Automation has permeated the workplace in advanced economies and this trend may accelerate in a post covid-19 world. Although there is burgeoning research on its electoral implications, we still know little about the social policy preferences of workers whose jobs are threatened by automation. Unfavourable public opinion on social policies in the form of political backlash may limit their implementation and effectiveness. Recent studies show that automation-threatened workers favour generous redistribution to compensate for potential economic loss. Yet, with the expansion of other types of social policies like social investment and demanding active labour market policies (ALMPS), or also known as workfare, these studies also prompt a further question: how do automation-threatened workers view these other social policies? In a recent study, I find that automation-threatened workers in Western Europe favour demanding ALMPs, which seems paradoxical given their own vulnerability to being subjected to the costs and burdens imposed by such polices.
Yet, this finding is less paradoxical when considering the employment trajectories of most workers in automation-threatened jobs. Automation-threatened jobs typically contain routine job tasks that are repetitive and easily codified. Although the employment shares for automation-threatened routine occupations have declined substantially in Western Europe, two recent studies demonstrate that most individual workers in such occupations remain employed in them. For instance, they find that more than half of these workers manage to cling onto their routine jobs; only a tiny fraction become unemployed in the UK, Germany and Switzerland. These studies attribute the decline in employment shares of automation-threatened jobs to falling rates of new workers entering these jobs rather than rising involuntary exits.
Presumably, two self-interest considerations may influence these workers’ opinions on demanding ALMPs. On one hand, they may seek to reduce barriers that impede their access to unemployment benefits or avoid additional costs that they may be burdened with, because they may come to rely on such benefits should they become unemployed. As demanding ALMPs impose such barriers and costs, they may oppose it. On the other hand, these workers may worry about the fiscal viability of their benefits, especially when permanent austerity renders welfare scarcity concerns even more acute. They may then support demanding ALMPs to impede current unemployed ‘competitors’ from accessing their unemployment benefits. It is likely that both considerations, welfare access and welfare competition, will weigh on unemployment-threatened routine workers’ minds. Ultimately, these workers’ support for such policies may depend on which consideration is more dominant on balance.
Using cross-national microlevel data from the European Social Survey (Round 8) for 15 West European countries, I find a significant association between occupational threat of automation (measured as the degree of task routineness) and support for demanding ALMPs which refer to cuts to unemployment benefits for refusing jobs paying lower wages or requiring lower education. Workers in automation-threatened occupations are significantly more likely to support demanding ALMPs than workers in automation-shielded occupations as illustrated in Figure 1. It also shows that workers whose occupations have middling occupation risks are ambivalent about workfare. These findings remain robust to other common alternative measures of automation threat found in the automation literature.
These results suggest that, on balance, concerns about welfare competition may influence automation-threatened workers’ views on demanding ALMPs more than considerations about their own barriers and costs to unemployment benefit recipiency. In this regard, the employment trajectories of workers in automation-threatened jobs help us to understand why the former consideration may be more salient than the latter. Even though workers in automation-threatened occupations face a higher risk of unemployment than workers in automation-shielded occupations, this higher unemployment risk is nevertheless unlikely to materialise as imminent unemployment. If it does occur, it is likely to take place in the long run as Kurer and Palier (2019) had argued; automation “gradually alters the employment structure [such that] jobs are eliminated over a long period of time”.
Consequently, routine workers may not view themselves to be under immediate threat of unemployment. They may therefore also not consider themselves to be in imminent need of unemployment benefits. Instead, the threat of future unemployment may render concerns about the future viability of benefit systems more salient, especially when individuals have become more anxious about the costs and sustainability of welfare programmes under permanent austerity. Routine workers may hence support demanding ALMPs to maintain their future benefit levels by imposing restrictive obligations on unemployed workers’ current usage of unemployment benefits.
These findings are consistent with others demonstrating that workers in automation-threatened jobs favour generous redistribution. They would like more redistribution and would thus support policies that appear to maintain their benefits in the long run. It is however important to highlight that demanding ALMPs, in practice, make no guarantee about maintaining benefit levels in the long run, even if it seems that curtailing benefit access for one group may imply ensuring it for another. Finally, it is foreseeable that there may be other mechanisms which spur automation-threatened workers’ support for demanding ALMPs. Future research may investigate them and evaluate their relative influence on support for demanding ALMPs. With competing mechanisms, an experimental setup may allow researchers to better tease out conditions under which some mechanisms are more dominant than others, and their corresponding effects on automation-threatened workers’ support for demanding ALMPs and other types of social policies. For instance, these workers may stop demanding ALMPs once they become unemployed and subjected to the costs and burdens of such policies. All in all, the findings here are a first step to understanding how the threat of automation may impact public support for various social policies which remains curiously understudied.
About the author
Zhen Im is a Postdoctoral Researcher at Copenhagen Business School.