratic formats); indeed, there are four chains representing “autoservicios”, big-box hypermarkets or supercenters that have become the dominant retail format in Mexico. This earlier Mexican study includes five convenience store chains, two chains of intermediate-sized food stores, and five department store chains as well as a range of other formats (analysis of which falls outside the scope of this study).3 Distinctive to Mexico is the fact that one of the autoservicio chains is the government-owned ISSSTE chain, run by the public sector workers’ retirement fund (later privatized in 2007). The longitudinal study includes all the food-selling formats. Pseudonyms are used to refer to both US and Mexican companies, except for the ISSSTE stores (whose identity would have been difficult to obscure in any case).
16We also draw on publicly available data as appropriate to characterize workers and companies in the two countries.
17Clearly the two countries differ along numerous dimensions of political and social institutions. We highlight the institutional features that most directly impact retail work and that contrast across the two countries.
18The systems regulating work hours and minimum wage laws contribute, in each country, to shaping important aspects of work, particularly retail work because of its low wages and nonstandard hours. US retail environments are shaped by the lack of regulation of work hours, an hourly minimum wage, and a federal overtime mandated pay premium over 40 weekly hours. The minimum wage, to which retail entry-level wages are pegged—slightly above—is low and has declined in real terms (Economic Policy Institute 2010), standing at 70 percent of its 1968 level in real terms in 2011 (calculated by authors from US Bureau of Labor Statistics 2012a, US Department of Labor 2012). Key social protection benefits, such as health insurance and pension, are employer-sponsored and are often not available to part-time and casual workers. The possibility to establish sharp cost differentials between full-time and part-time workers directly affects retailers’ staffing patterns (e.g. most cashiers are part-time) and impacts workers’ experience.
19There is no universal, public and free, child care system in the US; instead, parents use a patchwork of private pay services, subsidized child care in centers or family providers, and informal arrangements (Clawson and Gerstel 2002; Gerstel et al 2002). Therefore a ready supply of people with caretaking responsibilities, primarily mothers, will seek and maintain part-time jobs in spite of the limitations which they entail simply to handle their care responsibilities.
- 4 At the time of revision of this article, November 14, 2012, the Mexican Congress had just passed a (...)
20Mexico has a longer work week, a daily minimum wage (at the time the research was conducted), and universal social insurance. Mexico’s full-time work week is 48 hours (six days x eight hours), with overtime provisions only applying beyond this point. The Mexican minimum wage was until late 20124 set by daily pay, and not hourly pay. Mexican law also mandates universal social insurance (via employer contributions), though the level is low and workers with greater market power generally get plans above the minimum. Formal sector workers can in principle benefit from subsidized housing and child care schemes though, in reality, ease of access and availability vary across regions.
21Collective bargaining can play a significant role in affecting job quality. Unions and collective bargaining operate in distinct institutional environments in the two countries, with consequences for job quality. In the US, with its context of the limited government role in employment regulation, collective bargaining can play a significant role in affecting benefit access, hours of work, and procedures for promotion. As a result, there are differences in job quality between unionized and non-unionized retail jobs, albeit differences that have diminished over the past 40 years. Still, collective bargaining affects relatively few retail workers.
22Where they exist, US grocery unions have bargained hard, showing a willingness to strike, notably in the hard-fought 2003-04 Southern California grocery strike and lockout (Broder 2003) and retailers have vigorously resisted unionization (Greenhouse 2011). But US union coverage has fallen significantly over time, so that retail union density was 4.9% in 2011, reflecting a grocery store unionization rate of 16.7 % (more than twice as high as the private sector average) along with virtually no union presence in the rest of the sector (UnionStats 2012). For these reasons, collective bargaining has a limited impact, not only affecting increasingly fewer workers directly but also having diminishing “threat” effect on non-union workplaces.
23In Mexico, where labor law is, in principle, much more conducive to unionization than US law, unions have a limited impact. By law, a petition of 20 workers is sufficient to form a union, whereas US law requires a majority vote that in practice is typically contested by employers. In reality, in the retail sector, as in many other sectors, collective bargaining contracts consist overwhelmingly of “protection contracts” that affirm the benefits required by law and specify wages and fringe benefits somewhat above the legally required minima, but block entry by any more militant union (Bouzas and Reyes 2007, Bouzas and Vega 1999, Tilly and Álvarez 2006). Indeed, Arturo Alcalde (1999, p.73), one of Mexico’s leading labor law experts, asked rhetorically, “Who knows of a single democratic union of restaurants, janitors, offices, or retail stores?” Author Tilly (forthcoming), based on an analysis of Mexican retail contracts, has argued that there is more heterogeneity than this statement would suggest, but that the large majority of contracts are essentially protection contracts. In our sample, the only case in which there appears (based on worker and manager accounts) to be a union committed to serious bargaining is the government-owned ISSSTE stores. In other chains, unions were absent or had a minimal footprint. Even where a union was present, workers and even store-level managers were often unaware of it. Particularly poignant were the comments of three department store clerks (interviewed jointly) who, unaware they had a union in their store, commented on what difference a union could make:
Interviewer: Does having a union make a difference?
Employees: They help the worker more. They respect holidays. Without a union, sometimes we have to work on holidays. With a union, you receive a dinner, a transportation bonus, and uniforms. You get all the fringe benefits.
24In Mexico, retail union density was reported at 1.4 percent in 2002 (the most recent year available; Esquinca and Melgoza 2006), but this percentage (based on a worker survey) is suspect, given the fact that many unionized retail workers are unaware they have a union.
25Now, we turn to findings from case studies including field interviews on several key dimensions of retail job quality which we found to present the most salient contrasts between the two countries, contrasts that illustrate most vividly the interplay between retailers’ competitive strategies on one hand, and national institutions and social norms governing work as well as the reproductive sphere, on the other hand. These dimensions are: hours of work, compensation, and career ladders and mobility.
- 5 For a more extended discussion of retail hours levels and trends in the United States, Mexico, and (...)
26Table 1 shows average weekly hours of work in the two countries, for the entire economy and for retail5. Two generalizations emerge. First, Mexicans work longer. Mexico unambiguously has longer weekly hours, for the economy as a whole and within retail. Second, the cross country difference is more extreme for retail. In the United States, retail workers work shorter hours than others. In Mexico, retail workers work even longer hours.
27The differences stem from distinct management practices, which in turn respond to distinct national institutions regulating employment. In the United States, retailers use the full-time/part-time distinction as a legally and normatively acceptable way to exclude a large group of employees from the standard wage level and full benefit package. The distinction has become a status distinction that may have little to do with actual hours worked. A human resources officer at a Value Fresh noted, “There’s probably plenty of 30-plus-hours part-time employees that just are not full-time because the stores are not able to make anybody wholesale full-time if they want to, because of the expense of the benefit packages.” And a part-time shift supervisor at Megamart remarked, “It’s funny because when I started this job, I knew it was part time and I figured part time would be, you know, maybe 25, 30 hours. Pretty much since day one, I’ve been averaging almost 40 hours.”
28Retailers achieve the control of total labor costs through the control of labor deployment—that is, through achieving a close coupling of worker hours with customer flow while maintaining minimum adequate staffing levels. In the United States, where standard full-time workweeks are 37 to 40 hours, control of total labor cost is achieved in two ways: 1) with the use of part-time workers, some of whom who can flex up to 40 hours (and have lower unit cost); and 2) in some cases, with relatively low guaranteed hours for full-timers (32 to 35 hours) who can also flex up. As a part-time cashier at Food Chief summed it up, “I wouldn’t mind 40 full time but they don’t do that. You can work 31 ½ Monday through Saturday and then Sunday you can work 7 ½, so they keep you under [40 hours].” Retailers can meet several cost-control goals this way: a close match of staffing with customer flow over time, avoidance of the federal and state overtime premiums (which kick in at 40 hours) when hours must be flexed upward, and the ability to reduce hours to achieve savings at short notice. Labor market institutions shape this set of strategies: in particular, the overtime premium and the legitimacy of unilateral, real-time management decisions to alter an individual’s work hours.
29As a result, hours in US retail tend to be short and variable. This sets up a tension between part-time employees who want more hours, and managers trying to keep a large crew of part-timers on board for maximum flexibility, while minimizing the total hours budget. A part-time cashier at Homestyle reported that the most common reason for co-workers quitting is “they don’t get enough hours.”
30Unlike the case in the United States, Mexico’s institutions motivate businesses in general, and retailers in particular, to set longer hours. First, Mexico imposes a wage premium on overtime, as in the United States, but the premium only takes effect at 48 hours. While some industries have adopted a shorter workweek, the 48-hour week consisting of six 8-hour days is the standard in formal Mexican retail. A standard 48-hour workweek raises the average number of hours and also makes it relatively easy to cover Saturdays and Sundays (simply by staggering six-day schedules) without use of part-time workers. Second, Mexico’s minimum wage set a daily minimum at the time of the research. For jobs that pay the minimum wage or close to it, as many retail jobs do, there was no advantage to having workers cover just a few peak hours per day, because the employer had to pay the daily minimum regardless of the number of hours. Indeed, contractual wage rates have in general been set by the day. Third, employer-provided health and pension benefits are relatively rare in Mexico; most Mexicans depend on the inadequate and uneven but nonetheless nearly universal government-provided systems (Dion 2010, Levy 2008). Avoiding benefit payments is not a major consideration for employers, which removes another motivation for creating part-time jobs in the United States.
31Thus, the key incentives for businesses, and retailers in particular, to shorten hours in the United States are absent in Mexico. Moreover, given that pay is set by the day, there is an economic incentive for employers to press workers to work longer while paying the contractual daily wage. Fieldwork indicates that, in fact, retailers do save on labor costs by extending worker hours: it appears to be common for stores to require workers to work added hours without pay, in violation of the overtime law. Interchanges with a manager at a Día y Noche convenience store and a clerk at a Hipermerca supermarket show two perspectives on this:
Interviewer: How do you handle overtime? Do you pay for the extra hours, or are they included in the regular salary?
Día y Noche manager: For a shift—that is, if someone works on their day off—we just pay them for an extra day.
Interviewer: But if somebody has to work, for example, one more hour at the end of a shift—does that happen sometimes?
Manager: No, no, because here every employee has their own work, so if they don’t finish it then they have to stay and finish it up.
Interviewer: And do you sometimes work some extra hours?
Hipermerca clerk: Oooh, yes, always. We never leave when we’re supposed to.
Interviewer: And do you get paid for the additional hours?
Clerk: (Laughs) No.
Clerk: It depends on whether…. Supposedly if they tell us they’re going to pay us overtime, then yes. But if we stay because we choose to, or, if we have a lot of work we have to stay because no way we can leave all that work undone.
32The operative conception is thus “a day’s work for a day’s pay,” even if the day’s work takes more than eight hours. A number of workers at various retailers confirmed this practice, with estimates ranging from “four extra hours in the average week” to “one or two extra hours [daily].” Certainly this strategy occurs in the United States as well (see, for example, Associated Press 2009), and Bernhardt et al. (2009) estimate that in their sample of low-wage workers in New York, Los Angeles, and Chicago, nearly two-thirds of grocery store workers who worked more than 40 hours were not paid the overtime premium. Still, our US fieldwork pointed to managers seeking to avoid scheduling workers for over 40 hours, rather than making it a routine practice; no workers in our US sample reported unpaid overtime.
33This line of argument posits that some legal requirements bind more than others in Mexico: employers are more likely to violate the legal requirement to pay for all hours worked than the requirement to pay the minimum daily wage. This supposition is certainly consistent with our retail fieldwork: there were numerous reports of unpaid extra hours, and no reports of contractual wages below the statutory minimum. It also makes sense, because there is a written record of contractual wages in employment contracts and payroll records, whereas there is no built-in record-keeping of extra work hours in retail establishments. Also contributing to the incentive to dodge overtime payments is a high premium of 100% (i.e., double time) up to 57 hours, and 200% beyond that point.
34An interesting indication that (on average) Mexican businesses press less-powerful employees to work longer hours, whereas US businesses impose shorter hours, emerges from comparing the hours of nonsupervisory and supervisory workers in Table 1. In the United States, weekly hours are longer in the category with all employees (which includes supervisors and managers), which resonates with our own fieldwork finding that store managers work extremely long hours (typically 55 hours or more per week), whereas, in Mexico, weekly hours are longer for line employees than for managers.
35There are two reasons to think that hours extension may be more extreme in Mexican retail than in the rest of the economy. First, Mexican chain stores are typically open seven days a week and long hours, much like their US counterparts. Stretching out hours is one way to cover customer flows over a service day of 12 or 14 hours. Second, as discussed above, retail unions are notorious for being particularly prone to toothless “protection contracts”, leaving retail workers particularly unprotected (Tilly forthcoming).
- 6 US figure is the median for all workers, 2011 annual average. Median for hourly employees alone is (...)
36Compensation is a key characteristic and important quality indicator for retail jobs. In the broadest terms, retail wages reflect the overall productivity and standard of living of the two countries. Thus, average hourly retail pay is $15.86 in the United States, and Mexican pesos 20.22 (US $1.50) in Mexico in 2009 (the most recent data available)6 (US Bureau of Labor Statistics 2012, INEGI 2010).
37Particularly important, both in pay-setting differences across countries and variation within countries, is corporate strategy. To illustrate this, we first examine how competitive strategy shapes variable pay in the US and Mexican retail sectors. Then we look at how market, institutional, and strategic factors come together in determining pay hierarchies within US and Mexican stores.
38Outside the terrain of electronics sales (which we consider separately below), variable pay took the form of bonuses. In both countries, store-performance-based bonuses for store managers (and in some cases multiple layers of store supervision) were ubiquitous. But bonus policies for non-managerial employees ranged all over the map.
39In the US grocery sector, bonuses were historically limited to the managerial level. However, one company has long provided bonuses based on store-level performance at the hourly level, and four others recently added them. A Homestyle district manager commented:
One of the biggest issues that the rank and file had was, “The department manager gets a bonus, co-managers get a bonus, store manager gets a bonus, all you folks at corporate get a bonus. We don’t get a bonus. So we feel like we’re working for the good of the department manager, not necessarily for the good of Homestyle.” We listened. Hey, that’s a valid point…. So we developed this profit sharing program and the impact has been incredible.
40At Food Chief, where no such bonus is in effect, managers still have the discretion to give Food Chief Bucks (store credits) to workers who have done something outstanding. Six of the ten food retailers offer some form of variable compensation to nonsupervisory workers.
41At the same time, two other chains recently dropped variable pay. Megamart ended its variable pay program in response to a wage and hour lawsuit. Freshland for several years paid productivity bonuses—tied to cases stocked—in certain production departments. But the experiment was dropped because of dissension from employees who did not have access to the incentives.
42The Mexican retail landscape in terms of variable pay to non-supervisors has one baseline element: Mexican law requires profit-sharing with all employees, which typically amounts to an extra two weeks of pay per year (but obviously depends on achieving an accounting profit). Beyond this, however, practices vary widely, including group bonuses tied to store or chain performance, individual bonuses for perfect attendance and good performance, and “employee of the month” awards. This scattered set of practices suggests an ongoing process of experimentation. However, a closer look at individual-level bonuses among grocery chains points to some likely processes of organizational learning and imitation. The chains offering performance bonuses (Bodegas López, Gusto, ISSSTE, and Sabor) are all based in centrally located Mexico City. The chains with employee of the month competitions (Convenimax, Día y Noche, and Tiendas Wilson) are all based in Monterrey, the largest city in Mexico’s North. Neighboring companies are presumably learning from each other.
43Commission pay in consumer electronics sales is a special case of variable pay, and presents a striking contrast between the United States and Mexico. In the United States, our sample includes four “pure” electronics retailers (as opposed to those selling electronics as part of a broader set of merchandise lines). Of these, the two big-box chains, Electronix and Technology Source, had both dropped commissions several years earlier. The main driver seems to have been cost reduction, especially in the context of shrinking margins on electronic equipment. As an executive at one of the big boxes described it:
It was great for the payroll side of the business and being able to reduce…. We [were] throwing a lot of money at a very few individuals in our stores. The other problem, too, was you couldn’t leverage your payroll very well. As your sales increased, so did your payroll. (TS12)
44He pointed to the incentive problem resulting from the abolition of commissions: “Now [the worker is] not motivated to want to work on a Saturday…where prior, you wanted to work on a Saturday and Sunday. That’s when you made your money, because the customers were there.”
45In Mexico, we were only able to gain access to one electronics chain, Hogartónico, but it is one of the largest, a successful and rapidly growing enterprise. There we found that 100% of salespeople’s income was based on “pay for contribution to sales.” When asked about the obligation this imposes on sales staff, Hogartrónico’s head of training replied, somewhat dismissively:
We provide the marketing, products, a nice assortment, we provide all the means they need in order to sell. What do they have to put in? Just love for the team, affection for customer service, in order to start selling the products.
46Thus, both US and Mexican electronics retailers structured compensation so as to keep costs low. But the leading US electronics sellers opted for cost-cutting over direct sales incentives (they used supervision, team-building, and promotion opportunities to encourage sales). In Mexico, on the other hand, a leading electronics retailer found a way to combine incentives with low cost by eliminating the base salary altogether.
47One simple measure in which there is a striking contrast between the two countries, is pay hierarchy within the store, measured by the store general manager’s salary divided by the entry-level clerk salary (including any bonuses in both cases), which we were able to calculate for most chains in both countries (10 in the US, 13 in Mexico; not all chains provided sufficient wage and salary data). Not surprisingly, this ratio is higher for larger stores with more layers of management: full-size US grocery stores (with 200-450 employees) have ratios ranging from 4.3-5.4, whereas the two small format chains (20-40 workers) in our US sample both have ratios of 2.8. But in Mexico, the two chains of full-size grocers for which we have information both fall at the lower end of the large size category (about 200 employees), but have much higher ratios, of 11.0 and 29.2! Confirming the contrast, smaller Mexican grocery chains (6-60 workers per store) show ratios of 3.3-3.8 (with the exception of Convenimax, to be discussed below), well above the US ratio of 2.8.
48The difference seems to reflect the overall higher level of inequality and hierarchy in Mexican society. In Tiendas Wilson, which has the staggering top-to-bottom ratio of 29.2, the largest portion of the gap occurs at the level of department heads (the lowest level of management), who earn 5.4 times as much as a full-time clerk—whereas in large US grocery stores this ratio ranges from 1.1-1.6.
49But variations within each country are as revealing as the gaping difference between them. For example, unions evidently make a difference. Comparing the two large-format Mexican grocers, the lower management to clerk ratio of 11.0 is at ISSSTE, where there is a formidable union. At Tiendas Wilson, where HR officials commented, “We have a union-free life…even though there’s a union” because the union does not get involved in any management decisions, the ratio is nearly three times as large—and the pay ratio of department head to clerk in particular is exactly three times as large as at ISSSTE. Similarly in the US, comparing non-union Homestyle and unionized Food Chief, two otherwise similar chains, we find that Homestyle has a larger ratio of 5.4 compared to 4.9, but even more revealingly, Food Chief has a much larger gap between full-time and part-time clerks (full-timers are the core of union membership), but a much smaller gap between full-time clerks and managers.
50Corporate strategy, and in particular the effect of ownership structure, also enters into the determination of these hierarchies. Convenimax is the one smaller-format Mexican grocery chain where the manager-to-clerk ratio falls outside the neighborhood of 3.5: its ratio is 2.0 in franchised stores and 2.5 in company-operated stores. Clearly Convenimax management is marching to a different drummer; one possible reason is that the chain is US-owned. Also, we have wage information on one Mexican chain with mid-sized stores (about 100 employees), Gusto, which has a ratio of 2.6, lower than all the Mexican companies except Convenimax. Possible reasons for this outlier include the fact that Gusto is a small family-owned firm in the process of implementing more formal management systems, and that it is based in a sparsely populated state with little presence of the large chains. Thus, country-level differences, the degree to which institutions shape salaries, and company-level discretion all affect within-store pay inequalities.
51Retail chains in both countries follow similar strategies regarding promotion ladders and tradeoffs between promoting from within and hiring from outside. However, significant differences in norms regarding gender roles and reproductive sphere institutions result in somewhat different trajectories for Mexican and US women retail workers, as well as slightly different patterns of gender differences.
- 7 Source for education: INEGI, Censo de Población y Vivienda 2010: Tabulados del Cuestionario Básico (...)
52Before delving into similarities and differences in mobility trajectories for retail workers, it is worth noting two contrasts that are important parts of the context for mobility in the two countries. First, the general level of education is far lower in Mexico than in the US and workers with any post-secondary education are considered part of the socio-economic elite. Over 51% of the working age population (age 15+) does not have a high school diploma and only 17% has a college degree7.
53Also, formal sector jobs are far less available, even for workers with education. Retail jobs are considered appropriate for educated workers relative to other options. This pattern holds even given the caveats that jobs in retail are considered somewhat low pay and that, ultimately, most workers with education aspire to run their own business, not remain employed in a chain.
54Second, and of less importance in this cross country comparison, the lowest position in chains— bagger—is not a waged position In Mexico. Baggers —usually school age youths—work exclusively for tips. In the US, the job is often held by youths as well but is waged and is a common first step into retail job ladders. Similarly, uniformed parking lot assistants (directing motorists), a non-existent position in the US, work exclusively for tips; they are adult men. Otherwise job ladders within chains are fairly similar in the two countries.
55Turnover for entry-level workers is high in both countries, a hallmark of retail work. It is primarily due to quits, rather than firings, also a common pattern. Among entry-level workers, turnover is high among young workers overall, in both countries. It is also triggered by workers seeking higher pay (in Mexico) or longer work hours (US part-time workers). In Mexico, women quit due to family responsibilities (see below). Among Mexican higher level workers, turnover is due, in addition to seeking higher pay, to setting out to run a business on their own (sometimes managing a franchise store), and to moves and migration, while among their US counterparts, it is more likely triggered by the desire for easier work schedules (less weekend and night work).
56Where US managers occasionally mentioned advantages of turnover such as keeping a lid on wage growth through seniority or implementing reorganizations of work processes without friction or inertia, we saw only rare instances of this view in Mexico manager interviews. Mexican managers saw turnover rates above 25% as a liability and as eroding the ability of the business to operate efficiently. Manager statements illustrate the contrast. An assistant manager at The Market, a US grocery chain, stated, “High turnover can sort of give managers a chance to somewhat start anew.” A Food Chief regional HR manager put it more bluntly: “We like turnover because it keeps our average hourly rate down.”
57In contrast, a Mexican manager commented on the disadvantages of turnover and his company’s orientation to keeping workers, uttering words we heard in other interviews: "Of course turnover is a problem. There's healthy turnover and there's turnover that gives you trouble. It means doing it all over again--training and hiring--and all that costs money. Our main challenge is to retain our people. Here they're valued, so we have to take care of them."
58In spite of this apparent difference, turnover is on the whole higher in Mexican food chains in particular than in US ones. (There is only one electronics chain in the Mexican sample making comparison of turnover in this sub-sector impossible.) Perhaps it is this high turnover that generates managers’ greater focus on retention. In food chains for which information was available average turnover was 100% (as high as 113% if government owned ISSSTE is excluded). This level of turnover is substantially higher than the average in the US study sample (mean 53.4%).
- 8 The one exception was a food retailer with a low turnover of 20% that retains workers by exceeding (...)
59Beyond averages, there is cross firm variation in both countries. Entry-level turnover ranged widely across Mexican firms, from 1% yearly (in ISSSTE) to 230% in chains, and from zero to 80% in single store retailers. The availability of other (formal) employment options affected turnover rates directly, according to Mexican managers. For example, a Hipermerca manager noted: “Turnover has stopped [because] the shoe industry is in recession; those are businesses that offer much more manageable hours.” Respondents at Gusto noted that turnover had been halved from a high of 60-80% after maquiladoras relocated away. Cross-firm variation is evident in the US as well but the range of variation in the study sample was far narrower, 40-80% in food retail8 and 68-80% in consumer electronics.
60In both countries, the broad pattern of promotion has historically been to promote from within. Interviewed US companies reported fairly high rates of manager promotion from within —in parallel to the high turnover model for entry-level workers. Food chain respondents estimated that 60-90% of upper level hourly jobs as well as store level management were filled from within while electronics chains reported rates of 50%.
61While Mexican chains also promote from within, Mexican workers with formal education take jobs in retail more readily than in the US. They are considered for promotion ahead of those who started in entry-level positions and who have a low education level (junior high).
62At Convenimax, a US owned convenience store chain, 90% of managers started in lower positions and were promoted from within; so were 30-40% of supervisors and 60% of district managers. Promotion from within is seen as a means to ensure that there is congruence of practice and adherence to company “spirit” across franchises. If the job of the manager is to open new franchises, it is desirable to have someone promoted from within, who is dedicated to the company. At a supermarket chain, a frontline worker estimated that 70-80% assistant managers were promoted from within.
63This average reliance on promotion from within also encompasses significant variation between and within firms. For example, estimates from various interviewees ranged from one third of store managers to none promoted from within at Día y Noche but were as high as 66% in one region at Hipermerca. Also, mid-level managers (department managers and supervisors) are far more likely to be promoted from within than general managers, a pattern shared with US retailers.
64Even while mobility from within still distinguishes retail from other industries, recent changes in practices include increasing hiring from outside. Modern retailers in both countries have tapped increased availability of college-educated workers to hire for mid-level and higher level managers. In the United States, educational requirements for retail positions are higher than in Mexico (for example, a 4-year college degree for assistant store manager as compared to a two-year degree in Mexico), but the patterns are similar. At Convenimax, the Manager of Franchising noted that competitor 7-11 started raising the education level of some managers, and expected Covenimax to follow suit. For example, several years earlier the chain had started hiring regional managers with technical degrees. At ISSSTE, the government owned food chain, very few managers come from within because internal candidates “do not have the education level required.”
65While retail job ladders and hiring patterns look roughly similar across the two countries, gender differentiation of mobility patterns is more marked in Mexico than in the US.
66In both countries, broad patterns of gender typing operate. On the whole, men are more likely to be over-represented in management ranks (and rise faster) and, within these, in the higher echelons. Mexican men were able to parlay promotions achieved in small business to similar positions in retail chains. Taking these similar patterns into account, we find that Mexican women workers have work patterns more marked by their reproductive roles than US women. In the US, the gender differences appear to be ones of kind of job and degree of career achievement. In Mexico, in addition to a similar job gap, women are much more likely to withdraw from formal work during family formation and child bearing years.
67This difference reflects both Mexico’s gender hierarchy and its reproductive institutions. Mexico has much more clearly demarcated gender roles, with women expected to have primary responsibility as family caregiver, particularly in the raising of children. Related to these gender roles and further reinforcing them, child care availability and affordability are limited, making it difficult to hold a job while raising young children. For example, by law, formal workers can access publicly funded child care but, in reality, child care availability is limited and unevenly accessible across the country. Social forces steering women toward a full-time parenting role are further reinforced by the prevalence of full-time schedules in formal sector retailers. Where US women workers might adapt to limited child care by working part-time—while suffering significant immediate economic costs and longer term negative career consequences—Mexican women leave the workforce altogether because short hours are not an option in formal retail, except in a few multinational retailers who can afford to pay an effective hourly rate well above the minimum wage.
68As a result, we find two different trajectories for Mexican women. First, women with higher education are hired at mid-level or higher level manager positions which they seem to maintain with some interruption for child bearing but with options for child care later. Second, women hired in entry-level positions face two levels of challenges to mobility: they experience career interruptions from withdrawal from the labor force; and those with continuous work histories are hindered by education requirements for mid-level positions.
69The career interruptions are quite significant. Those who worked while raising children reported “family difficulties” as the reason for leaving positions over time. Similarly, managers alluded to family difficulties that triggered turnover. Commenting on sources of turnover, the HR manager at one Tiendas Wilson store said “recently we’ve had a series with family problems “. The HR head at another Wilson store listed “married women who don’t have anyone to take care of their children and family problems” as causes for turnover.Longitudinal study interviews –for women who are older and in lower level positions—likewise indicate stretches of time entirely dedicated to child rearing or, at least, to avoiding formal sector work.
70The US study did not gather detailed career histories but did gather some work histories as well as interview part-time workers (mostly cashiers). For part-time workers and for those who recount shifting back to part-time from a full-time schedule, family responsibilities figured significantly. Workers reported working part-time due to the care of young children and the need to reduce hours during school holidays. Given the limited availability of affordable child care, two parents may work alternating shifts in order to avoid having to use non-family child care at all and single parents find that the only way to watch over school children is to work part-time. (Elder care responsibilities also played a role in part-time work.)
71In the United States, day-time, steady, part-time schedules were filled by women with family responsibilities. These women did not turn over much. The accommodation of family that the job provides came with a cost: low pay, and small chances of moving up due to the accumulated years in entry-level positions. These women, however, could shuttle back and forth between short-hour part-time and time-adjuster status (seeking and waiting for full-time) when family responsibilities lightened.
72In addition, promotion paths were open for US women in entry-level positions. For example, cashiers with a steady work record could be promoted to shift supervisor, and eventually “front end” supervisor (who covers cashiers, customer service desk, and a few other functions). This promotion path was peopled primarily by women. It is much less certain that front end supervisors could migrate to supervisory positions in other sectors of stores. However, in chains with smaller stores, those hired as cashier (a multi task job that also entails being a floor clerk) could eventually manage a store. Beyond store manager, however, women were thinly represented. In sampled grocery chains, we did not encounter a female regional manager. US women workers beyond entry level and with higher education were found in HR, training, merchandising, and a few in operations functions.
73We inquired about the role of institutional differences, if any, in job outcomes in Mexican and US retail. Institutional differences indeed come into play. Exactly how and to what degree they do so, however, varies depending upon the particular job characteristics of concern. In some dimensions of job quality, within-country cross-firm differences are more salient for job outcomes than cross-national contrasts. That is, there is room for firm discretion within given national contexts. Furthermore, we note that where they have impacts on job outcomes, institutional effects may be direct (as with a minimum wage law) or indirect (as with an institution affecting labor supply characteristics). We elaborate these patterns below, focusing on the main dimensions of job quality.
74Regarding work hours, differences in labor market institutions have the most direct effects even while modern retailers in both countries operate with long hours and 7-day operation. The most salient contrast is that Mexico has higher average hours and low use of part-for most entry-level jobs while US retail has a high rate of part-time as well as lower average hours. We attribute these patterns to key features of employment regulation and social protection mechanisms in each country. Most important are the fact that in Mexico, the minimum wage and contractual wages were (until late 2012) based on the work day (rather than hourly), on one hand, and the employer based benefit provision system in the US which has built in incentives to use part-time, on the other hand.
75In the two countries, incentives run in opposite directions regarding work hours. In Mexico, the weekly overtime threshold of 48 hours, combined (until late 2012 when it was revoked) with the day-based minimum wage, discourage creation of part-time jobs, particularly for jobs close to the minimum wage level. The day-based wage also had a built-in incentive to extend unpaid hours, a practice that is feasible in an environment of limited enforcement and weak unions. Both of these processes led to high average hours. In the US, in contrast, the overtime threshold of 40 hours is enforced on the whole, and, combined with the benefit cost differential between full-time and part-time workers, yields incentives to use part-time as well as to keep hours at or below 40 hours for both full-time and part-time workers.
76The major institutional contrast between the countries that affects pay hierarchy is greater societal acceptance of hierarchy in Mexico than the US. We found greater gaps between managers’ and wage workers’ pay in Mexican retailers than in the US; interestingly, in some companies the gap was greatest between department heads (first level of management) and entry-level workers suggesting high tolerance for pay differentiation indeed. Similarly, where commission pay was used (in electronics), Mexican retailer Hogartrónico used 100 % commission pay thus exposing sales staff to considerable fluctuation in earnings. In contrast, US electronics retailers, having lost control of the costs of partial commission pay, had reverted to 100% base salaries for sales staff thus reducing labor costs but also eliminating large differences among workers.
77But company discretion also plays a role in pay hierarchy. There were differences due to size, with large stores having greater hierarchy in both countries. There were also differences due to ownership type. Family-owned chain Gusto and government-owned ISSSTE were able to implement more equal within-store wage distributions than their counterparts, presumably in part because they are sheltered from some of the financial market pressures to which other chains must respond. US-owned Convenimax also shows a different within-store wage distribution than its Mexican-owned convenience store counterparts. This US-owned chain was presumably less susceptible to societal norms regarding hierarchy. There also were indications that unionized retailers (Mexico’s ISSSTE and US Food Chief) had more compressed pay structures.
78Regarding variable pay practices, we find that there are greater within-country, cross-company, contrasts than cross-national differences illustrating a fair amount of firm discretion in this matter. While store performance based managerial bonuses are common practice in both countries, retailers differ significantly within-country as to their bonus practices for non-managerial employees. Retailers in both countries experimented with different types of pay, mixing forms of commission or bonus with base pay. There was variation over time for some retailers, indicating a fair amount of experimentation, thus making it difficult to discern reliable patterns. Also, within country variation might reflect joint experimentation. For example in Mexico, there were shared practices among firms in close proximity (and a concomitant contrast with practices of retailers across regions) pointing to the geographic diffusion of learning.
79Modern retailers structure job ladders similarly in the two countries; we find little difference and none that could specifically be attributed to national institutional settings. Of course, education requirements for particular job grades are higher on average in the US, reflecting the greater prevalence of secondary and post-secondary education in the population relative to the situation in Mexico. Also, retailers in both countries have relied on recruitment from within to populate the ranks of supervisors, store management and even regional management. In recent years, retailers in both countries have displayed a tendency toward more “lateral” recruiting (recruiting from outside) than in the past.
80Turnover is high in retail in both countries but we do find differences in the reasons for turnover that are attributable to national differences. Turnover levels appear to vary within country and across retailers and are driven by practices at the firm level.
81Even with similar job ladders, significant cross-national differences in gender roles in the family and in interactions with the education and labor market spheres indeed are reflected in mobility and turnover patterns for men and women in each of the two countries. Mexican women are more circumscribed in their role than US women workers. So even while there is gender-based stratification in US retailers, it appears less pronounced than in Mexico. Thus, societal institutions beyond those regulating the labor market operate in concert to have an indirect effect on jobs and contribute to shaping distinct outcomes in the two countries.
82Higher levels of education, particularly post-secondary education, are less accessible to women in Mexico; they are expected to become the primary family caretaker. We noted a two-tier pattern to women’s paths in Mexican retail. On one hand, women with professional and academic post-secondary degrees access retail jobs “from the outside,” being hired in at higher levels. Women in these positions are either young and without children or purchase caregiving in order to hold full-time jobs (thus adhering to a different norm). On the other hand, women hired in entry-level positions experience limited promotion opportunities due to low levels of educational attainment as well as truncated careers due to multiple interruptions caused by responsibility for family care. Family caregiving—and difficulty with accessing child care—were often noted as a reason for quits (causing turnover). The fact that Mexican retailers create few part-time jobs contributes to this pattern. Full-time work is difficult to handle for entry-level workers with family care responsibilities, particularly when transportation is problematic. In these ways, socially constructed gender roles operate early in life and throughout the work career, thus shaping women’s trajectories in retail.
83Without being free of the pressures of gender-based job stratification, US women retail workers, particularly entry-level workers, resolve the immediate difficulties of care giving caused by lack of affordable child care and sparse paid family leave by relying on part-time work, thus remaining connected to jobs (albeit with low earnings and scant benefits) over time. Also, there is more widespread acceptance of women’s work outside the home and of the use of non-family child care, even while the child care system is patchy. Without arguing that non-benefitted part-time retail jobs are in some sense “good” for US women, we do find that gender role norms are not quite as constraining to job mobility for US women in retail. Entry-level women workers are better able to move up to supervisory positions and store management.
84In brief, nation-specific institutional effects such as we have described typically interact with corporate strategy. Companies react resourcefully to each set of institutional constraints, and those reactions themselves compound the initial differences. So, for example, in terms of hours of work and electronics commission sales, US retailers lean toward cost-minimizing solutions (while of course seeking to ensure adequate effort), whereas Mexican retailers lean toward worker-effort-maximizing solutions (while of course seeking to contain costs).
85Beyond the national patterns of institutional frameworks and typical corporate responses, we also found variation within each country. Some of that variation has to do with regional differences in labor demand and regional level institutions, dimensions we did not explore in this paper. But there is also space for businesses to come up with varied solutions to the same basic problem as we have just demonstrated.
86We posed three questions about retail jobs in the United States and Mexico: To what extent can we explain broad-cross national differences in retail jobs with institutional differences? Which institutions seem to be most important in explaining which differences? And within the parameters set by institutional and market differences, how much discretion is left for companies to adopt different policies and practices?
87We sought to limit extraneous variation by restricting our attention to similar businesses: large chains, primarily in groceries and consumer electronics. Nevertheless, even within this restricted universe, we did find that national institutions —both legal and social norms— have discernible imprints on the characteristics of retail jobs. The simplest instances are cases where institutions act directly on employment as is the case with hours of work outcomes. But indirect institutional effects are also important, particularly involving institutions that surround the reproduction of labor such as child care systems and norms regarding mothers’ role in child rearing. At the same time, institutions are far from an iron cage: we find ample room for managerial discretion and experimentation in both countries’ retail sectors.
88These findings have implications for broader analysis and theorization of unequal employment outcomes, and for public policy. On the analytical side, these results bolster the argument that national institutions remain highly important in shaping workplaces even in conditions of globalization of production models and ownership. They highlight the less visible role of reproductive institutions, especially in a sector like retail with a disproportionate concentration of women as is true of many service sectors. And they flag the impact of strategic behavior by firms in determining final employment outcomes. These analytical conclusions also translate into insights on the policy side. Understanding the continuing weight of institutions remind us that public policy gives us important levers to affect job quality—including a broader menu of possibilities beyond what’s in place in any given country. At the same time, our findings make the cautionary point that institutions can have major unintended and unexpected consequences (widespread part-time work in the US and unpaid work hours in Mexico were certainly not envisioned by those designing the policies that motivate them). This comparative glimpse at one particular type of job in two particular countries suggests the rich possibilities for further comparative work on the determinants and processes underlying labor market inequality and job quality.
We thank the Russell Sage, Rockefeller, and Ford Foundations and the Upjohn Institute for Employment Research for support in various portions of this research. Thanks also to Patrick Adler and, especially, Diana Denham for excellent research assistance.