The horizontal axis is labeled “Birth year (S Ds)” and has markings ranging from negative 2 to 2 in increments of 1 unit. The vertical axis is labeled “Generalized trust (S Ds)” and has markings ranging from negative 0.1 to 0.15 in increments of 0.05 units. The graph shows two lines. The first line for “General population” starts from (negative 2, 0.04), slopes downward, and terminates at (2, negative 0.04). The second line for “Managers” starts from (negative 2, 0.11), slopes downward, and terminates at (2, negative 0.08). The two lines intersect at (0.5, negative 0.01). Note: All numerical data values are approximated.Birth year effects on generalized trust of managers compared to non-managers. Note(s): See Model 4 in Table 1. Source: Author’s own work
The horizontal axis is labeled “Birth year (S Ds)” and has markings ranging from negative 2 to 2 in increments of 1 unit. The vertical axis is labeled “Generalized trust (S Ds)” and has markings ranging from negative 0.1 to 0.15 in increments of 0.05 units. The graph shows two lines. The first line for “General population” starts from (negative 2, 0.04), slopes downward, and terminates at (2, negative 0.04). The second line for “Managers” starts from (negative 2, 0.11), slopes downward, and terminates at (2, negative 0.08). The two lines intersect at (0.5, negative 0.01). Note: All numerical data values are approximated.Birth year effects on generalized trust of managers compared to non-managers. Note(s): See Model 4 in Table 1. Source: Author’s own work
Measurement of variables and descriptive statistics
| Variable | Definition/description | Mean, standard deviation and no. of observations |
|---|---|---|
| Generalized trust (0–10) | This variable is measured using the standard single-item generalized trust scale (e.g. Bauer and Freitag, 2018; Delhey et al., 2011; Rosenberg, 1956; Uslaner, 2015). The item reads: Generally speaking, would you say that most people can be trusted, or that you can't be too careful in dealing with people? The answer scale is an 11-point Likert-type scale that ranges from 0 You can't be too careful to 10 Most people can be trusted | μ = 5.03 SD = 2.49 N = 448945 |
| Birth year | This variable is measured using the item that asks respondents in what year they were born, and I use this variable to operationalize individuals' generational or birth cohort. I use these latter two terms synonymously, even though the term generational cohort is often used more strictly to define birth cohorts that have experienced a similar (supranational) socioeconomic and historico-cultural context during their formative years | μ = 1962.5 SD = 18.1 N = 448,945 |
| Birth-year ventile | This variable is constructed using the item on birth year and involves dividing the individuals in our sample evenly over 20 cohorts, where the first cohort involves the 5% of individuals with the earliest birth year and so forth until the twentieth cohort. We use this variable as an alternative operationalization of individuals' generational cohort. In the analysis we use this variable as a continuous variable that happens to have a minimum score of 1 (the earliest cohort) and a maximum score of 20 (the latest cohort) | μ = 10.5 SD = 5.76 N = 448,945 |
| Birth-year fortile | This variable is constructed using the item on birth year and involves dividing the individuals in our sample evenly over 40 cohorts, where the first cohort involves the 2.5% of individuals with the earliest birth year and so forth until the fortieth cohort. I use this variable as an alternative operationalization of individuals' generational cohort. In the analysis I use this variable as a continuous variable that happens to have a minimum score of 1 (the earliest cohort) and a maximum score of 40 (the latest cohort) | μ = 20.5 SD = 11.5 N = 448,945 |
| Manager (1 = yes) | This variable is measured using the item that asks respondents whether they are/were responsible for supervising other employees: In your main job, do/did you have any responsibility for supervising the work of other employees? I identify individuals as managers (non-managers) if they respondent yes (no) to this item | μ = 29.0% SD = 45.3% N = 448,945 |
| Manager born after 1980 (1 = yes) | This dummy variable is measured by combining the item measuring managerial status and the item on birth year (see above). The variable has a value of 1 when an individual is a supervisor and was born after 1980 and a value of 0 otherwise. I have taken 1980 as the cutoff point because this birth year is one standard deviation above the average birth year in the sample. The percentages of managers born after 1980 and before 1981 add up to the total percentage of managers in the sample | μ = 3.94% SD = 19.5% N = 448,945 |
| Manager born before 1981 (1 = yes) | This dummy variable is measured by combining the item measuring managerial status and the item on birth year (see above). The variable has a value of 1 when an individual is a supervisor and was born before 1981 and a value of 0 otherwise. The percentages of managers born before 1981 and after 1980 add up to the total percentage of managers in the sample | μ = 25.0% SD = 43.43% N = 448,945 |
| Age | This variable is calculated as the difference between the year of the survey and respondents' reported year of birth (see above) | μ = 49.3 SD = 17.6 N = 448,945 |
| Female (1 = yes) | This variable is based on the item that codes respondents' sex as male or female | μ = 52.4% SD = 49.9% N = 448,945 |
| Years of education | This variable is based on the item that asks respondents about how many years of full-time equivalent education they have completed? | μ = 12.8 SD = 4.05 N = 448,945 |
| Educational degree | This variable is based on ESS coding of the highest level of education that respondents state they have. Coding is based on an ESS-adapted version of the standard ISCED classification 0 Not possible to harmonize into ES-ISCED; 1 ES-ISCED I, less than lower secondary; 2 ES-ISCED II, lower secondary; 3 ES-ISCED IIIb, lower tier upper secondary; 4 ES-ISCED IIIa, upper tier upper secondary; 5 ES-ISCED IV, advanced vocational, sub-degree; 6 ES-ISCED V1, lower tertiary education, BA level; 7 ES-ISCED V2, higher tertiary education, ≥ MA level 55 Other | Categorical |
| Industry | Based on ESS coding of answers to the item that asks respondents to describe their employer: What does/did the firm/organization you work/worked for mainly make or do? Industry is recorded using two-digit NACE (nomenclature statistique des activités économiques dans la Communauté européenne) codes, which are equal to two-digit SIC (Standard Industrial Classification) codes. Earlier rounds of the ESS used revision 1 and 1.1 of the NACE (which are the same) and later rounds used revision 2, which does not map onto the previous revisions. The industry fixed effects that I include take this change in measurement/definition into account | Categorical |
| Occupation | Based on ESS coding of answers to the item that asks respondents to describe their job: What is/was the name or title of your main job? In your main job, what kind of work do/did you do most of the time? What training or qualifications are/were needed for the job? Occupation is recorded using four-digit ISCO (International Standard Classification of Occupations) codes. Earlier rounds of the ESS used the 1988 version of the ISCO classification system and later rounds used the 2008 version. The occupation fixed effects that I include take this change in measurement/definition into account | Categorical |
| Trust in police (0–10) | This variable is measured using the item that asks respondents to rate their trust in a specific institution, namely the police. The answer scale is an 11-point Likert-type scale that ranges from 0 No trust at all to 10 Complete trust | μ = 6.03 SD = 2.61 N = 444,621 |
| Trust in legal system (0–10) | This variable is measured using the item that asks respondents to rate their trust in a specific institution, namely the legal system. The answer scale is an 11-point Likert-type scale that ranges from 0 No trust at all to 10 Complete trust | μ = 5.16 SD = 2.73 N = 439,743 |
| Trust in politicians (0–10) | This variable is measured using the item that asks respondents to rate their trust in a specific institution, namely politicians. The answer scale is an 11-point Likert-type scale that ranges from 0 No trust at all to 10 Complete trust | μ = 3.53 SD = 2.42 N = 441,948 |
| Religiosity | This variable is measured using the item that asks respondents how religious they are: Regardless of whether you belong to a particular religion, how religious would you say you are? The answer scale is an 11-point Likert-type scale that ranges from 0 Not at all religious to 10 Very religious | μ = 4.60 SD = 3.03 N = 445,342 |
| Religious denomination | This variable is measured using the item that asks respondents about the religion or denomination that they belong to at present, if any. Respondents who indicated that they do not belong to a religion of denomination are added as a separate category that we may refer to as “atheist.” Answers are thus coded as follows 0 Atheist; 1 Roman Catholic; 2 Protestant; 3 Eastern Orthodox; 4 Other Christian denomination; 5 Jewish; 6 Islamic; 7 Eastern religions; 8 Other non-Christian religions | Categorical |
| Employment relation | This variable is measured using the item that asks respondents about the type of employment relation that they are in. Possible answers for the item are: 1 Employee; 2 Self-employed; 3 Working for own family business | Categorical |
| Labor market status | This variable is measured using the item that asks respondents about their main activity during the last seven days. Answers are coded as follows 1 Paid work; 2 Education; 3 Unemployed, looking for job; 4 Unemployed, not looking for job; 5 Permanently sick or disabled; 6 Retired; 7 Community or military service; 8 Housework, looking after children, others; 9 Other | Categorical |
| Field/subject of study | This variable is measured using the item that asks respondents in which field or subject they have their highest educational qualification. This item has been included in Rounds 2–4 of the ESS (2004–2008). Possible answers for the item are 1 General/no specific field; 2 Art, fine/applied; 3 Humanities; 4 Technical and engineering; 5 Agriculture/forestry; 6 Teacher training/education; 7 Science/mathematics/computing etc.; 8 Medical/health services/nursing etc.; 9 Economics/commerce/business administration; 10 Social studies/administration/media/culture; 11 Law and legal services; 12 Personal care services; 13 Public order and safety; 14 Transport and telecommunications | Categorical |
Note(s): Data are from the European Social Survey (ESS). See https://www.europeansocialsurvey.org for further information including codebooks
Generalized trust and its nomological network
| Generalized trust | Perceived helpfulness | Perceived fairness | Trust in police | Trust in legal system | |
|---|---|---|---|---|---|
| Perceived helpfulness | 0.516 | 1 | |||
| Perceived fairness | 0.591 | 0.537 | 1 | ||
| Trust in police | 0.293 | 0.291 | 0.303 | 1 | |
| Trust in legal system | 0.339 | 0.306 | 0.315 | 0.664 | 1 |
| Trust in politicians | 0.344 | 0.316 | 0.316 | 0.491 | 0.591 |
Note(s): Table presents individual-level correlations pertaining to samples comprising between 490,000 and 510,000 individuals (see Table A1). Table A6 presents the correlations between generalized trust and the chief independent variables in the analysis. Perceived helpfulness (0–10) is measured using the item that asks respondents to rate the helpfulness of people: Would you say that most of the time people try to be helpful or that they are mostly looking out for themselves? Because this item appears to capture something related to generalized trust, I expect it to correlate positively with individuals' generalized trust. Perceived fairness (0–10) is measured using the item that asks respondents to rate the fairness of people: Do you think that most people would try to take advantage of you if they got the chance, or would they try to be fair? As before, this item appears to capture something related to generalized trust, and I therefore expect it to correlate positively with individuals' generalized trust
Testing for a generational trend in (managers') generalized trust with various control variables added
| Dependent = generalized trust | Using industry and occupation to control for supervised activities | Controlling for other trending factors | Controlling for employment relation and labor market status | Controlling for field/subject of study | Country-age and country-year intercepts |
|---|---|---|---|---|---|
| Model A1 | Model A2 | Model A3 | Model A4 | Model A5 | |
| Birth year | −0.016 (0.003) [p = 0.000] | −0.005 (0.002) [p = 0.021] | −0.030 (0.003) [p = 0.000] | −0.010 (0.004) [p = 0.012] | −0.021 (0.002) [p = 0.000] |
| Birth year * Manager (0/1) | −0.019 (0.003) [p = 0.000] | −0.022 (0.003) [p = 0.000] | −0.029 (0.003) [p = 0.000] | −0.030 (0.006) [p = 0.000] | −0.020 (0.003) [p = 0.000] |
| Manager (1 = yes) | 0.001 (0.004) [p = 0.835] | 0.013 (0.003) [p = 0.000] | 0.018 (0.003) [p = 0.000] | 0.006 (0.006) [p = 0.321] | 0.015 (0.003) [p = 0.000] |
| Industry fixed effects | Yes | No | No | No | No |
| Occupation fixed effects | Yes | No | No | No | No |
| Trust in police | – | 0.072 (0.002) [p = 0.000] | – | – | – |
| Trust in legal system | – | 0.113 (0.002) [p = 0.000] | – | – | – |
| Trust in politicians | – | 0.167 (0.002) [p = 0.000] | – | – | – |
| Religiosity | – | 0.016 (0.002) [p = 0.000] | – | – | – |
| Religious denomination | No | Yes | No | No | No |
| Employment relation | No | No | Yes | No | No |
| Labor market status | No | No | Yes | No | No |
| Field/subject of study | No | No | No | Yes | No |
| Country-age and country-year intercepts | No | No | No | No | Yes |
| Baseline controls | Yes | Yes | Yes | Yes | Yes |
| Obs. | 428,554 | 412,659 | 420,839 | 114,685 | 448,945 |
| -2Loglikelihood | 1102020.8 | 1027261.0 | 1084086.6 | 292951.8 | 1155098.8 |
Note(s): Table reports results for multilevel models with individuals cross-classified by age and period. Standard errors are in parentheses and p-values are in square brackets. Baseline control variables are sex, years of education, education level and country fixed effects (see Table 1). The dependent variable and continuous independent variable are standardized (μ = 0; SD = 1). To save space, the table reports a selection of coefficients, but complete results as well as a replication package are available on request
Generational trends in (managers') generalized trust estimated with outliers removed and for different measures of generational cohort.
| Dependent = generalized trust | Outliers removed (age and birth year observations below the 1st and above the 99th percentile excluded) | Ventile cohorts (generational cohort numbered 1–20) | Fortile cohorts (generational cohort numbered 1–40) | |||
|---|---|---|---|---|---|---|
| Model A6 | Model A7 | Model A8 | Model A9 | Model A10 | Model A11 | |
| Birth year | −0.029 (0.002) [p = 0.000] | −0.022 (0.002) [p = 0.000] | – | – | – | – |
| Birth year * Manager (0/1) | – | −0.024 (0.003) [p = 0.000] | – | – | – | – |
| Birth-year ventile | – | – | −0.028 (0.003) [p = 0.000] | −0.021 (0.003) [p = 0.000] | – | – |
| Birth-year ventile * Manager | – | – | – | −0.024 (0.003) [p = 0.000] | – | – |
| Birth-year fortile | – | – | – | – | −0.028 (0.003) [p = 0.000] | −0.021 (0.003) [p = 0.000] |
| Birth-year fortile * Manager | – | – | – | – | – | −0.024 (0.003) [p = 0.000] |
| Manager (1 = yes) | 0.016 (0.003) [p = 0.000] | 0.015 (0.003) [p = 0.000] | 0.015 (0.003) [p = 0.000] | 0.013 (0.003) [p = 0.000] | 0.015 (0.003) [p = 0.000] | 0.013 (0.003) [p = 0.000] |
| Baseline controls | Yes | Yes | Yes | Yes | Yes | Yes |
| Obs. | 4,31,423 | 4,31,423 | 4,48,945 | 4,48,945 | 4,48,945 | 4,48,945 |
| -2Loglikelihood | 1111523.9 | 1111466.5 | 1157878.3 | 1157819.3 | 1157875.5 | 1157816.9 |
Note(s): Table reports results for multilevel models with individuals cross-classified by age and period. Standard errors are in parentheses and p-values are in square brackets. Table A1 provides details on the two measures of individuals' generational cohort. Baseline control variables are sex, years of education, education level and country fixed effects (see Table 1). The dependent variable and continuous independent variable are standardized (μ = 0; SD = 1). To save space, the table reports a selection of coefficients but complete results as well as a replication package are available on request
Comparison of older and younger managerial cohorts and the general population
| Dependent = generalized trust | Model A12 |
|---|---|
| Manager born after 1980 (1 = yes) | −0.063 (0.008) [p = 0.000] |
| Manager born before 1981 (1 = yes) | 0.028 (0.004) [p = 0.000] |
| Reference category: Non-manager (neither a manager born before 1981 nor a manager born after 1980) | 0 |
| Birth year | −0.033 (0.003) [p = 0.000] |
| Reference category: Born after 1980 (1 = yes) | 0 |
| Born before 1981 (1 = yes) | −0.034 (0.006) [p = 0.000] |
| Baseline controls | Yes |
| Obs | 4,48,945 |
| -2Loglikelihood | 1157743.2 |
Note(s): Table reports results for multilevel models with individuals cross-classified by age and period. Standard errors are in parentheses and p-values are in square brackets. Baseline control variables are sex, years of education, education level and country fixed effects (see Table 1). The two managerial cohort dummies are constructed so that there is one younger cohort comprising the managers whose birth year is at least one standard deviation above the average birth year in the sample (see Table A1 for details). The dependent variable and continuous independent variable are standardized (μ = 0; SD = 1). To save space, the table reports a selection of coefficients but complete results as well as a replication package are available on request
Correlations between generalized trust and chief independent variables
| Generalized trust | Birth year | Age | Year | Manager (1 = yes) | Female (1 = yes) | |
|---|---|---|---|---|---|---|
| Birth year | 0.024 | 1 | ||||
| Age | −0.017 | −0.950 | 1 | |||
| Year | 0.023 | 0.244 | 0.071 | 1 | ||
| Manager (1 = yes) | 0.073 | −0.062 | 0.063 | −0.001 | 1 | |
| Female (1 = yes) | −0.028 | −0.036 | 0.035 | −0.005 | −0.153 | 1 |
| Years of education | 0.188 | 0.260 | −0.222 | 0.139 | 0.185 | −0.031 |
Note(s): Table presents individual-level correlations pertaining to samples comprising between 453,000 and 510,000 individuals (see Table A1; see also Table A2)