Welfare state support during the COVID-19 pandemic: Change and continuity in public attitudes towards social policies in Germany.

Our analysis asks whether the pandemic situation affects welfare state support in Germany. The pandemic has increased the health and income risks calling for welfare state intervention. While increased needs, more deservingness, and higher state responsibility during such a crisis would suggest augmented support generally and among those at risk, this might be a short-term effect and cost considerations could reverse this trend. We study public attitudes towards four key social policy areas based on the German Internet Panel (GIP). We use three waves prior and further three waves since the pandemic had been declared in March 2020. The analysis shows both continuity in the popularity of social policies, in particular health and pensions, and some short-term increase in support for unemployment and family policies. The results after nearly 2 years suggest rather continuation with some thermostatic short-term boosts in support instead of any long-lasting change.

INTRODUCTION

The welfare state is meant to protect against social risks. This is particularly the case during a major health and economic crisis. The current COVID‐19 pandemic and the policy responses can be seen as an exogenous shock, forcing governments to take unprecedented measures to contain its spread and to mitigate its impact on society and the economy. Social policies have been key in complementing nonpharmaceutical interventions that governments used to slow the spread of COVID‐19. Health and long‐term care policies are crucial for coping with such a public health challenge. In addition, short‐time work benefits, parental leave policies, or home office regulations allowed workers to stay at home and reduce contact with others while receiving income support. Moreover, social protection such as unemployment benefits and pensions are automatic stabilizers, while further redistributive social policies can mitigate the uneven impact of the pandemic. The welfare state is thus essential in sustaining social cohesion and fostering acceptance of lockdown measures.

The public debate saw a renaissance of the welfare state as a crisis management tool, but whether public opinion has shifted towards social policies during the pandemic remains unclear. Theoretical arguments provide competing expectations on how public support for social policies might react to an increase in social needs and policy changes during the crisis. The new politics of the welfare state approach suggests that welfare state expansion fosters its support with long‐term implications for attitudes on social policies. In contrast, the attitudes‐like‐a‐thermostat‐thesis claims that welfare state expansion puts pressure on its financial sustainability and that citizens as taxpayers adapt their support downward to limit costs.

Our aim is to evaluate whether welfare state support has changed in response to the crisis and policy responses in Germany or whether public opinion remains unaltered. Using policy trackers, we provide an overview of the main social policy instruments to address the COVID‐19 pandemic. We analyze six waves of the German Internet Panel (GIP) covering 2 years before (1/2018, 1/2019, and 1/2020) and one and a half years during the COVID‐19 pandemic (6/2020, 1/2021, 9/2021) to investigate any changes compared with prepandemic times across four social policy areas: health care, pensions, unemployment benefits, and family policies.

Our survey analysis over time provides initial evidence on how the pandemic might change the politics of the postpandemic welfare state. For our analysis, Germany as a “crisis light” country serves as a least likely case to expect changes in social policy support, but if there is an increase, we would expect such a trend to apply even more in severe cases. By contrast, if social policy attitudes remain largely unchanged, such continuity may be due to the relatively smaller extent of the crisis and the well‐functioning welfare state in place.

THEORETICAL FRAMEWORK

Public attitudes towards the welfare state and particular social policies have been a well‐established field in international public opinion research, focusing on the long‐term trends in values and short‐term attitude cycles (Kumlin et al., 2021). Although research has been done on the effect of economic and political crises on changing attitudes, the Covid‐19 pandemic and its containment since 2020—the Great Lockdown—provides a unique external shock not only to welfare state capacity but also to the beliefs around social risks and its mitigation by social policy. There are competing theory‐based expectations on the possible reaction of public opinion to the role of the welfare state in the context of a crisis such as this pandemic.

On the one hand, there is the claim of a “Corona moment” for welfare state revival in public debate. Due to the exogenous nature of the pandemic, those who lose their job should not be perceived as having control over their situation, therefore we assume that this deservingness criterion is salient, while the others (attitude, reciprocity, identity) are rather unaffected by the pandemic; therefore we would expect that increased need and the perception of little control leads to more support for social policies (Knotz et al., 2022; Oorschot, 2000; Oorschot et al., 2017). Furthermore, risks and risk perceptions are key in shaping social policy attitudes (Rehm et al., 2013). If perceived unemployment risk goes up and more people expect to rely on welfare, public support should raise out of self‐interest (Rehm, 2016). People who rely on social benefits during the pandemic might further increase their support and oppose any retrenchment of these benefits (Margalit, 2013; Naumann et al., 2016; Pierson, 1993). Also, prosocial behavior, solidarity, and the willingness to cooperate are higher in situations of crisis and disaster (Quarantelli & Dynes, 1977; but see also Prainsack, 2020), though these reactions hold only if the pandemic is perceived to remain a crisis.

In contrast, some public opinion theories suggest that welfare state support will weaken in times of crisis or be rather short‐lived. Most prominently, proponents of a negative feedback claim that “in effect, the public would behave like a thermostat; when the actual policy ‘temperature’ differs from the preferred policy temperature, the public would send a signal to adjust policy accordingly, and once sufficiently adjusted, the signal would stop” (Wlezien, 1995:, 1995: 981). The model explains policy preferences as a function of a fixed, individual preference for an optimal level of policy output (for example, spending) and the current level (Soroka & Wlezien, 2010). Due to higher social expenditure and anticipated tax pressure, more citizens than before would agree that the current level is above their preference. If policy moves in one direction, e.g., provides more benefits and becomes more expensive than citizens want, public opinion swings in the other direction, signaling a preference for lowering benefits.

These contrasting perspectives provide competing expectations about the effect of the COVID‐19 pandemic on public attitudes, yet they are not mutually exclusive. It is possible (and likely) that people see the need for more benefits and perceive higher risks but are at the same time concerned about the anticipated costs. Hence, such an evaluation entails a trade‐off between need and costs. It is thus an empirical question which of the two mechanisms of attitude formation dominates during a crisis. Our research design and data do not allow us to identify the specific causal paths between the COVID‐19 pandemic and welfare state support, but they allow us to explore whether shifts occur and by whom.

In our analysis, we focus on how assumed shifts in risk pools coincide with possible changes in welfare state support. The starting point of our argument is that needs and risks, and costs, are not equally distributed across society. The overall change in social attitudes might therefore depend on how a crisis affects different sections of society. “The basic idea is that a crisis is a change to the risk pool. Different types of changes to risk pools represent different crises and can be expected to lead to different social policy consequences.” (Rehm, 2016: 184).

Rehm (2016) distinguishes three types of crises and four respective consequences for the risk distribution. The basic assumption is that the risk distribution is right‐skewed: Low‐income people have a higher risk to rely on social policies, while those of higher income are the net payers. In the first type of change, the risk distribution remains right‐skewed, yet it becomes more heterogeneous, thus risk inequality further increases. In the second type, the risk distribution remains right‐skewed, but it becomes more homogenous and risk inequality decreases. The third type is more fundamental, caused typically by economic depressions or hyperinflation, when risk distributions flip and risks are mainly concentrated among the high‐income earners. In the fourth type, crisis effects on the risk distribution are unknown and risk is replaced by uncertainty.

The first type of change in which low‐income groups are mostly affected should lead to waning welfare state support. Imagine a typical economic recession in which unemployment tends to increase disproportionally for low‐skilled workers. As risk inequality increases, social insurance becomes more expensive for a majority of net payers and the thermostatic counter‐reaction dominates. If crises also affect the middle or even the top income groups (risk changes type 2 and 3), a broad majority is expected to be favorable to risk socialization and therefore support social policy expansion. Examples are the introduction of unemployment insurance and old‐age pensions after critical emergencies or deep recessions (Castles, 2010; Cutler & Johnson, 2004). Finally, if uncertainty prevails, everyone independent of their income perceives an augmented risk and will tend to support social insurance against these risks. Wars and deep depressions are typical crises where risk is replaced by uncertainty. At least during its first wave, the COVID‐19 crisis induced a similar pattern of uncertainty as individual risks were incalculable. While unemployment experience and health risks quickly became stratified, the medium‐term income and childcare situation remained uncertain.

Empirical studies of COVID‐19 effects on social policy attitudes do not provide a clear picture so far and they are mainly restricted to short‐term effects. Overall, there is little evidence of profound shifts in attitudes towards social policies, other than increased trust in government in the short run (e.g., Bol et al., 2021), which drops as the crisis unfolds (Jørgensen et al., 2021). In a Dutch panel study (from 2017 to May 2020), Reeskens et al. (2021) found decreasing support for state intervention in social affairs in general, but slightly increasing support for redistribution while most other core political values remained stable. In contrast, Ares et al. (2021) analyze data in Germany, Sweden, and Spain (2018 vs. June 2020) and conclude that the pandemic has not led to a shift in support for state intervention but reinforced polarization with respect to (re)distributive politics. Moreover, they found some evidence that citizens positively updated their views on state capacity and trust in politicians. For the UK (April–September 2020) Blumenau et al. (2021) find no evidence that exposure to pandemic‐induced shocks affected attitudes towards the role of government in economic or social policy. Busemeyer (2021) finds stable preferences for health care spending during the pandemic. Hence, the preliminary evidence from these studies suggests that public opinion did not move as much as public debates about the revival of the welfare state and an “end of austerity” suggest.

SOCIAL PROTECTION IN GERMANY

The German welfare state is based on Bismarckian principles, particularly contributory social insurances with pay‐as‐you‐go financing of earnings‐related benefits against social risks ranging from sickness, old age, and unemployment to long‐term care. However, recent reforms have merged means‐tested minimum income benefits, made activation measures more conditional, and advanced voluntary funded pensions, thus strengthening liberal market but also universalist principles (Seeleib‐Kaiser, 2016). The pandemic revealed the scope of insufficient protection coverage for various socio‐economic groups, such as the long‐term unemployed, the self‐employed (e.g., freelancer), women with mini‐jobs or care breaks, or first‐generation immigrants (Cantillon et al., 2021). Following several pension reforms, some groups of retirees see increased poverty risks compared with those with long contributions during their working lives. While child benefits are part of German family policy, increased childcare provision and enhanced parental leave policies have been developed more recently.

While the pandemic was unprecedented and spread rapidly across Europe, Germany was affected less severely than many other large European countries, both with respect to public health outcomes and socio‐economic repercussions (Wieler et al., 2021). Mortality rates due to Covid were held at a comparatively lower level, particularly during the first wave. Partly due to forward‐looking pandemic management informed by expert advice and characterized by a strong health care system. While the national lockdown suppressed Covid‐19 infections relatively effectively during the first pandemic wave in spring 2020, the later waves resulted in higher infection and hospitalization rates, pushing the health care system temporarily to its limits. The economy did not experience as severe a decline in production as other competitors, for instance, the unemployment rate increased only modestly thanks to the use of short‐time work and other mediating measures. Data collection for the surveys during the pandemic took place in June 2020 after the first pandemic wave, in January 2021 at the end of the second wave, and in September 2021 before the fourth wave during a low infection period (Figure 1).

Figure 1

Evolution of Covid‐19 cases and stringency of lockdown policies. Source: Hale et al., 2021

Containment policies and health care

Public health responses to contain the spread of the virus were largely taken at the subnational (state) level and local public health authorities, but the federal lawmaker set the legal framework. The authorities effectively mobilized the country's laboratory capacities and established one of the strongest monitoring and test capacities in Europe, relying widely on PCR tests. Within weeks, Germany set up large‐scale local contact tracing despite the failed attempts to introduce an effective contact‐tracing app. A relatively effective containment strategy managed to reduce the viral spread among the older population, in particular limiting Covid transmission in long‐term care facilities, which reduced hospital admissions and resulted in a comparatively lower fatality rate (Wieler et al., 2021). While there are multiple social, political, and psychological effects of the containment policies, public support of the measures was particularly important as evident from the early pandemic wave (Naumann et al., 2020).

The pandemic spotlighted the strengths and weaknesses of Germany's health care system financed through sickness insurances, while the partly state‐subsidized hospital systems are governed at the subnational level. Entering the pandemic with the highest density of hospital beds and intensive care units (ICUs) in Europe, the government further expanded ICUs from 28,000 to 40,000 within a few months (Bariola & Collins, 2021). Contrary to many other countries, ICUs have not been overburdened on a broad scale. However, health care workers were confronted with temporary equipment shortages despite Germany being one of the largest producers with its strong chemical and medical equipment export industry (Bahnsen & Wild, 2021).

Even though the German company BioNTech developed one of the vaccines, the population remained relatively reluctant to get vaccinated (or had doubts about the low‐cost Astra‐Zeneca vaccine). By December 2021, only around 70% of German residents received full double vaccination placing the country third last in Western Europe only trailed by the German‐speaking neighbors Austria and Switzerland (OWiD, 2021). While the pandemic increased health risks for the entire population, the prevalence of serious illness and death is strongly skewed towards the older population and those with preconditions. Excess mortality rates were comparatively low during the first wave (spring 2020) and most working‐age citizens supported containment policies to protect the elderly and hospitals from overcrowding without much concern about their personal health from a Covid infection. While the mean health risk has increased due to the pandemic, most individuals have perceived their health risk as low. We expect the risk distribution to have become more bottom‐heavy and unequal, thus preventing increases in support for health policy despite the increase in the mean health risk.

Pensions

Pensions provided stable income maintenance during the pandemic without major changes due to safeguards. While incomes of current workers were threatened, retirement benefits were secured as an automatic stabilizer for the elderly's household income. A debate ensued on whether retirees with stable pensions would be the hidden winners of the pandemic (Bernau, 2021; Hagelücken, 2021; Steffen, 2021). However, pensions were not increased for the first time in a decade, this may negatively affect the perceived living standard of recipients. In addition, the automatic benefit uprating has been delayed to eventually catch up with wage developments. A more direct impact of the pandemic was that the health risks and labour market situation pushed some workers close to retirement age into earlier retirement than otherwise.

The financial risk of older people, particularly those already retired, remained stable and incomes were less affected by the pandemic than those of the working‐age population. However, older people were more affected by the increased health risks in case of a Covid infection, while also suffering severely from social isolation. Overall, we expect public support for the elderly to have increased, although this rests on the assumption that overall higher support due to health risks would translate into higher support for income via pensions.

Unemployment and short‐time work

The most visible social policy response addressing the employment crisis was the massive expansion of short‐time work (“Kurzarbeit”) to avoid a surge in unemployment, encompassing nearly 18% of the dependent labour force during the first lockdown (Ebbinghaus & Lehner, 2022). The model of labour hoarding, a tested measure during the previous economic crisis, was rapidly reactivated and supported by both social partners (Ebbinghaus & Weishaupt, 2021). Employers reduce employees' working hours instead of laying them off, thereby preserving the employer‐employee link. The scheme paid nearly the full labour costs of furloughed workers to employers, while wage replacement for workers was in line with unemployment insurance benefits but comparatively less generous than in other European large economies (Pusch & Seifert, 2021: 101–102). For the hours lost, employees receive at least 60 percent of their net pay as a short‐time allowance, while since Summer 2020, the benefits were topped up to 80% for longer short‐time work, while there were also sectoral collective agreements for around half of the workforce (Herzog‐Stein et al., 2021). Keeping workers in short‐time work also ensured continued payments of pension contributions securing future pension claims (Geyer et al., 2021).

While the largest part of the employment shock has been absorbed by short‐time work, the programme has some gaps, such as for marginal employment (“mini‐jobs”) or freelancers and other self‐employed (Herzog‐Stein et al., 2021). With 28% in marginal employment, the distributional impact has been particularly unequal for women (Cook & Grimshaw, 2021). The self‐employed received lump‐sum payments (“Corona Soforthilfe”), tax reductions, and deferrals of social security and tax payments. Redundant workers received contributory unemployment benefits with extended entitlement and suspended wealth conditionality, though they ended up with substantially lower income replacement compared with short‐time work. Those without or at the end of their contributory unemployment benefits received the means‐tested minimum income benefits with little uprating during the pandemic. Self‐employed who lost their income could also access unemployment assistance (“Hartz IV”) due to the suspension of the wealth test.

The distribution of unemployment risk is more bottom‐heavy than in other social policy areas and citizens face massively different probabilities of becoming unemployed depending on their occupation and skills. As a result, the unemployment policy generally enjoys lower support than other social policies. However, with the pandemic and stay‐at‐home orders, large parts of the workforce faced uncertainty about their continued employment. Following the “veil of ignorance” thesis, we expect support for unemployment policy to have increased in response to the first lockdown. With the expansion of short‐time work implemented swiftly and generously, massive dismissal was prevented and employment contracts were safeguarded for a large share of the workforce, thus easing uncertainty. However, not everyone benefitted equally and workers in specific sectors (such as hospitality or those in temporary employment) faced a substantially higher chance to be laid off. Risk inequality has thus widened for the working‐age population, which should have limited support for the unemployed as the pandemic continued.

Family policy

Families, in particular children, were affected severely by the pandemic with closures of schools and early childcare facilities. Younger children remained at home under the care of their parents. While crucial to contain the spread of the virus at the onset, this containment policy resulted in a double burden for family caregivers, mostly women. In reverse, elevated caring duties limited availability for paid employment, thereby negatively impacting household income and career progression (Hipp & Bünning, 2021; Reichelt et al., 2021). School closures were implemented at the state level, which resulted in increasing regional divergence as the pandemic progressed. Closures faced considerable popular dissatisfaction due to the massive burden on home carers, erratic decision‐making at times, and inconsistent differences between states. Policy‐makers responded by increasing time and financial resources for families, in particular, extending leave policies and increasing the universal child bonus (€300 per child offset by tax allowance for high‐income families), and the allowance for low‐income families (€185 per month) and additional support for single parents (Cantillon et al., 2021). Following existing regulations, short‐time work aimed to offset the negative impact on families with increased benefits for employees with children, topped up by 7 percent of the forgone wage. However, the adverse impact on families with children and mothers is far from being compensated. Especially those who are already affected by heavy financial burdens, such as lone parents were put under strain, thus elevating their social risks. School closures caused the strongest disruption for families, affecting virtually all children and adolescents. Risks among families flipped to top‐heavy and the distribution of risks narrowed, even though families with dual working parents and smaller living spaces were hit harder. We expect support for a family policy to have increased reflecting the flip in the distribution of risk.

PUBLIC ATTITUDES TOWARDS SOCIAL POLICIES BEFORE AND DURING COVID‐19

Data

To capture the social policy attitudes of the population in Germany, we use data from the German Internet Panel. The GIP is based on a random probability sample of the general population in Germany aged 16–75. The panel started in 2012 and was supplemented with additional participants in 2014 and 2018; participants were recruited offline using strict statistical procedures (Blom et al., 2015). Every other month, panel participants are invited to take part in a voluntary online survey. For the Mannheim Corona Study (MCS), the GIP launched a special survey (Cornesse et al., 2021) in which GIP respondents were invited to participate in weekly surveys for 16 weeks between March 20 – July 10, 2020.

In our analysis, we use the state responsibility item available for four policy areas as an indicator of public support across different social policies. Respondents are asked whether they think that it should be the “government's responsibility” to provide (1) health care for the sick, (2) a reasonable standard of living for the old, (3) a reasonable standard of living for the unemployed, and (4) a reasonable standard of living for families with children. Answer categories ranged from 0, should not be the government's responsibility at all to 10, and should be entirely the government's responsibility. Table A1 in the Appendix provides frequency distributions of these variables over time.

These items are available three times before the COVID‐19 pandemic (January 2018, except for family policies, January 2019 and January 2020), are part of the MCS in June 2020 at the end of the first lockdown, and are available in two regular GIP waves after the first wave and lockdown (January 2021, September 2021). Case numbers with valid responses vary between 2669 and 4894 per wave, while overall, our analysis relies on 23,732 responses.

Our key indicators to capture risk are age, associated with the severity of COVID‐19 but also an important group characteristic in attitude research, and income as used in many previous studies (e.g., Rehm, 2016). We recode age into three categories (35 years old and younger, 36–59 years old, 60 years and older). Relying on the self‐reported household income of respondents and the number of household members, we calculate the equivalised household income following the OECD‐modified equivalence scale, which assigns a value of 1 to the household head, 0.5 to each additional adult member, and 0.3 to each child.

We further add a set of control variables and distinguish men and women, people with and without children in the household, and the employment situation (employed or self‐employed, unemployed, not in the labour force [e.g., retired, studying, or doing housework]). Political ideology is measured with an item on left vs. right self‐placement (11‐point scale). We also include the place of residence and distinguish respondents living in the Eastern from the Western parts of Germany.

The survey results

Focusing on the overall support for government intervention with respect to key areas of the German welfare state, the patterns of public attitudes in Germany before and during Covid‐19 confirms previous comparative research: Health care is the most popular social policy area, followed by pensions, whereas family policies and support for the unemployed is less popular. The trends show increased support for social policies across the four areas, though at different levels (Figure 2, see Table A2 in the Appendix for the regression models). We find the strongest increases between short before the COVID‐19 pandemic (i.e., January 2020) and the first survey during the pandemic (June 2020) for family policies and unemployment benefits (+0.5 points). Support also increases for the more popular areas of pensions (+0.4 points) and for healthcare (+0.2 points) with already very high levels of support. In the short run (January vs. June 2020) our findings confirm the expectation that the pandemic has induced a “veil of ignorance” and led to the widespread perception that everyone might be at risk of losing a job or requiring health care. Hence, support that the government should be responsible to insure against these risks increased, and this effect is stronger in social policy areas with previously less support.

Figure 2

Support for social policies in Germany before and during the COVID‐19 pandemic. Note: Average support on an 11‐point scale, 95‐CI in gray shadings, vertical line indicates the beginning of the first COVID‐19 wave in Germany

However, these reactions to the pandemic shock are not long‐lasting. For healthcare and unemployment protection, we see public attitudes returning to their prepandemic levels within a year. With decreasing uncertainty and increasing risk inequality along already existing social divisions, the cost argument—social insurance becomes more expensive for a majority—weights in, and social solidarity with the vulnerable is receding to normal levels. We observe a similar readjustment to previous levels of support for pension and family policies in January 2021, yet in September 2021 (around the federal election) support for the government's responsibility to provide benefits for these groups reclimbs to the levels observed during the first lockdown. One limitation for interpreting attitudinal changes over time is the higher likelihood that other events than the ongoing COVID‐19 pandemic shape attitude formation. For example, in spring 2021, the German parliament announced that state pensions will not increase in 2021 (in line with stagnating wages), for the first time since 2010. While our findings show overall a short‐term increase in support that levels off within a year, only the increase in support for family policies suggests a slight recalibration towards a more family‐friendly welfare state.

These aggregate trends might hide changes in attitudes of subgroups of society. In this article, we examine two of the most widely researched conflict lines. Hence, focusing on age and political ideology allows us to explore whether the COVID‐19 pandemic has increased the generational conflict or fuelled ideological polarization over the welfare state. In the following analyses, we look at these potential cleavages and examine whether the strength of these conflicts over the role of the state has changed.

The left column of graphs in Figure 3 shows the predicted support for three different age groups: the young (<35 years old), the middle‐aged (35–59 years old), and the old (60 years or older). The right column shows the difference in support between young and middle‐aged and between young and old, it serves as an indicator of the strength of the generational conflict (controlling for other socio‐demographic differences, which might exist between age groups, see regression models in Table A2). We do not find a generational conflict over health care. In the three other areas of the welfare state, the elderly show lower support than the young. This difference is slightly increasing over pensions and unemployment benefits, yet this seems to be a general trend that already emerged pre‐COVID. The difference between old and young in support for families is decreasing over time. Again, there is no evidence that this trend has been affected by the COVID‐19 pandemic.

Figure 3

Age differences in support for social policies—predicted values based on regression models with control variables (Table A2)

Following the same logic as Figure 3, we explore risk polarization over social policies in Figure 4 and show predicted support for low‐income (at risk of poverty), middle‐income (3rd to 8th income decile), and high‐income (9th and 10th decile) earners in the left column of graphs. In the right column, we focus on the difference between high‐ and low‐income earners. In all areas of social policies, low‐income earners are more supportive than those with middle or high incomes. These attitude differences between rich and poor are strongest for pensions. Yet, the time trends in these differences do not provide any evidence for a risk shift after the pandemic and the income groups rather move in tandem over time following the overall time trend.

Figure 4

Income and support for social policies—predicted values based on regression models with control variables (Table A2)

CONCLUSION

The COVID‐19 pandemic led to intensified needs for social protection and created uncertainty about the future, it has been seen in public debate as a critical juncture for the revival of the welfare state. Theoretical expectations that such a crisis entails increased needs, enhanced deservingness, and more state responsibility would suggest an increase in public support for more government responsibility across major social policies. However, more cautious voices point at the short‐term nature and thermostatic cycles in response to increased state activities. Indeed, this is supported by recent studies across Europe suggesting that welfare state support has not changed much except for trust in government.

In our study, we compared the development of German public opinion, relying on high‐quality panel data and three waves prior to and three waves during the first 1 ½ years of the pandemic. This allows us to reliably examine whether immediate, short‐term reactions to the COVID shock are lasting beyond the initial shock. We found evidence for a strong tendency of continuation in welfare state support and some indication of rather short‐term boosts in additional support during the pandemic. While health care and pensions have always been popular, we noted more significant changes among the less supported unemployment policies and family‐related measures. In particular, the first pandemic wave has led to an increase in three of the four policy areas, but not in health care, most likely reflecting the already relative widespread support for the German healthcare system.

There has been little evidence that socio‐demographic or political differences have been altered and thus, the thermostatic up and downs are reflecting rather general trends than substantial shifts in particular groups. These results suggest a boost of support in areas that were more visibly negatively affected during the initial crisis. With declining uncertainty, a soberer public opinion tended to return to the long‐held value‐orientation about welfare state support along the four social policy areas studied here. Germany is seen as a least likely case to find attitude changes because both the economic and the health effects of COVID‐19 were rather limited in comparison to other European countries that have been hit much harder by the pandemic.

Hence, while our study provides evidence for a thermostatic reaction, we would not entirely disregard the other theoretical approaches. This holds also because our data cover a rather short period. While we can capture short‐term changes, we lack thus far evidence on the medium‐ or long‐term trends. A more long‐term analysis might not only provide a fruitful perspective for future empirical studies, but we would suggest adding a distinction between short‐term preference changes and long‐term value evolution to the set of theoretical explanations.

Table A1

Click here for additional data file.