Rethinking precariousness and its evolution: A four-country study of work in food retail

Abstract

This article analyses the expansion of precarious work in industrialized economies, integrating the welfare regime, risk shift and segmentation literatures. I examine trends in wages, working hours, pensions and healthcare in food retail in Canada, Germany, Sweden and the USA between 1980 and 2016. Precariousness increased in each country, but the form and degree of change differed markedly, reflecting the effects of product market competition, bargaining centralization and labour regulation.

Keywords

Atypical workers Canada comparative employment relations food retail Germany precarious work risk shift segmentation Sweden trade unions USA welfare

Introduction

Precariousness is perhaps the most negative change facing workers today. Some writers have explored its features explicitly (Kalleberg, 2009; Standing, 2011); others have done so implicitly by examining how welfare state reconfiguration, shifts away from risk pooling and labour market segmentation impose risks on certain types of workers.

‘Welfare regime’ research examines the generosity of public and private social arrangements and their impact on workers’ dependency on the market for survival (Esping-Andersen, 1990; Seeleib-Kaiser, 2008). ‘Risk shift’ research examines how collective actors transfer responsibilities for managing risk onto individuals (Beck, 2000; Hacker, 2006). ‘Segmentation’ research examines how labour market dynamics concentrate risks on specific categories of workers (Crouch, 2015; Palier and Thelen, 2010). While each literature focuses on a specific dynamic, there have been no analyses of how they fit together. Furthermore, we need to understand better how trends towards precariousness diverge across national contexts. I argue that each stream illuminates one dimension of precariousness: retrenchment (cuts in resource allocation), individualization (increased individual responsibility for risk-sharing) and segmentation (transfer of risks to non-core workers), but that these approaches should be combined. I aim to show that precariousness is a variegated concept with different dimensions; integrating these leads to a more nuanced understanding of its trajectories. I apply this composite framework to case-study evidence from food retail in Canada, Germany, Sweden and the USA. Sub-national jurisdictions are also compared to highlight within-case variations.

I first present an overview of the literatures on retrenchment, individualization and segmentation. Next, I apply an integrated framework to trends in precariousness facing supermarket workers in four countries and assess its contribution to our understanding of these trends. Finally, I review the findings and their implications.

Rethinking precariousness

Precariousness has been defined as an increase in unpredictable work arrangements, including wage-setting, scheduling and employee benefits that affect the material welfare of workers (Doellgast et al., 2018; Standing, 2011). It links closely to the concept of risk, also involving uncertainty (Kalleberg, 2009). The key emphasis here is on the disruptive effects of variable work outcomes. When work arrangements are stable, workers can plan for the future, make life investments (e.g. buying property) and avoid unpredictable debts. Those whose work arrangements are uncertain are forced to live in the short term and are vulnerable to debt.

While considerable research has found that work arrangements have become less predictable, there is little conceptualization of the different ways in which this has taken place. My premise is that we need to examine various aspects of precariousness in order to understand it.

Retrenchment

Retrenchment denotes cuts in resources that alleviate risks. Esping-Andersen (1990) examined the effects of different welfare regimes on the decommodification of workers and citizens. The more generous the welfare provisions, the greater the capacity for mitigating undesired risks and reducing dependency on the market for survival. This work created a point of departure for further research on the importance of both state-provided and occupational forms of welfare for shielding citizens and workers from uncertainty (Powell and Barrientos, 2004; Seeleib-Kaiser, 2008). Public retrenchment in OECD countries made it increasingly evident that private actors – firms and unions – would face new pressures in alleviating workers from insecurity (Johnston et al., 2011), and heightened precariousness when private actors are unable or unwilling to offset declining public contributions.

While the welfare state literature focuses on conventional forms of social protection, such as public and private benefits, retrenchment includes cuts in all resources guaranteed to individual workers by their employers and the state, such as salaries, working hours, retirement incomes and healthcare benefits. Linking this broader set of resources to risky work outcomes can be found in many literatures (Fudge, 2005).

Individualization

Individualization involved the reduction of risk pooling by collective actors: a ‘risk shift’ in which private companies and governments abdicate their responsibilities for managing risks (Beck, 2000; Hacker, 2006). When collective actors reduce guarantees to help workers cope with uncertain living expenses, healthcare costs and retirement, individualization increases.

The form of individualization depends on the context. At the level of the firm, it may involve replacing guaranteed wage increases with performance-based pay, more variable scheduling, increased use of temporary contracts or a shift from defined-benefit to defined contribution pensions. Collective bargaining may facilitate individualization by permitting working hour variability through the use of opening clauses (Palier and Thelen, 2010). The state has also contributed to individualization by weakening employment protection legislation, privatizing pensions and healthcare services or attacking collective regulation. The consequence is that worker debt is now the main economic stabilizer, creating a regime of ‘privatized Keynesianism’ (Crouch, 2009).

Segmentation

Segmentation by cleavages relating to both employment status (part-time, temporary, contract) and social location (gender, race, or age) is shaped by rules emanating from within and outside firms. Within-firm rules include company policies, norms and customs that regulate recruitment (limiting ‘ports of entry’) and promotions, and provide basic job security (Osterman, 1982; Osterman and Burton, 2005). Whether intended or not, these rules enforce the power of specific groups within the firm and consequently affect the distribution of precarious working conditions between core and non-core workers.

Rules external to the firm matter as well. These include collective bargaining and public policy. For example, union strategies in North America have favoured the segmentation of risks to younger workers through two-tier bargaining (MacNeil, 2013). The Hartz reforms in Germany exemplify how the deregulation of agency work and the reconfiguration of unemployment insurance have precipitated the growth of precarious work (Baccaro and Howell, 2017; Palier and Thelen, 2010). The combined effects of facilitating the use of more precarious work contract arrangements while diminishing social protection provided by the government contribute to creating new segments of insecure workers.

Towards an integrated framework

We must integrate these three dimensions for a more nuanced understanding of precariousness. First, this can show that there are multiple paths to precariousness. Second, this highlights that the trends affecting precariousness can be contradictory. While the focus thus far has been on the negative effects of retrenchment, individualization and segmentation, countervailing trends are also possible. Precariousness is mitigated when actors increase resources to buffer against insecurity, provide more guarantees from economic hardship and prevent segmentation from being a path to more precarious work. Finally, an integrated framework elucidates the importance of both private and public actors in the complex interactions that shape precariousness.

Research design

The research design is a multi-case comparison of food retailers in four countries: two coordinated economies (Germany and Sweden) and two liberal market economies (Canada and the USA). They exhibit different levels of bargaining coverage and centralization (see Table 1). This diversity makes for an interesting comparison of precariousness. All except Sweden are federal states, so I also examined differences within countries, comparing Québec and Ontario (Canada), Nordrhein-Westfalen (NRW) and Bremen (Germany) and New York (especially New York City, NYC) and Massachusetts (USA).

Table 1.

Bargaining coverage (percent) and centralization in food retail.

	United States		Canada		Germany	Sweden
	NYC	Mass.	Ontario	Québec	Germany	Sweden
Coverage
1980	80	70	80	80	100	95
2016	30	50	50	80	30	95
Centralization
1980	Industry	Chain-wide	Chain-wide	Chain-wide	Sectoral	Sectoral
2016	Chain-wide	Chain-wide	Chain-wide	Store-based	Sectoral	Sectoral

Source: Bargaining coverage is based on estimates provided by the major union in each country, centralization on interviews and an analysis of collective agreements. Only national estimates were available for Germany and Sweden.

I focus mainly on supermarkets, given their historical importance in the sector, but also address experiences in hypermarkets and discounters. The expansion of the latter store formats has been linked to heightened cost pressures on traditional supermarkets and new forms of segmentation in the industry (Carré and Tilly, 2017; Geppert et al., 2015).

Food retail represents the epitome of worker insecurity. The ILO (2015) reports that ‘no industry has experienced greater diversification in its waged employment over the last four decades than retail commerce, where different forms of atypical, non-standard employment practices have now attained standard status’ (p. 1). Job quality in food retail also varies considerably across and even within countries (Carré et al., 2010; Carré and Tilly, 2017; Coulter, 2014). While this industry is notorious for spreading poor working conditions across large segments of workers, it has also been subject to much experimentation across national and sub-national boundaries.

I focus on trends in wages, working hours, pensions and healthcare from 1980 to 2016, conducting 97 semi-structured interviews with managers, union representatives and experts with insider knowledge of major chains in each country. I include interviews with older and retired workers to capture a historical perspective. These data were complemented by an analysis of 20 collective agreements from the two time-periods, to explore changes in key provisions that affect insecurity. Official documents (corporate, union and government) and secondary literature were used to corroborate details.

Research findings

I now present an overview of the main trends towards precariousness facing supermarket workers in the four countries. In 1980, few workers were precarious in these countries: full-time schedules were the norm, wage rates were stable and increased with inflation, traditional benefits buffered workers from the costs of retirement and poor health and segmentation was not a severe problem. This changed in the ensuing years; but there were clear cross-country differences in trends along the three dimensions. As regards segmentation, there was also significant variation within countries.

USA

American food retail workers experienced negative trends across all dimensions. In the 1980s, supermarkets operated like cartels (Lichtenstein, 2009), able to control competition and guarantee gain-sharing with employees. The most powerful voices in bargaining units were those in more skilled occupations, such as bakers and butchers, who could expect full-time hours, real wage increases and generous health and pension benefits. However, less skilled workers such as check-out workers (known as cashiers in North America) also did reasonably well. In recent years, however, instability is omnipresent: fierce competition from non-union operators, including discounters and superstores (such as Wal-Mart), has undermined the pricing strategies of traditional supermarkets.

Unionized supermarkets have cut labour costs to compete with non-union competitors and new store formats. As a result, working-hours arrangements have become less generous across most employers. On average, 80 percent of unionized supermarket workers were part-time in 2016, while the reverse was true in the 1980s. This has combined with wage stagnation, which translates low hours into poverty incomes for most workers. Unionized chains are investing less in health and pensions, especially for part-timers and new hires. For example, one agreement in Massachusetts provides workers hired after 2013 with one-third of the health benefits received by those hired before 2010, and even this is not sufficient to meet basic health needs. ‘Our employees need to find insurance on the market’, notes one union representative, and these benefits are hardly sufficient to ‘cover the costs’. With weak union presence, many workers in other store formats, especially discounters, often have access to inferior benefit provision or no private scheme whatsoever. While some hypermarkets offer superior work arrangements, their focus on bulk purchases and low costs still adds considerable pressure on traditional formats to retrench. Combined with weak state welfare protections, these workers have altogether fewer resources to buffer against adversity than decades ago.

The individualized nature of risk-sharing arrangements facing American food retail workers is severe. Before the 1980s, unions in northeastern USA enforced pattern bargaining (or centralized bargaining in the case of NYC) to pool risks across employers; these could commit to real wage growth, strong scheduling protections and traditional employee benefits because their competitors were similarly committed. With declining bargaining coverage and decentralization, these employers now perceive these protections as a liability and require increased labour cost flexibility to respond to competition from discounters, and more recently ecommerce. Food retail workers are now scheduled only days in advance and have few controls over their schedules in the long term. Moreover, defined benefit (DB) pension arrangements have been supplanted with uncertain contributory plans. As one union representative argued, ‘many pension funds have not recovered from the recession/depression that we had in 2008–09. I know for a fact we lost over 25 percent of our assets, which was a tremendous hit [on] our pension fund’. With store closures, fewer companies participate in joint-employer benefit arrangements, which creates a disequilibrium between the ratio of retirees and incoming funding and threatens the long-term viability of these plans. Health savings account contributions or annual benefits continue to replace traditional insurance for part-timers, reflecting a national trend across industries (Hacker, 2006). However, the pace of individualization is not uniform across the country. For instance, unionized stores in New England have experienced a radical shift towards contributory healthcare schemes, whereas traditional healthcare arrangements are still offered in some NYC chains, albeit with rising co-payments and deductibles. Moreover, some countervailing trends are notable in unionized chains. For example, working-hour guarantees of 15–21 hours have become the norm for part-time workers in many locals, forcing employers to bear risks for the asymmetries between customer traffic and employee schedules, but they fail to compensate for the large-scale individualization of risks in most other areas.

Segmentation is also ensconced in American food retail, as risks have passed on to specific groups. As indicated above, part-timers are most affected by retrenchment and individualization in unionized settings. However, collective agreements in the industry typically include clauses that provide newer hires (usually younger) with riskier social arrangements such as lower hourly rates, reduced sick and holiday leave and inferior scheduling guarantees. One local union president describes most workers as ‘career part-timers’. As in most countries, there are few full-time positions available and specificities in hiring requirements mean that existing part-timers cannot expect to secure full-time employment in the foreseeable future. They must thereby choose between accepting hectic scheduling and lower compensation or leaving the company altogether.

The expansion of the non-union workforce represents another important form of segmentation. First, this is linked to the growth of non-union discounters whose focus on low labour costs and union avoidance has undermined the competitive strategies of traditional supermarkets, most of which emulate discounters by resisting unionization and eliminating services, for instance, through the outsourcing of more sophisticated baking and meat-cutting functions to third parties. Second, traditional labour contracts are being undermined by market competition. For example, most unionized supermarkets succumb to supplier pressure for more control over shelving products such as milk, bread and soft drinks; this is now done by external operators, many of whom are false self-employees.

Canada

Canadian food retailers experienced similar changes: the decline of the cartel-like model and the entry of non-union competition also exposed supermarket workers to negative pressures across each dimension. However, these trends are limited by the presence of resilient public programmes designed to protect workers from adversities in old age and illness. They are also limited by higher union densities in the industry, particularly since most discounters in Canadian food retail are bound by collective agreements.

Trends in retrenchment are nonetheless troubling. Heightened product market competition has spurred the growth of part-time work from 20 percent in 1980 to 80 percent in 2016, a trend exacerbated by plummeting real wages. Supermarket workers in Ontario and Québec in 2016 generally earned only 50–80 percent of their 1980 earnings. Since universal public pensions and most private pension plans in Canada are contributory, lower current incomes provide less generous incomes in retirement. Furthermore, the main private pension plan in the industry has shifted towards greater employee contributions, while employer contributions have not kept pace with inflation. One former executive describes this as ‘another thing that has hurt unions and has hurt companies. Some of these multi-employer pension plans are so under-funded, it’s shameful’. The generosity of health-care has been relatively stable, because of universal healthcare. However, rising co-payments and deductibles in private drug benefits have been harmful.

Conditions have also become more individualized in Canada. Pattern bargaining has eroded since the 1980s, leaving unions unable to pool risks by enforcing common standards across employers. This is most evident in scheduling practices. Most supermarket chains have adopted computer-operated, just-in-time scheduling, to reflect last-minute inventory fluctuations and customer traffic rather than employee needs and lifestyles. Senior workers have priority on the number of hours they receive, but they need to accept whatever is offered. According to one union official:

that’s our problem… with people that have kids, people that have parents that are sick, people that are studying at school. You’re telling them… look we’ve got a lot of hours to give you. If you don’t take them all, you’re out… Crazy schedules. That’s the worst part of our industry now.

Furthermore, with no minimum-hour guarantees, part-timers suffer from not knowing their income from week to week. Private pension schemes available to food retail workers tend to be contributory, making them particularly risky. However, federal Old Age Security providing workers with a guaranteed minimum retirement remains secure and has proven its effectiveness at reducing poverty by international standards (Fritzell and Ritakallio, 2010). Apart from pharmaceuticals, universal access to free healthcare has remained stable.

Furthermore, there are some countervailing trends. For example, some chains in Ontario have negotiated ‘minimum-wage plus escalator clauses’, compelling companies to augment the salaries of all workers proportionally to increases in the minimum wage, obliging employers to bear some of the risks associated with legislative changes to employment standards. Interviewees reported positive real wage increases for most in the industry as a result, as provincial minimum wage increases (most recently to CAD 14 an hour or €9.5) have had ripple effects for workers in food retail. Thus, while individualization has been the dominant trend, there are signs of certain forms of risk mitigation.

Trends in segmentation differ markedly by province. In Ontario, precariousness is strongly segmented. Part-timers receive lower wage and benefit arrangements, while younger workers are indirectly affected by tiered agreements that segment working conditions by date of hire. As one employer negotiator described, ‘over time, we had to reduce some labour rates and benefits. So you negotiated new hire rates, as people left the business you hired at a lesser rate, and you changed some of the benefits that they enjoyed’. Québec has barred such segmentation through provisions in the legislation on labour relations and standards, often called ‘orphan clauses’, which prohibit companies from discriminating against workers based on their date of hire. Non-union workers in both provinces do not benefit from various wage, benefit and working-hour arrangements provided by collective agreements. Yet surprisingly, most Canadian discounters in food retail are covered by collective agreements, limiting the extent that working conditions are segmented by store format. Small non-union segments within stores, however, are linked to the outsourcing of shelf-filling and other functions, leading to false self-employment in many cases. The gaps between union and non-union workers are narrowed, however, by universal Old Age Security and health-care programmes which provide protections to all workers in the country, eliminating gaps produced by cleavages based on store format and union-affiliation.

Germany

The major difference between the German experience and that of North America is how negative trends in retrenchment and individualization are segmented to workers not covered by collective agreements. In addition, the highly contributory nature of the welfare system multiplies the negative individualizing and segmentation effects of declining bargaining coverage.

Retrenchment is evident: while product market competition has been important, the most glaring factor is the rise of low-road competition ensuing from employers’ refusal to support the legal extension of sectoral agreements and associated declines in bargaining coverage from 2000 onward (Behrens, 2011). Stores bound by sectoral agreements in their region have struggled to compete against the growth of low-cost competition. The main negative trend is the retrenchment in working hours (from 20 percent part-time in 1980 to 67 percent in 2016). As a senior union official put it:

the problem in retail is that many employees work only part-time. Contrary to claims by the employers, this is often not the desire of the employees. The ever-expanding opening hours of the stores entail that employers are increasingly resorting to more flexible staffing and therefore part-time or mini-jobs, € 450 jobs.

The expansion of part-time work has corollaries for retirement incomes: contributory pension schemes provide low retirement security. One union official commented: ‘for a large part, employees in the retail sector clearly cannot live from their public legal pension, and by a long shot. Many employees are at risk of poverty in retirement’. Many low-wage workers in the country are thereby forced into social assistance.

However, collectively bargained real wage increases provided an important counter-trend. For instance, a low-skilled check-out operator in NRW earned 36 percent more per month in 2016 than in 1980, while the minimum salary of skilled butchers was 30 percent higher. There are some inequalities across regions. For example, the starting salary of an unskilled worker in Bremen is nearly 20 percent lower than in Länder where unions negotiated higher rates of pay. However, wage outcomes become more equal for those with job tenure. According to a senior union official, ‘there is a special feature in retail: even people without formal training are paid according to the collective agreement rate for trained employees after they have a few years of professional experience’. Thus, despite differences in starting salaries, all employees in the country reached the highest scale (€2471 a month) within 7 years, and skilled workers sooner. The relatively high rates of pay are amplified by supplements for ‘unsocial hours’ (evenings, nights and weekends). In NRW, for example, they vary from 20 to 120 percent of the standard rate. This creates a dynamic where some part-timers fare better than full-timers did 30 years ago, though those working few shifts struggle to maintain an income needed to survive, especially mini-jobbers whose incomes remain below €450 monthly for tax exemption purposes.

Individualization trends are mixed. Collective agreements and works council representation still provide important forms of collectivization. Workers under sectoral agreements are guaranteed real wage increases, whatever the profitability of their store or chain. They also limit firms’ ability to deviate from an employee’s average weekly hours. As one union representative put it:

if the contract says 20 hours, but regularly the company demands 28 hours, then you can’t just go back on it all of a sudden. Then you have a right to obtain at least 28 hours. Courts have frequently upheld this right.

Moreover, proactive works councillors often negotiate agreements providing 2–4 weeks’ notice of scheduling changes, allowing them to plan their expenses and family obligations accordingly. In contrast to just-in-time scheduling in North America, German retailers use software that adapts schedules to employee preferences (Carré and Tilly, 2017), undoubtedly because of the powers of works councils.

However, public pensions and healthcare are highly individualized. The public pension regime is contributory, and hence detrimental for low-wage workers who receive scant earnings in retirement. Controversially, public healthcare was traditionally funded through a parity arrangement in which employees and employers both contributed equally; but the employer contribution has been frozen at a rate of 7.3 percent since 2005. Employees are thus responsible for the other 7.3 percent, but also for any shortfalls in contributions needed to keep the system solvent, adding to workers’ financial uncertainty.

Segmentation trends are highly problematic. Service industries have become increasingly fragmented, contributing to the growth of workers in the periphery (Holtgrewe and Doellgast, 2012). Union officials estimate that only 30 percent of the sector was covered by collective agreements in 2015. This is, first, because of growing numbers of agency workers and subcontractors, both means of union avoidance, particularly since the Hartz reforms in 2002. As one union official commented:

Temp agency workers and subcontractors, who are often used for activities such as shelving, or even as strike-breakers, are significantly more precarious than permanent employees who are subject to social insurance. Temp agency workers and contractors also have fewer rights if a works council exists. A works council cannot represent any of the contract workers in the company at all.

Second, a growing number of stores are not covered by sectoral agreements. The decline of bargaining coverage is linked to the growth of franchising by leading supermarkets, accounting for 40 percent for the largest chain and 80 percent for the second largest. Unlike discounters who pay above the collectively agreed rate, these franchise operations pay substantially less. For example, one works councillor claimed that check-out staff in an independent franchise earned €6000 a month less than in corporate stores bound to a collective agreement. In addition, they do not benefit from wage supplements, scheduling guarantees and other protections associated with collective agreements or works councils.

The decline of bargaining coverage is also linked to store formats. While worker voice and associated protections in supermarkets are split between franchises and corporate stores, hypermarkets are usually bound by agreements and have high works council presence, because of their large size and the historically high numbers of union members within their stores. By contrast, the small size and anti-union strategies of discounters makes them hard to organize (Geppert et al., 2014, 2015). Consequently, they offer compensation exceeding collectively bargained pay scales as part of their union avoidance strategy, yet face no constraints in other areas of risk, including scheduling. This adds pressure on other stores to engage in union avoidance. According to one union official, ‘the discounters make the rules of the game, they make the prices, and others must follow’. In addition, they cut costs by offering fewer services (e.g. less in-store food preparation), spending less on store appearance and maintaining aggressive supply chain strategies.

Sweden

Swedish supermarket workers have fared comparatively better than in other countries. The remarkable stability of Swedish industrial relations and welfare state institutions is undoubtedly accountable for its many successes. Nonetheless, the rise of part-time work in an industry with no minimum working hours has been critical for many.

The major form of retrenchment is the reduction in working hours. While the entry of foreign discount chains has intensified product market competition since the early 2000s, the comprehensive sectoral agreements and state social policies have left employers with few cost-cutting alternatives to working-hours reduction. Part-time scheduling has prevailed for several decades, yet has nonetheless increased from 58 to 70 percent of the sector’s workforce, with a negative impact on incomes. Furthermore, since a significant pillar of the Swedish retirement system is its contributory pension scheme, any reduction in current incomes has effects that endure after retirement. However, the reduction of working hours and its consequences for current and retirement incomes is to some extent offset by trends in wage rates. The typical starting rate for an adult retail worker in Sweden is 52 percent greater than it was in 1980. Moreover, collective agreements oblige employers to compensate workers with a 50–100 percent supplement for work during unsocial hours. As an employers’ association official put it, ‘part-timers can make as much as full-timers, since the wage supplements [for] evening and weekends [are] high. However, full-time workers enjoy this too’. Thus, retrenchment in the form of reduced working hours has had negative effects for some workers, but this is often offset by positive increases to employees’ wage rates.

The highly collective nature of private and public risk-sharing arrangements makes Swedish retail workers less vulnerable than elsewhere. The continuity of encompassing sectoral bargaining has ensured that workers are guaranteed wage increases, independently of trends in product markets or store profitability. As a former company negotiator indicated:

we haven’t been able to raise the prices on food in any way. If you can buy food elsewhere that’s cheaper, you’ll go there. So as a consumer, you benefit. But every year you have a higher cost of salary of 2–3 percent. This creates a pressure on business to be very productive.

Collective agreements have also obliged companies to provide extensive scheduling protections, including 1 month’s notice of changes, limits on week-to-week working time reductions and rights to refuse work on weekends. Schedules thereby reflect workers’ needs for stability. The fact that all employers are bound to respect these conditions, independent of market dynamics, is akin to a form of risk pooling. Interestingly, employers interviewed were largely supportive of these encompassing agreements, as they shield them from market disruptions that would occur if new entrants could avoid collective bargaining.

The Swedish social-democratic welfare state also provides significant guarantees for workers. The pension system, while partly contributory, provides workers with a guaranteed minimum retirement income that is internationally noted for its success at reducing poverty (Fritzell and Ritakallio, 2010). Universal healthcare, while under pressure, is still largely government-funded and provides access to all who need it. The agglomeration of collectivized social arrangements in Sweden is significant, as it provides supermarket workers with buffers from a litany of risks. Thus, even those most negatively affected by decreased working hours enjoy important protections from adverse events pertaining to incomes, working hours and old age.

Segmentation has been limited, principally because of the comprehensive sectoral agreements which extend uniform terms to workers under different labour contracts and employers. According to a senior economist with the retail union; working conditions are almost the same in every chain and workplace, which prevents the growth of non-union competition that would drive a race to the bottom. In terms of employment contracts, part-timers receive the same benefits as full-timers. Many of them are young workers aged 16–19, many of whom are among the growing number of temporary workers in the industry and whose real wages have grown at a slower pace than average. For example, the starting rates for those aged 16 were 16 percent lower than for older workers in 1980, whereas the gap now stands at 57 percent. However, this minority of workers are guaranteed pay raises with age and benefit from all the privileges enjoyed by other workers.

Short-term contracts, renewable for terms of 6 months, are on the rise in the industry. The segmentation effect of these contracts is limited because they must conform to the terms of the collective agreement. In addition, they may only be renewed four times, after which such employees must be integrated into the regular workforce.

Moreover, the retail union has the power to extend the terms of the collective agreement to every major chain in the country. Swedish unions are famous for their power in introducing ‘blockades’, coordinating sympathy actions across multiple unions to shut down recalcitrant employers (Woolfson et al., 2010). For example, action against Toys‘R’Us in 1995 forced the company to sign an agreement with the union Handels. Such blockades prevent low-road competitors, such as foreign discounters, from making union avoidance a precondition of competitiveness, so union officials reported that German discount chain Lidl was ‘rather good’ at complying with the terms of the Swedish retail agreement when it entered the country in 2002. This was surprising, as the chain has been notorious for its union avoidance strategies in other European countries, including its home country Germany (Hamann, 2006). Finally, it is worth noting the importance of the Swedish welfare state in preventing the segmentation of working conditions relating to retirement incomes and healthcare. Being both universal and generous, Swedish social programmes set a floor for social welfare in the industry, preventing these aspects of risk from being segmented along occupational lines (Pavolini and Seeleib Kaiser, 2018).

Comparative analysis

A comparison of these four cases indicates that precariousness is rising everywhere. This finding resonates with the literature confirming that food retail is an exemplar of global trends in low-wage work (Carré et al., 2010; Standing, 2011). Retrenchment in working-hours was the single common denominator across the four countries. Contrary to trends towards voluntary part-time employment in some countries (O’Reilly and Fagan, 1998), part-time work has been largely employer-imposed.

However, precariousness varies across the national cases (see Table 2): there are more differences than similarities along the three dimensions. The USA has shown growing precariousness in all three. While there are notable differences across states and municipalities, it is nonetheless true that most American workers have been negatively affected by the retrenchment of resources that mitigate precariousness. In addition, firms provide fewer forms of collectivization to weather economic uncertainty than before. This has occurred along segmented lines, as collective agreements and human resource practices often bar part-timers and younger workers from protections available to older full-time workers. Furthermore, most food retail workers are now non-union, especially in discount chains, with protections often inferior to that provided by collective agreements.

Table 2.

Comparative analysis of main trends (1980–2016).

	USA	Canada	Germany	Sweden
Retrenchment	High	Medium	Low	Very low
Real wages	Small decrease	Large decrease	Moderate increase	Large increase
Working hours	Large shift to P-T	Large shift to P-T	Large shift to P-T	Large shift to P-T
Pensions	Large decrease	Moderate decrease	No change	No change
Individualization	High	Medium	Low	None
Segmentation	High	Medium	High	Very low

Canadian trends are similar, but the impact is significantly reduced by higher bargaining coverage and universal government-provided retirement and health benefits. Retrenchment has ensued as wages and working hours have been cut drastically across the industry. Individualization has also taken place, but its effects most poignantly affect scheduling, as employees with reduced private benefits still have access to public retirement and health services. Segmentation has also been less. In terms of bargaining coverage, most supermarkets and discounters are unionized in both provinces, reducing the gap in working conditions across workers in different store formats and according to union-affiliation. In terms of social policies, private benefits have weaker effects on the segmentation of workers, since universal programmes for retirement and health provide a minimum standard of benefits accessible to all workers. Segmentation is more prevalent in Ontario than Québec, given legislated differences in the acceptance of two-tier employment standards.

The German path to precariousness mainly involves segmentation. Full-time workers covered by collective agreements have experienced some improvements in income and no significant individualization. The growing number of part-time workers must grapple with fewer working hours but still receive important protections afforded by collective agreements and benefit from highly collectivized scheduling arrangements when they have a proactive works council. However, a growing number of workers are in chains operating outside traditional bargaining structures. Franchises and discount formats are inimical to unionization and the formation of works councils. Such workers have greatly inferior working conditions, and until 2015 had no statutory minimum wage. This is exacerbated by a welfare regime which favours individualization and segmentation, by relying on contributory schemes to fund public retirement and health benefits. While the introduction of the minimum wage in 2015 has diminished the wage gap somewhat (Bosch, 2018), the gap between workers covered and not covered by collective agreements is nonetheless far greater than in the other countries.

Sweden has mainly experienced negative trends in retrenchment, and even these are less severe than elsewhere. The major impact on precariousness is the decline in working hours. Beyond this, workers can still count on strong forms of security. Risks are still collectivized, as guarantees in wages, working hours and benefits by firms and the state remain largely unchanged. They are integrated because collective bargaining structures have remained stable and ensure that similar working conditions are applied throughout the industry.

Certain factors affect each dimension of precariousness. First, retrenchment is a consequence of product market competition. Driven by the growth of new store formats and the increase in non-union competition (except in Sweden), unionized stores have experimented with cost reductions in desperate attempts to remain profitable. Their methods have been constrained by the scope of collective bargaining. For instance, Swedish agreements contain strict limitations on wage-setting and scheduling for all employers, so much so that hiring fewer employees seems to be the most effective way to reduce costs. While similar constraints were once more common in North American agreements, collective bargaining has now become a tool for reducing labour costs, as unions face few alternatives but to align work practices in their firms with those of non-union operators. German firms have faced similar competitive pressures, seeking to diminish jobs and working hours where agreements exist, and pursuing cuts in pay and other areas where they do not.

Second, bargaining centralization affects the collectivization of risks. Centralized bargaining was critical to setting industry- or sector-wide guarantees for working conditions in Germany, NYC and Sweden. Pattern bargaining was instrumental in Massachusetts, Ontario and Québec, as it also served to maintain minimum standards. Aside from Sweden, different forms of individualization have taken shape from decentralization. In 2016, German stores bound to sectoral agreements faced drastically weaker forms of precariousness than those that were not. The enforcement of pattern bargaining in NYC also sustained certain forms of collectivization that do not exist in most US union locals, including traditional retirement and health benefit arrangements. Other union locals in the USA and Canada have since resorted to company- or store-level bargaining, preventing any significant collectivization through risk pooling with other employers. Negative effects were least pronounced in settings where welfare states provide universal retirement and health benefits.

Third, segmentation is linked to union capacities to extend collective agreements and regulations limiting the differentiation of working conditions across diverse types of workers. The near-absence of segmentation was most obvious in Sweden: Handels was able to extend collective agreements across all major chains within the sector, including foreign discounters with an anti-union reputation. Segmentation took many forms in the other countries. It was most obvious in Germany where newly subcontracted workers, franchise workers and workers in chains not bound to sectoral bargaining have radically fewer protections than those in stores bound to agreements. Segmentation is also problematic in North America, especially in the USA where lower union density and weaker welfare state provisions contribute to a larger gap between unionized and non-unionized workers.

Finally, the variations across sub-national units show how trends in precariousness are not always determined by national institutions. Bargaining decentralization poses challenges for union coordination in enforcing common wage floors across Länder, which explains why wage retrenchment is higher in Bremen compared to NRW. The continued enforcement of pattern bargaining in NYC has enabled the union to pool risks across employers in ways that are simply not feasible in Massachusetts. Provincial legislation explains why segmentation has been lower in Québec than in Ontario, yet why wages have become more generous in Ontario.

Conclusion

I have argued for a multi-dimensional framework to analyse precariousness. The literatures which focus on welfare regimes, risk shift and segmentation contribute to our understanding of how precariousness evolves over time; but their insights need to be integrated. My study of food retail in four countries shows how incorporating different dimensions of precariousness is important for exploring its trajectories. Doing so provides insights on national comparisons of precariousness. For example, social policy research has often focused on how the generosity of national welfare regimes affects insecurity (Esping-Andersen, 1990), neglecting how trends in individualization and segmentation fit into the picture. The framework presented in this study goes beyond trends in resource provision. In comparing Germany and Sweden, for instance, it shows how long-standing trends in individualization and the rise of segmentation were critical to heightened precariousness in Germany, contrary to Sweden where risks were more collectivized and integrated. Negative trends in retrenchment, individualization and segmentation were evident in both Canada and the USA but were far less severe in the former.

The framework also provides insights into shifts in precariousness within countries. The examination of sub-national variations illustrates the limits of national models when examining precariousness. For example, federal retirement incomes in Canada have remained generous, collectivized and integrated in that they are accessible to all, while trends in private retirement incomes for food retail workers vary considerably across provinces. Sub-national trends were evident in Germany and the USA as well, because of different political dynamics. The emphasis on segmentation also demonstrates how precariousness features differently across organizational lines: discount formats are most amenable to precariousness while hypermarkets are least. Of course, precariousness is ameliorated across the board when collective bargaining and welfare state institutions are strong and comprehensive.

Of great interest is the existence of dynamic and contradictory trends. As illustrated in my analysis, although precariousness has risen in each country, there were also counter-movements in most. The retrenchment of working hours in Sweden is counteracted by more generous wage rates. Wage stagnation in US unionized stores has coincided with improved scheduling protections for part-timers. Segmentation in Germany highlights how positive and negative trends may occur simultaneously, diverging across stores which are covered by collective bargaining and those which are not. Many forms of segmentation in Québec have diminished, even as retrenchment and individualization have been rampant throughout Canada.

My research opens new territory for exploring how actors affect each dimension of precariousness. It uncovers separate drivers of precariousness relating to each: retrenchment is most affected by product market competition, individualization by bargaining decentralization and segmentation by the power of unions to extend collective agreements and state labour regulations. Of course, these causes are interrelated and by no means exclusive of other factors. Nonetheless, my study has attempted to unpack the multi-causal nature of precariousness by identifying its varied sources and outcomes at different levels of analysis.

Future research could examine firms, unions and the state, their importance was identified in this study, but other actors may be relevant, including individual workers, managers and non-governmental organizations. Furthermore, the framework need not be bound to the parameters set by my study which used food retail as a testing ground. One could emphasize other units of analysis by comparing occupations, sectors or even national trends. Researchers can also address precariousness in specific settings, such as how it takes shape through organizational restructuring or with the rise of new global value chain configurations. New avenues for research should draw on many disciplines and fields, including sociology, industrial relations, management, political science, labour studies and economics. To conclude, my main argument is that precariousness is multifaceted in nature. Therefore, theory and research must be sufficiently rich and robust if we are to explore all its facets and formulate effective solutions for its mitigation.

Footnotes

Acknowledgements

Special thanks go to Gregor Murray, Guylaine Vallée, Mélanie Laroche and Gerhard Bosch for their comments on earlier drafts of this paper. I also thank Raoul Gebert and Mathieu Dupuis for language support. Finally, many thanks to the reviewers for their comments, and to Richard Hyman for his excellent editorial work.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research would not have been possible without research funding from the Interuniversity Research Centre on Globalization and Work (CRIMT) and Fonds de Recherche du Québec, Société et culture (FRQSC).

Author biography

Sean O’Brady is a Postdoctoral Researcher in the ILR School at Cornell University, USA.

References

Baccaro

Howell

(2017) Trajectories of Neoliberal Transformation: European Industrial Relations: Trajectories of Neoliberal Transformation. Cambridge: Cambridge University Press.

Beck

(2000) Brave New World of Work. Cambridge: Polity Press.

Behrens

(2011) Das Paradox der Arbeitgeberverbände: Von der Schwierigkeit, durchsetzungsstarke Unternehmensinteressen kollektiv zu vertreten. Berlin: Sigma.

Bosch

(2018) The making of the German minimum wage: A case study of institutional change. Industrial Relations Journal 49(1): 19–33.

Carré

Tilly

(2017) Where Bad Jobs Are Better: Retail Jobs across Countries and Companies. New York: Russell SAGE Foundation.

Carré

Tilly

Van Klaveren

Voss-Dahm

(2010) Retail jobs in comparative perspective. In: Gautié

Schmitt

(eds) Low-wage Work in the Wealthy World. New York: Russell SAGE Foundation, pp. 211–268.

Coulter

(2014) Revolutionizing Retail: Workers, Political Action, and Social Change. Basingstoke: Palgrave Macmillan.

Crouch

(2009) Privatised Keynesianism: An unacknowledged policy regime. British Journal of Politics and International Relations 11(3): 382–399.

Crouch

(2015) Labour market governance and the creation of outsiders. British Journal of Industrial Relations 53(1): 27–48.

10.

Doellgast

Lillie

Pulignano

(2018) From dualization to solidarity: Halting the cycle of precarity. In: Doellgast

Lillie

Pulignano

(eds) Reconstructing Solidarity: Labour Unions, Precarious Work, and the Politics of Institutional Change in Europe. Oxford: Oxford University Press, pp. 1–41.

11.

Esping-Andersen

(1990) The Three Worlds of Welfare Capitalism. Cambridge: Polity Press.

12.

Fritzell

Ritakallio

(2010) Societal shifts and changed patterns of poverty. International Journal of Social Welfare 19(S1): S25–S41.

13.

Fudge

(2005) After industrial citizenship: Market citizenship or citizenship at work? Relations Industrielles/Industrial Relations 60(4): 631–656.

14.

Geppert

Williams

Wortmann

(2015) Micro-political game playing in Lidl: A comparison of store-level employment relations. European Journal of Industrial Relations 21(3): 241–257.

15.

Geppert

Williams

Wortmann

Czarzasty

Kağnicioğlu

Köhler

H-D

Royle

Rückert

Uçkan

(2014) Industrial relations in European hypermarkets: Home and host country influences. European Journal of Industrial Relations 20(3): 255–271.

16.

Hacker

(2006) The Great Risk Shift: The New Economic Insecurity and the Decline of the American Dream. Oxford: Oxford University Press.

17.

Hamann

(2006) The European Black Book on Lidl. Berlin: Ver.di.

18.

Holtgrewe

Doellgast

(2012) A service union’s innovation dilemma: Limitations on creative action in German industrial relations. Work, Employment and Society 26(2): 314–330.

19.

International Labour Organization (ILO) (2015) Employment Relationships in Retail Commerce and Their Impact on Decent Work and Competitiveness. Geneva: ILO.

20.

Johnston

Kornelakis

d’Acri

(2011) Social partners and the welfare state: Recalibration, privatization or collectivization of social risks? European Journal of Industrial Relations 17(4): 349–364.

21.

Kalleberg

(2009) Precarious work, insecure workers: Employment relations in transition. American Sociological Review 74(1): 1–22.

22.

Lichtenstein

(2009) The Retail Revolution: How Wal-Mart Created a Brave New World of Business. New York: Metropolitan Books.

23.

MacNeil

(2013) Two-tier workplace compensation: Issues and remedies. Canadian Labour and Employment Law Journal 17(1): 181–224.

24.

O’Reilly

Fagan

(1998) Part Time Prospects: Part Time Work in Europe, North America and the Pacific Rim. London: Routledge.

25.

Osterman

(1982) Employment structures within firms. British Journal of Industrial Relations 20(3): 349–361.

26.

Osterman

Burton

(2005) Ports and ladders: The nature and relevance of internal labor markets in a changing world. In: Ackroyd

Batt

Thompson

Tolbert

(eds) Oxford Handbook of Work and Organization. Oxford: Oxford University Press, pp. 425–445.

27.

Palier

Thelen

(2010) Institutionalizing dualism: Complementarities and change in France and Germany. Politics & Society 38(1): 119–148.

28.

Pavolini

Seeleib Kaiser

(2018) Comparing occupational welfare in Europe: The case of occupational pensions. Social Policy & Administration 52(2): 477–490.

29.

Powell

Barrientos

(2004) Welfare regimes and the welfare mix. European Journal of Political Research 43(1): 83–105.

30.

Seeleib-Kaiser

(2008) Welfare state transformations in comparative perspective: Shifting boundaries of ‘public’ and ‘private’ social policy? In: Seeleib-Kaiser

(ed.) Welfare State Transformations. Basingstoke: Palgrave Macmillan, pp. 1–13.

31.

Standing

(2011) The Precariat: The New Dangerous Class. London: Bloomsbury.

32.

Woolfson

Thörnqvist

Sommers

(2010) The Swedish model and the future of labour standards after Laval. Industrial Relations Journal 41(4): 333–350.