The Shift of U.S. Public Attitudes Toward Welfare State: A Cohort Comparison between Pre and Post Economic Crisis Era

Eun Kyung Lee, Adelphi University School of Social Work

Tae Kuen Kim, Adelphi University School of Social Work

1 Introduction

The United States is considered a reluctant welfare state as its social policies and welfare programs are less comprehensive and less extensive than those of other developed countries (Jansson, 2019). The U.S. has not adopted basic welfare system, such as universal health insurance, child allowance, and sickness benefits, at the national level that most industrialized countries implemented decades ago. As Esping-Anderson (1998) mentioned, the U.S. is the epitome of a liberal welfare regime, with limited state intervention and an emphasis on personal responsibility. For decades, scholars have questioned why the U.S. welfare state is exceptionally retarded and why the federal government is hesitant to intervene in the private sector, compared to other developed countries (Prasad, 2016). While there have been possible explanations including strong federalism, fragmented political institutions, and insubstantial cooperative tradition, scholars argued that the U.S. has a unique socio-political culture that highly emphasizes self-reliance and individual freedom, and this individualism underlies the lagged U.S. welfare state, or American exceptionalism (Bobo, 1991; Feldman and Zaller, 1992; Iversen and Soskice 2009, Katznelson, 2014; Turner, 1996).

As individualism emphasizes that individuals in a society should be responsible for their own well-being, it argues that governments should play a limited role in providing public assistance (Cheng and Ngok, 2020). Therefore, societies dominated by individualistic ideologies are more likely to have lower levels of support for collective intervention, which is at the heart of the welfare state. Research empirically shows that higher levels of individualism lead to negative attitudes towards the state’s welfare responsibility among citizens (Toikko and Rantanen, 2020). In this context, prior comparative studies also revealed that the public’s welfare attitudes in the U.S. are generally less supportive than in many European counter countries where egalitarian value is relatively prevailing (Andreß and Heien, 2001; Arikan and Bloom, 2014).

Since the turbulence of 2009 Economic Crisis, however, there has been a growing skepticism of individualism and free-market economics in the U.S. society, as well as an expanding sentiment in favor of government intervention and socialized protection. In response, the results of current opinion polls indicate majority of the U.S. citizens support redistributive tax system and progressive social policies such as universal childcare, expanded unemployment benefits and free college education (Kim, 2021). The significant shift in public consensus among U.S. citizens from absolute trust in individualism to an increased concern for the role of government and common well-being was one of the key factors that enabled the Biden administration to propose and implement large package bills of progressive welfare programs during the COVID-19 pandemic (Kim, 2020). In this sense, the U.S. welfare state is in the midst of a transition.

Against this backdrop, this study exploratively examines how public’s attitudes toward welfare state have changed in the U.S. over the past decade. Whereas most previous research (Baranowski and Jabkowski, 2022; Bean and Papadakis, 1998; Blekesaune and Quadagno, 2003; Goossen, 2020; Jakobsen, 2011; Kurowska et al., 2019; Larsen, 2008; Ng, 2015) used a cross-sectional approach to identify individual or country level factors affecting welfare attitudes, this study employs a longitudinal approach to trace the dynamic of welfare attitudes in the U.S. If welfare attitudes of a society are stable over time, the findings of cross-sectional comparison would be valid (Van Heuvelen, 2014). However, assuming the U.S. society has undergone a shift in public’s attitudes toward the welfare state since the 2009 Economic Crisis, this study seeks to identify the trajectory of American attitudes toward welfare states by longitudinally comparing welfare attitudes in the pre-crisis period with those in the post-crisis period. Through a cohort comparsion, we also pay attention to how individuals’ characteristics determine welfare attitudes and how the influences of these characteristics differ between two cohorts, pre-crisis cohort and post-crisis cohort.

2 Theoretical Background and Research Aim

2.1 Self-Interest theory and welfare attitudes

While there have been several theories that explain determinants of attitudes toward welfare state, the self-interest perspective has been a prominent theoretical framework in welfare attitudes research. The self-interest perspective focuses on how individuals' economic and social interests influence their attitudes toward welfare state. This perspective suggests that as welfare attitudes are primarily motivated by self-interest, or egocentric concerns, individuals are more likely to support welfare state when they expect the benefits of welfare policies and programs to outweigh the costs they should bear (Emilsson, 2022). According to this perspective, individuals who are more likely to face social and economic risks such as unemployment and financial insecurity have more supportive attitudes toward welfare state (Calzada, 2014; Wong et al., 2009). On the other hand, those who are in socio-economically stable circumstance are less likely to support welfare state, because they may perceive welfare policies and programs which usually involve higher tax as a threat to their own economic advantage and view government intervention as counterproductive (Breznau, 2010; Jæger, 2006). The self-interest perspective also argues that demographic factors, such as race and gender, influence welfare attitudes. Due to their past experiences of discrimination and economic disadvantage, historically marginalized groups, such as the racial minorities, immigrants, and women, may have more supportive attitudes toward welfare state.

However, critics of the self-interest perspective argue that it oversimplifies the complex mechanism shaping publics’ welfare attitudes and underestimates the role of individuals’ internalized norms such as preferences to specific social and cultural values as well as political ideology. Empirical research shows welfare attitudes are significantly associated with religious beliefs and the political spectrum of individuals and these value systems sometimes mediate the relationship between self-interest and welfare attitudes (Kulin and Svallfors, 2013; Van Heuvelen, 2014). Thus, welfare attitudes study needs to consider individuals’ value systems.

In addition, current literature suggests the conditional self-interest perspective, focusing on the interaction between self-interest and institutional or economic context of a society (Sachweh, 2018). According to the conditional self-interest perspective, the impact of self-interest on welfare attitudes can be moderated by social context. For example, individuals in a society providing universal entitlements and comprehensive social welfare benefits with all citizens, such as social democratic welfare regimes, are expected to be supportive of welfare state, regardless of their social class and position in a labor market. In contrast, in a society where selective and means-test based welfare policies and programs are dominantly developed, support for welfare state is expected to be stratified by socio-economic groups because welfare benefits are only delivered to those who are in need. As a result, institutional or economic context influences individual attitudes towards welfare state through the conditioning of individual self-interest (Arikan and Bloom, 2014; Arts and Gelissen, 2001).

2.2 2009 Economic Crisis and the U.S. welfare attitudes

The conditional self-interest perspective provides a valid theoretical framework when examining public welfare attitudes during economic crises. As Blekesaune (2007) argued, when economic conditions are favorable, individuals tend to take more personal responsibility and accept less government intervention, but they are more likely to demand government protection during tough economic times. According to the conditional self-interest perspective, economic crises can reinforce public’s welfare attitudes because individuals become more supportive of the welfare state when they perceive social benefits as helping to mitigate the negative effects of economic crises (Blekesaune, 2013). Empirical research confirms that experiences of crisis-induced reductions in material standard of living such as lay-offs and wage reductions increase individuals’ support for welfare state (Margalit, 2013; Owens and Pedulla, 2014).

In this context, scholars in the field of social welfare have focused on how the 2009 Economic Crisis, or the Great Recession, influences public’s welfare attitudes toward welfare state. While several empirical studies (Brito Vieira et al., 2017; Sachweh, 2018; Sachweh, 2019; Wang et al, 2017) examined the impact of the Economic Crisis on welfare attitudes among European or Asian countries, previous research has paid little attention to the U.S. case. As the epicenter of the Economic Crisis and one of its biggest victims, the U.S. experienced tremendous socioeconomic problems such as housing and stock market crashes, personal bankruptcies, mass unemployment, and wage cuts. In addition, the U.S. federal government implemented unprecedentedly large-scale market interventions and social relief programs to mitigate the negative effects of the crisis, including bailing out banks, expanding unemployment benefits, and distributing massive economic stimulus (History of the Federal Reserve System, 2013). The U.S. experience thus provides important clues for understanding the impact of economic crises on public’s welfare attitudes in societies experiencing rapid economic and institutional changes.

Based on the above discussion, supposing that changes in the economic and institutional contexts triggered by the Economic Crisis in the U.S. have led to the change in Americans’ attitudes toward welfare state, this study specifically hypothesized that public welfare attitudes increased in the U.S. after the Economic Crisis (H1). This study also hypothesized that the influence of individuals’ socio-demographic characters and values on welfare attitudes changed after the crisis (H2).

3 Methodology

3.1 Data Source

This study used data from the World Values Survey (WVS). The WVS is a global longitudinal research project that investigates people's values and beliefs. Since its inception in 1981, The WVS has collected data from approximately 100 countries around the world. As the largest international survey of human values and attitudes, the VWS is designed to measure people's attitudes and beliefs about diverse areas such as politics, religion, family, work, gender roles, and social issues (World Values Survey Association, 2020). As the survey uses a standardized questionnaire to ask respondents' views on a variety of topics in each wave, it allows researchers to understand changes in the beliefs, values, and motivations of people over time. The WVS uses a combination of multistage and stratified random sampling method, which ensures that the sample of respondents is representative of the population being studied and allows for generalization of the survey findings to the broader population. The survey has been conducted every five years, and the most recent wave, Wave 7, was conducted from 2017 to 2020 (World Values Survey Association, 2020). We employed Wave 5 conducted in 2006 as the pre-crisis cohort and Wave 7 conducted in 2017 as the post-crisis cohort.  The sample sizes of 2006 cohort and 2017 cohort are 1,249 and 2,596, respectively.

3.2 Measurement and variables

To measure attitudes toward welfare state, the dependent variable of the study, we developed an index of generalized welfare attitudes. Following Jakobsen’s (2011) approach, we aggregated three questions (indicators) from the VWS, which ask respondents’ attitudes regarding (1) income equality, (2) government responsibility, and (3) redistribution through tax, respectively. The questions are:

How would you place your views on this scale? 1 means you agree completely with the statement on the left; 10 means you agree completely with the statement on the right; and if your views fall somewhere in between, you can choose any number in between.

Q1: Incomes should be made more equal (1) vs. There should be greater incentives for individual effort (10)

Q2: People should take more responsibility to provide for themselves (1) vs. The government should take more responsibility to ensure that everyone is provided for (10)

Q3: “Governments tax the rich and subsidize the poor” is not an essential characteristic of democracy (1) vs. “Governments tax the rich and subsidize the poor” is an essential characteristic of democracy (10)

Higher values for Q2 and Q3 indicate a pro-welfare attitude, while the opposite is true for Q1. To create a unidirectional index, we reverse-coded Q1. The result is that all three items range from 1 to 10, with higher values indicating a pro-welfare attitude. We calculated the unweighted average of these three items and used this average as the level of supportive attitudes toward welfare state. Table 1 shows the intercorrelation of the three items that were combined to measure attitudes toward welfare state.

Table 1. Correlation matrix between items measuring welfare attitudes

	Income Equality	Government responsibility	Redistribution through tax
Income Equality	1	.399**	.300**
Government responsibility	-	1	.380**
Redistribution through tax	-	-	1
*p<.01, **p<.001

This study employed socio-demographic variables as predictors of welfare attitudes, including age (Young: 18-40, Middle: 41-65, Old: Over 65) sex (Female, Male), race (White, Black, Hispanics, Others), education level (High school or below, College or Bachler degree, Master or Professional degree), marital status (Married, Unmarried), employment status (Employed, Unemployed, Retired, Others), health condition (Poor/Fair-Good-Very good condition), social class (Lower-Middle-Upper class) and income level (Low-Medium-High level). This study also included political affiliation (Democrats, Republican, Others) and religiosity (Religious, Non-religious, Atheist) to measure individuals’ values. To measure subjective social class, respondents were asked “Would you describe yourself as belonging to (1) Lower class (2) Middle class (3) Upper class?” For the income level variable, respondents were asked “We would like to know in what group your household is (1) Lower level (2) Medium level (3) High level. Please, specify the appropriate number, counting all wages, salaries, pensions and other incomes that come in.” To measure the religiosity variable, respondents were asked “Independently of whether you attend religious services or not, would you say you are (1) a Religious person (2) Not a religious person (3) an Atheist?” We employed social class and income level variables as indicators to measure socially disadvantage groups and used other socio-demographic variables as control variables. Appendix 1 shows descriptive statistics of all these variables.

3.3 Analytical procedure

To test research hypotheses, we compared the means of generalized welfare attitudes of pre-crisis cohort to that of post-crisis cohort, using the independent sample t-test. We also analyzed the changes of welfare attitudes by socio-demographic and value groups. To examine the degree of variability and the effect size of predictors on welfare attitudes, we calculated ranges and Eta squared (η2) values. Finally, to determine how individuals’ characteristics determine welfare attitudes and how the influences of these characteristics differ between pre-crisis cohort and post-crisis cohort, we ran two separate multiple linear regression analysis for each cohort. All statistical analysis used a two-tailed test with the 99% confidence interval.

4 Results

4.1 The change of welfare attitudes in the U.S.

As shown in Table 2, while the mean generalized welfare attitudes score for 2006 cohort, pre-crisis period, was 5.01 (with SD=1.74), it rose to 5.69 (with SD=2.07) for 2017 cohort, post-crisis period. The result of an independent sample t-test confirms that there is a statistically significant increase in welfare attitudes after the Economic Crisis. The mean scores of all three indicators, income equality, government responsibility, and redistribution through tax, also significantly increased after the crisis. Considering Cohen’s d, income equality indicator saw largest increase (with Cohen’s d=.42), while government responsibility has the smallest increase (with Cohen’s d=.09). This result means that Americans' views on income equality have changed the most progressive throughout the Economic Crisis, while their perspectives on government responsibility have marginally changed.

Table 2: Descriptive statistics for welfare attitudes indicator and 3 sub-scales

	2006 Cohort		2017 Cohort		Test scores
	Mean	SD	Mean	SD	t-score	Cohen's d
Income Equality	4.92	2.26	6.07	2.89	12.09**	0.42
Government responsibility	5.06	2.67	5.32	2.98	2.58*	0.09
Redistribution through Tax	5.01	2.61	5.72	2.51	7.94**	0.28
Welfare Attitudes	5.01	1.74	5.69	2.07	9.91**	0.35
*p<.01, **p<.001

Additionally, we compared changes in welfare attitudes before and after the crisis across countries to see if the shift in welfare attitudes in the U.S. was an isolated or global phenomenon. We analyzed 15 Organization for Economic Co-operation and Development (OECD) member countries available in VWS.

Table 3: Welfare attitudes comparison by countries

Countries	2006 Cohort		2017 Cohort		Mean Change
	Mean	Rank order	Mean	Rank order
Australia	5.53	13	5.61	11	0.08
Finland	6.01	10	5.73	7	-0.28
France	5.75	12	5.50	13	-0.25
Germany	6.71	1	6.07	4	-0.64
Japan	6.13	9	6.54	2	0.41
Korea	6.49	2	5.62	10	-0.87
Netherlands	5.86	11	5.51	12	-0.35
Norway	6.14	7	5.74	6	-0.40
Poland	5.51	14	4.86	15	-0.65
Slovenia	6.17	6	5.22	14	-0.95
Spain	6.14	8	6.11	3	-0.03
Switzerland	6.34	4	5.69	8	-0.65
Turkey	6.44	3	5.93	5	-0.51
USA	5.01	15	5.69	9	0.69
Average (SD)	6.03 (.44)		5.75 (.45)		-0.27

Table 3 shows the average score for each country's generalized welfare attitudes for the pre-crisis and post-crisis cohorts. Of the 14 countries, only three - the U. S., Japan, and Australia - have seen an increase in welfare attitudes since the Economic Crisis, while the rest have seen a decrease. Furthermore, the overall average welfare attitude of the countries analyzed was 6.03 (with SD=.44) before the crisis and dropped to 5.75 (with SD=.45) after the crisis. The United States, however, shows the largest increase with .69 in welfare attitudes among the 15 countries. These results indicate that the impact of the Economic Crisis on welfare attitudes varies across countries, suggesting that the exceptionally large increase in welfare attitudes in the U.S. is not a global phenomenon, but rather unique to the U.S.

4.2 Welfare attitudes by socio-demographic and value groups

In the previous analysis, we found that welfare attitudes in the U.S. increased significantly after the crisis. To explore this more specifically, we analyzed how welfare attitudes changed by socio-demographic and value groups. Table 4 shows the change in welfare attitudes by groups in the pre-crisis and post-crisis cohorts. In almost all groups, welfare attitudes significantly increase after the crisis. The findings suggest that the Economic Crisis has led to a more positive attitude toward welfare state among the entire U.S. population, regardless of socioeconomic characteristics or political and religious values. However, the growth rate of welfare attitudes varies by group.

In particular, the ‘old age group’ saw the largest increase in welfare attitudes compared to other age groups, with a 1.02 increase. Among racial groups, the change in welfare attitudes was largest for minority groups (other groups), with a 1.25 increase. By social class, the ‘middle/upper class’ had an increase of 1.19, which was three times that of the ‘lower class/worker class’, which had an increase of 0.43. Political variables also showed differences in welfare attitudes between groups, with a 1.05 increase among ‘Democrats’, almost double the increase among ‘Republicans’ and ‘other groups’. By religious groups, the ‘atheist group’ had the largest increase in welfare attitudes at 1.21. We also analyzed in more detail how disagreements about welfare attitudes have changed in each socio-demographic and value group since the Economic Crisis.

Table. 4: Welfare attitudes changes in the U.S. between 2006 and 2017

Variables	Groups	2006 Cohort	2017 Cohort	Mean Difference	t-score
	Young	5.10	5.94	0.84	8.57**
Age	Middle	5.09	5.55	0.46	4.56**
	Old	4.45	5.47	1.02	6.08**
Sex	Male	4.95	5.47	0.52	5.47**
	Female	5.05	5.91	0.86	9.53**
	White	4.84	5.55	0.71	8.79**
Race	Black	5.65	5.92	0.27	1.53
	Hispanic	5.51	5.97	0.46	2.95**
	Others	4.78	6.03	1.25	5.05**
	High or below	5.11	5.64	0.53	5.22**
Education Level	College / BA	4.82	5.67	0.85	8.50**
	Master or above	5.32	5.83	0.51	3.04*
Marital Status	Married	4.85	5.59	0.74	8.31**
	Unmarried	5.17	5.83	0.66	6.75**
	Employed	4.99	5.69	0.70	8.35**
Employment	Unemployed	5.31	6.15	0.84	3.04*
Status	Retired	4.39	5.37	0.98	6.06**
	Others	5.4	5.91	0.51	3.08*
	Fair/Poor	5.24	5.88	0.64	4.44**
Health Condition	Good	4.92	5.67	0.75	8.40**
	Very good	4.99	5.49	0.50	3.72**
	Upper/Upper Middle	4.31	5.5	1.19	8.41**
Social Class	Middle	5.03	5.78	0.75	6.89**
	Lower/Working	5.36	5.79	0.43	4.06**
	Low	5.35	5.99	0.64	4.31**
Income Level	Medium	4.91	5.62	0.71	9.31**
	High	4.86	5.51	0.65	2.17
	Republican	3.97	4.41	0.44	3.85**
Political Affiliation	Democrat	5.59	6.64	1.05	11.09**
	Others	5.24	5.79	0.55	4.54**
	Religious	4.9	5.37	0.47	5.49**
Religiosity	Not religious	5.19	6.05	0.86	6.35**
	Atheist	5.35	6.56	1.21	4.26**
*p<.01, **p<.001
1. Mean difference is calculated by subtracting the mean of 2006 from the mean of 2007 for each cohort

Table 5 shows the degree of homogeneity in welfare attitudes in each group and the effect size of each predictor on welfare attitudes. Since the Economic Crisis, the variability of welfare attitudes by sex, political affiliation, and religiosity has increased, while the variability of all other predictors has decreased. These results suggest that the homogeneity of welfare attitudes by sex, political affiliation, and religiosity has decreased, while homogeneity has increased in other groups. These results indicate the difference in welfare attitudes between socially disadvantaged groups and socially advantaged groups had decreased after the economic crisis.  In order to examine the effect size of each predictor on welfare attitude, we analyzed Eta squared (η2) values. While Eta squared values of most predictors have decreased after the Economic Crisis, those of sex, political affiliation, and religiosity has increased. In the pre-crisis cohort, race, employment states, social class and political affiliation have relatively large effect size on welfare attitudes. However, after the Economic Crisis, the effect size of political affiliation was magnified (with η2=.217), the effect size of the remaining predictors was reduced, and religiosity emerged as a newly important predictor (with η2=.038).

Table 5: Degree of homogeneity and effect size

Variables	2006 Cohort			2017 Cohort
	Variability	F-score	η^2	Variability	F-score	η^2
Age	0.61	29.86**	0.017	0.47	12.09**	0.010
Sex	0.10	0.98	0.001	0.43	27.55**	0.011
Race	0.86	13.08**	0.033	0.49	8.03**	0.010
Education Level	0.39	4.68*	0.008	0.19	1.60	0.001
Marital Status	0.32	9.82*	0.008	0.23	7.97*	0.003
Employment Status	1.01	12.64**	0.032	0.78	8.05**	0.010
Health Condition	0.31	2.53	0.004	0.39	5.61*	0.005
Social Class	0.93	27.26**	0.046	0.29	5.14*	0.004
Income Level	0.49	6.50*	0.012	0.48	7.09*	0.006
Political Affiliation	1.62	105.54**	0.153	2.23	345.25**	0.217
Religiosity	0.45	3.96	0.007	1.19	49.50**	0.038
*p<.01, **p<.001
1. Variability is calculated by subtracting the minimum score from the maximum score among groups in a given variable. 2. Eta squared (η^2) is estimated based on the unconditional fixed effect model.

4.3 Multivariate analysis

We built three multiple linear regression models to find the predictors with the significant impact on welfare attitudes. To test the first hypothesis, public welfare attitudes increased in the U.S. after the Economic Crisis, we pooled the two data sets (2006 and 2017 samples) and included “wave (time)” variable, a dichotomous variable using 2006 as a reference category. The result of Model A in the Table 6 shows that the regression coefficient of the wave variable is .697 with p<0.01, while controlling for the covariates. This result indicates that the welfare attitudes of the post-crisis cohort (2017 sample) significantly increased, compared to the pre-crisis cohort (2006 sample), confirming the first hypothesis.

To test the second hypothesis, the influence of individuals’ socio-demographic characters and values on welfare attitudes changed after the crisis, we ran two separate regression models (Model B and Model C), and compared these two models. As shown in Table 6, in the pre-crisis cohort, age, education, employment status, social class, and political affiliation were statistically significant predictors, controlling for other variables. In particular, younger and middle-aged people, those with higher education (master's degree), those not employed, and those belonging to the lower/working class were more likely to have pro-welfare attitudes. In addition to these socio-demographic attributes, political affiliation is an important determinant for welfare attitudes among pre-crisis cohort. Controlling for all other factors, Democrats show more supportive welfare attitudes compared to Republicans and others.

Table 6: Fixed-effect model on welfare attitudes

Variables	Groups	Estimates: regression coefficients (s.e.)
		Model A	Model B	Model C
		Pooled Sample	2006 Sample	2017 Sample
Age	Young	.174 (.114)	.463 (.201)*	-.037 (.146)
	Middle	.010 (.102)	.403 (.184)*	-.211 (.128)
	Old (ref.)	-	-	-
Sex	Male	-.141 (.061)	-.003 (.095)	-.270 (.078)**
	Female (ref.)	-	-	-
Race	Black	.082 (.114)	.340 (.199)	-.036 (.141)
	Hispanic	-.182 (.141)	.424 (.242)	-.451 (.178)
	Others	-.012 (.133)	.580 (.239)	-.257 (.163)
	White (ref.)	-	-	-
Education Level	High or below	-.189 (.089)	-.492 (.164)**	-.103 (.109)
	College / BA	-.139 (.082)	-.372 (.154)*	-.073 (.099)
	Master (ref.)	-	-	-
Marital Status	Married	-.032 (.063)	.011 (.095)	-.029 (.080)
	Unmarried (ref.)	-	-	-
Employment Status	Unemployed	.270 (.127)	.080 (.257)	.341 (.148)
	Retired	-.110 (.102)	-.085 (.185)	-.146 (.122)
	Others	.258 (.086)	-.420 (.123)*	-.178 (.116)
	Employed (ref.)	-	-	-
Health Condition	Fair/Poor	-.200 (.086)	.103 (.149)	.289 (.108)
	Good	-.099 (.074)	.002 (.107)	.134 (.095)
	Very good (ref.)	-	-	-
Social Class	Upper/Upper Middle	-.336 (.086)**	-.752 (.137)**	-.143 (.110)
	Middle	-.109 (.074)	-.242 (.117)	-.052 (.095)
	Lower/Working (ref.)	-	-	-
Income Level	Low	.034 (.135)	-.145 (.216)	.138 (.172)
	Medium	-.010 (.115)	-.266 (.187)	.086 (.147)
	High (ref.)	-	-	-
Political Affiliation	Democrat	1.950 (.072)**	1.503 (.115)**	2.155 (.092)**
	Others	1.211 (.081)**	1.162 (.125)**	1.226 (.104)**
	Republican (ref.)	-	-	-
Religiosity	Religious	-.612 (.125)**	-.245 (.258)	-.668 (.146)**
	Not religious	-.388 (128)*	-.214 (.264)	-.337 (.128)
	Atheist (ref.)	-	-	-
Wave (Time)	2017	.697 (.065)**
	2006 (ref.)
Intercept		5.378 (.247)	5.841 (.462)	6.632 (.321)

The post-crisis cohort, on the other hand, shows a different pattern in the relationship between the predictors and welfare attitudes. The effect of age, educational level, employment status, and social class on welfare attitudes disappear after the Economic Crisis. In the post-crisis cohort, sex and religiosity emerged as important predictors, along with political affiliation. According to the analysis, females are more like to have pro-welfare attitudes than males and religious individuals are less supportive to welfare state, compared to non-religious and atheist individuals. Political affiliation was still a significant determinant of welfare attitudes in the post-crisis cohort, but it is worth noting that the gap in welfare attitudes between Democrats and Republicans widened from 1.4 in the pre-crisis cohort to 2.2 in the post-crisis cohort. These results confirm the second hypothesis.

5 Conclusion and Implications

This study empirically examined the change of welfare attitudes in the U.S. before and after the Economic Crisis. The findings of the study can be summarized in three ways. Frist, Americans’ support for the welfare state increases after the Economic Crisis. The results of the cross-country comparison show that welfare attitudes in the U.S. increased significantly compared to other countries after the Economic Crisis. This implies that the post-crisis increase in welfare attitudes is a uniquely American phenomenon, considering the Economic Crisis and its consequences pose a challenge for welfare state solidarity in Europe. As shown in the result section, only three countries, the U.S., Australia, and Japan saw an increase in welfare attitudes. According to Esping-Anderson’s typology (1998), these three countries belong to the liberal welfare regime which emphasizes minimal state intervention. Social welfare benefits in the liberal welfare regime are generally modest and intended to provide a basic safety net rather than comprehensive support. Other countries used in the analysis belong to the conservative or social-democratic welfare regimes where the state’s role in welfare provision is relatively strong and there is a comprehensive social security system. These results imply that the impact of the economic crisis on welfare attitudes varies across welfare regimes. Thus, future research should compare welfare attitudes across welfare regimes to explore the dynamics in more detail.

Second, research shows that after the Economic Crisis, welfare attitudes in the U.S. have improved across all groups, regardless of socio-demographic background and political and religious beliefs. The study also found a convergence in welfare attitudes across all groups. For example, before the crisis, the difference between welfare attitudes of the upper middle class and lower class was .75, but after the crisis, the difference significantly decreased to .14. Theses results imply that support for the welfare state has spread across the U.S. society since the economic crisis. Third, the factors that influence welfare attitudes have changed significantly before and after the Economic Crisis. Before the crisis, controlling for other covariates, social class and political affiliation were important variables deciding welfare attitudes. After the crisis, however, political affiliation, and religiosity stood out as critical factors. There results suggest that welfare attitudes have shifted from issues related to social class to issues related to value system, such as political affiliation and religiosity after the economic crisis. Moreover, it is worth noting that welfare attitudes have been polarized by political affiliation and religiosity since the economic crisis. As political affiliation and religiosity are leading measure of ideology in the U.S., the results arguably imply that the ideological gap between conservative and liberal groups has widened, and as a result, welfare attitudes are polarized. Religious individuals, especially those affiliated with conservative denominations, are more likely to align with conservative political ideologies that advocate for limited government intervention and lower taxation. This political stance translates into opposition to expansive welfare state. We recommend future research investing interaction of political affiliation and religiosity and its impact on welfare attitudes in the U.S.

Public support for the welfare state is crucial for its development and sustainability, as positive welfare attitudes among citizens can encourage policymakers to allocate more social resources to welfare programs and expand the social security net. The findings of the study imply that the Economic Crisis has increased Americans' supportive attitudes toward welfare state and therefore the likelihood of an expanded U.S. welfare state in the future. Meanwhile, the study reveals that welfare attitudes are increasingly polarized by political affiliation and religiosity, suggesting that a big challenge for the U.S. welfare state is how to reconcile political and cultural conflicts in future welfare reforms.

While this study explored the changes in welfare attitudes after the economic crisis, it has several limitations.  First, as the unit of analysis of this study is country-level, not state-level, we could not identify regional differences in attitudes towards the welfare state. Considering the importance of federalism in the nature of social welfare provision in the U.S., welfare attitudes are expected to vary by state. Thus, future research needs to examine the different patterns in welfare attitudes by the state in the U.S. Second, while this study shows that religiosity has emerged as a major determinant of welfare attitudes in the U.S. after the economic crisis, it does not examine different welfare attitudes by specific religions. People belonging to other religions may have different attitudes toward the welfare state. Future research needs to include the type of religion as a variable along with religiosity. On a final note, as this research is a correlational study, a kind of cross-sectional study, it is impossible to identify the true causal relationship. Therefore, we suggest future research pay special attention to a longitudinal approach.

References:

Andreß, H., & Heien, T. (2001). Four Worlds of Welfare State Attitudes? A Comparison of Germany, Norway, and the United States. European Sociological Review, 17 (4), 337-356.

Arikan, G., & Bloom, P. B. (2014). Social Values and Cross-National Differences in Attitudes towards Welfare. Political Studies, 63(2), 431-448.

Arts, W., & Gelissen, J. (2001). Welfare States, Solidarity and Justice Principles: Does the Type Really Matter?. Acta Sociologica, 44(4), 283–299.

Baranowski, M., & Jabkowski, P. (2022). How many worlds of welfare state attitudes are there? European experiences in a comparative perspective of cross-national survey research. Economics and Sociology, 15(1), 160-177

Bean, C., & Papadakis, E. (1998). A Comparison of Attitudes towards the Welfare States in Different Institutional Regimes, 1985–1990. International Journal of Public Opinion Research, 10(3), 211–236.

Blekesaune, M. (2007). Economic conditions and public attitudes to welfare policies. European Sociological Review, 23(3), 393–403.

Blekesaune, M. (2013). Economic strain and public support for redistribution: A comparative analysis of 28 European countries. Journal of Social Policy, 42(1), 57-72.

Blekesaune, M., & Quadagno, J. (2003). Public Attitudes toward Welfare State Policies: A Comparative Analysis of 24 Nations. European Sociological Review, 19(5), 415-427.

Bobo, L. (1991). Social Responsibility, Individualism, and Redistributive Policies. Sociological Forum, 6(1), 71-92.

Breznau, N. (2010). Economic equality and social welfare: policy preferences in five nations. International Journal of Public Opinion Research, 22(4), 458-484.

Brito Vieira, M., Carreira da Silva, F., & Pereira, C. R. (2017). Waiting for Godot? Welfare Attitudes in Portugal before and after the Financial Crisis. Political Studies, 65(3), 535-558.

Calzada, I., Gómez-Garrido, M., & Moreno, L. (2014). It is not only about equality: a study on the (other) values that ground attitudes to the welfare state, International Journal of Public Opinion Research, 26(2), 178–201.

Cheng, Q., & Ngok, K. (2020). Welfare Attitudes Towards Anti‑poverty Policies in China: Economical Individualism, Social Collectivism and Institutional Differences. Social Indicators Research, 150(3), 679-694.

Emilsson, K. (2022). Attitudes towards welfare and environmental policies and concerns: A matter of self-interest, personal capability, or beyond?. Journal of European Social Policy, 32(5), 592-606.

Esping-Andersen, G. (1998). The Three Worlds of Welfare Capitalism. Princeton, New Jersey: Princeton University Press.

Federal Reserve History (2013) The Great Recession and Its Aftermath. Retrieved March 2023, from https://www.federalreservehistory.org/essays/great-recession-and-its-aftermath#:~:text=Effects%20on%20the%20Broader%20Economy,-The%20housing%20sector&text=The%20decline%20in%20overall%20economic,recession%20since%20World%20War%20II.

Feldman, S., & Zaller, J. (1992). The Political Culture of Ambivalence: Ideological Responses to the Welfare State. American Journal of Political Science, 36(1), 268-307.

Goossen, M. (2020). The gender gap in welfare state attitudes in Europe: The role of unpaid labour and family policy. Journal of European Social Policy, 30(4), 452-466.

Iversen T., & Soskice, D. (2009). Distribution and redistribution: the shadow of the nineteenth century. World Politics, 61(3), 438–486.

Jakobsen, T. G. (2011). Welfare Attitudes and Social Expenditure: Do Regimes Shape Public Opinion?. Social Indicators Research, 101(1), 323-340.

Jansson, B. S. (2019). The Reluctant Welfare State: Engaging History to Advance Social Work Practice in Contemporary Society (9th ed). Boston, Massachusetts: Cengage.

Jæger, M. (2006). What makes people support public responsibility for welfare provision: self-interest or political ideology? A longitudinal approach. Acta Sociologica, 49(3), 321-338.

Katznelson, I. (2014). Fear Itself: The New Deal and the Origins of Our Time. New York: Liveright.

Kim, T. K. (2020). US November Election and Social Policy Direction under the New Presidency. Global Social Security Review, 15(1), 103-112.

Kim, T. K. (2021). American Progressive Agenda and Social Policy: California’s Guaranteed Income and Debate on $3.5Trillion Mega bill. Global Social Security Review, 18(1), 65-78.

Kulin, J., & Svallfors, S. (2013). Class, values, and attitudes towards redistribution: A European comparison. European Sociological Review, 29(2), 155–167.

Kurowska, A., Eisele, O., & Kiess, J. M. (2019). Welfare Attitudes and Expressions of (Trans)national Solidarity. American Behavioral Scientist, 63(4), 492-505

Larsen, C. A. (2008). The Institutional Logic of Welfare Attitudes: How Welfare Regimes Influence Public Support. Comparative Political Studies, 41(2), 145-168.

Margalit, Y. (2013). Explaining social policy preferences: Evidence from the great recession. American Political Science Review, 107(1), 80–103.

Ng, I. Y. H. (2015). Welfare Attitudes of Singaporeans – Ambiguity in Shifting Socio-political Dynamics. Social Policy & Administration, 49(7), 946-965.

Owens, L. A., & Pedulla. D. S. (2014). Material welfare and changing political preferences: The case of support for redistributive social policies. Social Forces, 92(3), 1087–1113.

Prasad, M. (2016). American Exceptionalism and the Welfare State: The Revisionist Literature. Annual Review of Political Science, 19(1), 187-203.

Sachweh, P. (2018). Conditional Solidarity: Social Class, Experiences of the Economic Crisis, and Welfare Attitudes in Europe. Social Indicators Research, 139(1), 47-76.

Sachweh, P. (2019). Crisis experiences and welfare attitudes during the Great Recession: A comparative study on the UK, Germany and Sweden. Acta Sociologica, 62(2), 135-151.

Toikko, T., &Rantanen, T. (2020). Association Between Individualism and Welfare Attitudes: An Analysis of Citizens’ Attitudes Towards the State’s Welfare Responsibility. Journal of Social and Political Psychology, 8(1), 132–150

Turner F. J. (1996). The Frontier in American History. Mineola, New York: Dover.

Van Heuvelen, T. (2014). The Religious Context of Welfare Attitudes. Journal for the Scientific Study of Religion, 53(2), 268-295.

Wang, K. Y., Wong, C., & Tang, K. (2012). Citizens' attitudes towards economic insecurity and government after the 2007 financial tsunami: A Hong Kong and Taiwan comparison, International Journal of Social Welfare, 22(2), 152-163.

World Values Survey Association (2020) World Values Survey: Round Seven -USA-Pooled Datafile Version. Retrieved June 2022, from www.worldvaluessurvey.org/WVSDocumentationWV7.jsp.

Wong, T. K., Wan S. P., & Law, K. W. (2009). Welfare attitudes and social class: the case of Hong Kong in comparative perspective, International Journal of Social Welfare, 18(1), 142-152.

Appendix 1: Descriptive statistics of variables in each cohort

Variables	Groups	2006 Cohort Frequency (%)	2017 Cohort Frequency (%)
	Young	515 (41.2)	1,041 (40.1)
Age	Middle	557 (44.6)	1,104 (42.5)
	Old	177 (14.2)	452 (17.4)
Sex	Male	604 (48.3)	1,256 (48.4)
	Female	646 (51.7)	1,340 (51.6)
	White	870 (69.7)	1,671 (64.4)
Race	Black	142 (11.4)	307 (11.8)
	Hispanic	161 (12.9)	409 (17.7)
	Others	76 (  6.1)	209 (  8.0)
	High School or below	584 (46.8)	825 (32.2)
Education Level	Some College / BA	527 (42.2)	1,172 (45.7)
	Master or above	138 (11.0)	567 (22.1)
Marital Status	Married	669 (53.6)	1,488 (57.3)
	Unmarried	580 (46.4)	1,099 (42.3)
	Employed	738 (59.1)	1,533 (59.0)
Employment	Unemployed	46 (  3.7)	195 (  7.5)
Status	Retired	194 (15.5)	498 (19.5)
	Others	271 (21.7)	332 (13.0)
	Fair/Poor	228 (18.3)	725 (28.1)
Health Condition	Good	652 (52.2)	1,252 (48.6)
	Very good	368 (29.5)	601 (23.3)
	Upper/Upper Middle	315 (26.7)	797 (31.1)
Social Class	Middle	413 (35.0)	929 (36.3)
	Lower/Working	451 (38.3)	832 (32.5)
	Low	252 (22.0)	526 (20.8)
Income Level	Medium	818 (71.3)	1,793 (70.8)
	High	77 (  6.2)	212 (  8.4)
	Republican	356 (28.5)	885 (33.0)
Political Affiliation	Democrat	503 (40.3)	1,107 (42.7)
	Others	390 (31.2)	633 (24.4)
	Religious	862 (69.0)	1,507 (58.0)
Religiosity	Not religious	292 (23.4)	865 (33.3)
	Atheist	42 (  3.4)	206 (  7.9)
Total sample size (N)		1,249	2,596

Author´s Address:
Eun Kyung Lee, LMSW, Ph.D.
Adjunct Professor, Adelphi University School of Social Work
1 South Ave. Garden City, New York, 11530, USA
eklee@adelphi.edu

Author´s Address:
Tae Kuen Kim, Ph.D.
Associate Professor, Adelphi University School of Social Work
1 South Ave. Garden City, New York, 11530, USA
tkim@adelphi.edu