(p.249) Appendix A Measuring Concepts
(p.249) Appendix A Measuring Concepts
This appendix discusses the operationalization of variables in alphabetical order. The complete questionnaire used in the Berlin surveys is presented in Appendix C.
Authoritarianism. ‘Here is another card with statements about political problems, but also about your personal situation. Would you tell much how much you agree or disagree with each statement?’ MPs' expressed their agreement or disagreement with the following five statements using a seven-point indicator: (1) ‘To compromise with political adversaries is dangerous because it frequently leads to the betrayal of one's own side’; (2) ‘A country's general welfare should always override the special interests of groups and organizations’; (3) ‘It will always be necessary to have a few strong, able individuals who know how to take charge’; (4) ‘If we don't defend ourselves against rabble-rousers and troublemakers, our political order will be replaced by disorder and chaos’; (5) ‘Discipline and obedience are essential in educating children into responsible citizenship.’ High values represent authoritarian responses. The additive index ranges from 5 to 35.
Communist Party: A dichotomous variable coded ‘1’ if the MP is a member of the reformed communist party (PS) and ‘0’ if not.
Comparative Performance Evaluations (1995). ‘If you think of the time in the GDR before the fall of the Berlin Wall, do you think people in the former GDR are, all in all, better off now, worse off, or does it make no difference?’ MPs expressed their summary evaluation by means of a seven-point indicator. In addition to the overall evaluation, MPs also evaluated: (1) ‘And when you just think of the personal freedom of people in the former GDR?’; (2) ‘And when you think of the social conditions of people in the former GDR?’; (3) ‘And when you think of the economic conditions of people in the former GDR?’
Conservative Ideology. ‘In a number of studies, people have been asked to place themselves on a left-right scale indicating their overall political position. For purposes of comparison, would you please tell me where you would place (p.250) your views on this scale?’ MPs used a ten-point, left-right self-placement indicator to summarize their ideological views.
Contemporary Economic Problems. ‘What are the two most important problems in Berlin?’ For the most important problem, MPs received a score of ‘1’ if they mentioned the general economic situation, unemployment, the budget deficit, or economic development. MPs received a score of ‘0’ if they did not mention any of these problems. The same scoring procedure was applied to the second most important problems. I thus obtained two dummy variables. I then created an additive index, adding the two dummy variables and the unemployment category (#21) in Table A1 (see below).
Democratic Ideals: ‘The term democracy is frequently used without further specifications these days. What seem to you personally the essentials of a democracy?’ Based on detailed notes taken during the interview, I coded the responses into a maximum of five variables on the basis of a modified coding scheme developed by Putnam (1973). The categories presented in Table 5.3 consist of additive indicators summarizing more detailed code categories. The following description lists the general categories as used in Table 5.3 and their constituent categories (each is preceded by a letter). Government by the People: (a) government by the people, popular control, control by the people; (b) popular interest in and awareness of politics; (c) responsibility or answerability of the government to the people; government by consent; government based on electoral mandate. Most responses could actually be captured by two or three variables. That is, after coding MPs' responses into about three variables, the essence of most responses were captured. Each democratic component was coded only once so that repetitive answers do not influence the number of components coded. Thus, the ‘government by the people’ indicator has a range from 0 (a respondent did not mention any democratic component matching the three categories) to 3 (a respondent mentions components a, b, and c. Social Equality: (a) just standard of living; freedom from want; social and economic security for all; (b) classless society; less social distance; (c) fewer rich and poor; (d) social and economic security for all. Active Participation: (a) popular participation; an active role for the people, popular involvement in decision-making; direct democracy within the framework of parliamentary democracy. Direct democracy: (a) direct democracy; referenda for important decisions; public should be able to recall ministers at any time. Equality of opportunity: (a) equality in general; (b) political equality; one man, one vote; equality of opportunity; each person has the possibility of developing him/herself as far as possible; participation of citizens in all areas of society. Civil Right/!Limited Government: (a) liberty; freedom in general; (b) political or civic liberties in general; (c) freedom of expression (speech, free press); (d) minority rights; consideration of the minority; (e) limited government; checks and balances; no arbitrary power; (f) laissez (p.251) faire, socially and economically; freedom from government interference in socio-economic affairs; Institutions: (a) elections; (b) majority rule; (c) representative or parliamentary government in general; (d) parliamentary or legislative control over the executive; (e) rule of law; legal due process. Political Competition: (a) possibility of government changes; minority can become majority; (b) party competition; more than one party; (c) strong, critical opposition; (d) elite competition; ruling oligarchy; Societal Competition: (a) pluralism; variety of private associations and institutions; (b) consultation by the government with groups and organizations; citizens' responsibility: (a) Mature educated, intelligent citizens; (b) freedom to do what is right; individual self-control; (c) assumption of responsibility and duties; (d) action in the interest of collective, not only of individual; reciprocal respect and tolerance.
The percentages in Table 5.3 represent the proportion of respondents mentioning at least one constituent component for each category. For the multivariate analyses (Table B1), I used an additive index of all the social-egalitarian responses. This indicator has a range from 0 to 4. Given the skewed distribution of this variable particularly in the West, I also analysed this model using probit, but the results generally lead to the same substantive conclusions (see Rohrschneider 1994 for details).
In order to examine the stability of responses (Fig. 5.3), I recoded the general categories into ‘0’ (a respondent did not mention any component) and ‘1’ (respondents mentioned at least one component). Then, the 1992 and 1995 responses were cross-tabulated.
Democratic Rights: ‘In order to get a comparable picture of the distribution of opinion with earlier studies, I am now using a standardized format to simplify matters a little bit. I am aware that these statements capture your views only incompletely. Nevertheless, would you tell me for each of the statements whether you agree or disagree?’ MPs evaluated the statements shown in Table 5.1 by means of a seven-point indicator. In Table 5.1 the percentages are based on respondents who agree (7, 6, 5) with statements A and B and who disagree (1, 2, 3) with the other statements. For the multivariate analyses, each statement was coded so that high values represent a democratic response; the original metric was used. The index ranges from 5 (undemocratic views) to 35.
Economic Values. ‘What do you personally feel are the major advantages and disadvantages of establishing a social market economy in East Germany? [PROBE, IF NEEDED.] Some people argue that socialist economies, despite several limitations, offered certain advantages, such as social security, job security, or greater collegiality at the workplace. How would you evaluate this argument?’ I coded the responses to the open-ended economy question in up to six variables on the basis of the coding scheme presented in Table A1.
Table A1. Coding economic ideals
‘What do you personally feel are the major advantages and disadvantages of establishing a social market economy in East Germany? (probe, if needed). Some people argue that socialist economies, despite several limitations, offered certain advantages, such as social security, job security, or greater collegiality at the workplace. How would you evaluate this argument?’
Long-term disadvantages (coded here if characteristic is viewed as a serious disadvantage immanent to market economies i.e. these seem unfixable within systemic boundaries)
Positive aspects of GDR economy (coded here if GDR advantage is explicitly mentioned)
Miscellaneous Critique (coded here if market economies are principally accepted, but that fundamental reforms are needed)
MPs' views about welfare policies are measured by the following question: ‘On this card is a list with statements regarding the economic system. Would you please tell me how much you agree or disagree with each of the following statements?’ MPs then evaluated the statements shown in Table 8.1 by means of a seven-point indicator, ranging from agree completely (7) to disagree completely (1). The percentages in Table 8.1 represent MPs who agree (7, 6, 5) with a statement. For the multivariate analyses I constructed an additive indicator ranging from 7 (free market values) to 49 (welfare values).
Education: (1) Volkschule/8. Klasse; (2) Mittlere Reife/10. Klasse/Mittel Schule; (3) Abitur/EOS; (4) Fachhochschule/Fachschule; (5) Universität.
Future Personal Finances. ‘In a few years, do you think that you will be better off, about the same, or worse off?’
Future National Economy. An additive index of characteristic #17 and #18 (Table Al).
Gender: Coded ‘1’ for male MPs, and ‘0’ for female MPs.
Income: (1) Below DM 4,000; (2) DM 4,000–5,999; (3) DM 6,000–7,999; (4) DM 8,000–9,999; (5) above DM 10,000.
Institutional trust: ‘Here is a list with important political institutions in our society. Would you tell me how much trust you have in each of these institutions?’ MPs' evaluated the five institutions shown in Table 9.1 on a seven-point trust-distrust indicator. The responses were collapsed into those trusting (7, 6, 5), ambivalent (4), and distrusting (3, 2, 1) institutions. For the OLS analysis I created an additive index using the original metric. The indicator ranges from a low of 5 (expressing distrust) to 35 (expressing trust).
Party Membership. A dummy variable of MPs' party membership. In the East the excluded parties are the PDS and the also Bündnis ‘90/Greens; in the West the excluded party is the Bündnis '90/Greens. When these dummy variables are used (because the theoretical argument requires a refined breakdown of respondents’ party preference), I did not employ the PDS-party membership variable described earlier.
Plebiscitarian Views. ‘Would you please indicate your opinion on the plebiscitarian involvement of citizens by indicating whether you find each of these [show card] procedures involving citizens directly meaningful or not meaningful?’ MPs evaluated the statement presented in Table 5.5. I used a dichotomous response format in order to keep the results comparable to the study of national-level East and West German MPs (Herzog et al. 1990; (p.255) Werner 1991): For the plebiscitarian index used in multivariate analyses, I first included the small number of missing values as a neutral middle category, and then created an additive index, ranging from 3 (reject all procedures) to 15 (endorse all procedures).
Pluralism. ‘You mentioned earlier —— as the most important problem in Berlin. There are a number of conflicts of interests involved in solving this problem, since different groups and people have different conceptions of how to solve it. Do these conflicts of interests have a positive or negative effect on the governability of Berlin? [PROBE, IF NEEDED.] Why? [or] Why not?’ Table A2 presents the coding scheme used to code MPs' responses to this question. The following recodes were used to prepare the frequency distribution in Table 7.2. Positive Views: categories #11, 12, 13, 14, 15 in Table A2; Statist Aversion: #31, 32, 44; Socialist Aversion: #35, 45, 46.
The statements in Table 7.1 were evaluated by political elites by means of a seven-point indicator ranging from agree strongly (7) to disagree strongly (1).
The pluralism index employed in the multivariate analyses is composed of two indicators. First, I recoded indicator B in Table 7.1 into three categories (5–7 = 0; 4 = 1; 1–3 = 2). Secondly, the pluralism score in Table 7.3 is collapsed into three categories (0 = 0; 1 = 1; else = 2). This recoding entails that high scores reflect pluralist orientations. I then combined both indicators into an additive index ranging from 0 (anti-pluralist views) to 4 (pro-pluralist views).
Post-war Generation: A dichotomous variable where ‘1’ represents MPs born in 1945 or later, and ‘0’ represents MPs born in 1944 and earlier. I chose this cut-off point for generations because it is theoretically meaningful and because it leaves enough cases within each cohort to conduct the cohort analyses. The decision to divide MPs into two groups—those born in 1944 or before and those born later—means that MPs reaching the age of 15 before 1960 are coded as a pre-war cohort. This cut-off point parallels the degree to which the quasi-laboratory conditions had been established in Germany, because the Berlin Wall was not built until 1961. Before that time East and West Germany were less insulated from each other. I have conducted a sensitivity analysis using eleven adjacent cut-off points for the cohort definition. For example, I first defined a pre-war cohort by combining MPs born in 1938 and before into the pre-war category. A second pre-war variable includes those born in 1939 and before, a third contains MPs born in 1940 and before, etc. I then conducted the analyses presented in Table B1 (Appendix B), using these different cohort variables in separate analyses. The results parallel the historical reality in Germany: the largest cohort differences within the East emerge when the post-war cohort contains predominately MPs who grew up after the construction of the Berlin Wall. While a more refined breakdown of cohorts would undoubtedly be desirable, these analyses suggest that the (p.256)
Table A2. Coding pluralist values
You mentioned earlier (Most Important Problem) as an important problem in Berlin. There are a number of conflicts of interests involved in solving this problem since different groups and people have different conceptions of how to solve the problem. Do these conflicts of interests have a positive or negative effect on the governability of Berlin? (PROBE, IF NEEDED) Why? (OR) Why not? (CODE UP TO FOUR MENTIONS OF WHY OR WHY NOT PROBLEMS REDUCE GOVERNABILITY).
Positive effect of conflicts of interest on governability
Negative effect on governability
No effect on governability
Postmaterialism: I used a slightly modified coding procedure of the original items Inglehart developed. While materialists and postmaterialists are coded as Inglehart suggests, I distinguish between mixed materialists and mixed postmaterialists, depending on whether respondents first mention the materialist or the postmaterialist item. This coding procedure uses more information than the three-point indicator. (The results are almost identical when the mixed categories are combined.)
Political Tolerance. ‘I would like to ask you a few questions about social groups and movements which are viewed by some people as threatening to the political and social order in the united Germany. Would you please select from this list [show card] the group or organization that you like the least?’ After selecting the first group, I asked: ‘What is your second least-liked group?’ After selecting the groups: ‘I would like to know your personal opinion on the following statements regarding the activities of each group. Let us begin with [LEAST-LIKED GROUP]. Independent of the existing legal framework, would you tell me, how strongly you agree or disagree with the following statements?’ MPs used a seven-point scale to express their views about the political activities contained in Table 6.1. Entries in Table 6.1represent agreement (7, 6, 5) with indicators 1 and 3 and disagreement (1, 2, 3) with the other indicators. For the multivariate analyses, I first recoded each of the indicators so that high values represent a tolerant response. I then combined the four questions about each group into two separate indices. Each of the two tolerance indicators thus ranges from 4 (representing intolerance) to 28.
For the stability analyses (Fig. 6.4), I first recoded the two summary indicators into three categories: intolerant (4–13); ambivalent (14–18); tolerant (19–28). These recodes reflect the intuitively accessible original metric where 1–3 reflects rejection of a statement, 4 represents undecided views, and 5–7 reflects supportive views. Then the 1992 and 1995 responses were cross-tabulated. Reasons for Intolerance (based on code categories presented in Table A3). Percentages in Table 6.2 represent MPs who mention at least one category. Historical reasons, first group (the focus in Table 6.2 is on fascists): categories 11, 12, 13, 14, 15, 16, 17; Historical reasons, second group: categories 11, 12, 13, 14, 16, 18. (2) Defendable democracy (first and second groups): categories 21, 22, 23, 24, 25. Reasons for Tolerance. (1) Principle-based tolerance: categories 41 and 44; (2) Control-based tolerance: category 42; (3) Group is harmless: categories 43, 45, 46, 47.
Religiosity: ‘How often do you attend church? Would you say: At least every Sunday; almost every Sunday; sometimes; once a year; less than once a year; never.’ High values represent religious responses.
Table A3. Coding reasons for/against tolerance
Historical reasons for refusing to grant rights
Reasons for tolerance
(p.259) Right-wing Target Group. A dichotomous variable where ‘1’ represents the choice of right-wing groups (fascists, anti-abortionists, and displaced peoples' organizations) and ‘0’ any other group.
Satisfaction with democracy's performance: ‘On the whole, are you very satisfied, fairly satisfied, not very satisfied, or not satisfied at all with the way democracy works in Germany?’
Social Change. ‘Here is another card with statements about political problems, but also about your personal situation. Would you say how much you agree or disagree with each statement?’ MPs expressed their views on a seven-point indicator when evaluating the following statements measuring social change: (1) ‘Everything is changing so fast these days that it is difficult to find firm reference points’; (2) ‘Everything is so uncertain these days that one has to be ready for anything.’ An additive index was created ranging from 2 to 14.
Threat Perception. ‘Please evaluate the two groups you selected in terms of the following characteristics [SHOW CARD]’. Please select a 7 if you feel the characteristics mentioned on the left side is most accurate. A 1 means that you think the characteristics mentioned on the right side is most accurate. (Members of Group): (1) ‘are violent or nonviolent’; (2) ‘are democratic or non-democratic’; (3) ‘partially employ illegal methods or only employs legal methods.’ The responses were recoded to represent the same polarity and then combined into two additive indexes. High values represent high threat perceptions.
Age: Measured in years.
Democratic Ideals. ‘In your opinion, what is most important about democracy. Which things on this list are absolutely necessary for one to be able to say of a country, This is a democracy?’ Multiple responses were allowed. Percentages in Table 5.4 are based on the number of respondents which selected a statement. The number of cases in the Allensbach surveys is between N = 1,000 and N = 1,100.
Democratic Rights. ‘On this card, there are several statements about politics, state, and society. We would like to ask you to indicate your views about these statements.’ Respondents were shown the statements presented in Table 5.2. Respondents evaluated these statements by means of a six-point scale which runs from +3 to –3 without a neutral category. Entries in Table 5.2 are either those agreeing with a statement (1, 2, 3) or disagreeing with a statement (–1, –2, –3). For the summary index, each statement was coded so that high values represent a democratic response.
(p.260) Economic Values, The eastern and western German public were asked in the 1991 and 1994 Allbus surveys (foreign nationals are excluded from the Allbus-based analyses in every chapter): ‘Here is a list with different opinions about social status differences in Germany in terms of existing ones as well as how it ought to be. Please tell me for each statement whether you agree completely, agree somewhat, disagree somewhat, or disagree completely.’ Respondents were then shown the statements presented in Table 8.4. After including the small number of missing data as a neutral middle category, I summed the responses to create an overall welfare economic indicator for mass publics. This indicator is used in the analyses presented in Table 8.6. Opinion about the Market Economy: ‘Do you have a positive or negative opinion about the economic system in Germany.’ Response categories are: (1) A good opinion; (2) A bad opinion; (3) Undecided. Welfare versus Socialist Views: The International Social Justice Surveys asked respondents to respond to two statements (Tab. 8.7): ‘The government should guarantee everyone a minimum standard of living’ and ‘The government should place an upper limit on the amount of money any one person can make.’ The response categories are: (1) Agree completely; (2) Agree somewhat; (3) Neither agree nor disagree; (4) Disagree somewhat; (5) Disagree completely.
Education: Higher values represent a higher degree. The few respondents who still attend school, gave no answer, or who obtained other school degrees, are excluded.
Freedom versus Equality: ‘Two people are talking about what is ultimately more important, freedom or as much equality as possible. Would you please read this and tell me which of the two comes closest to saying what you also think?’
A. ‘I think that freedom and equality are equally important. But if I had to choose between the two, I would say personal freedom is more important; that is, for people to be able to live in freedom and not be restricted in their development.’
B. ‘Certainly both freedom and equality are equally important. But if I had to choose between the two, I would consider as much equality as possible to be more important; that is, for no one to be underprivileged and class differences not to be so strong.’
Gender: Male ‘1’ and female ‘0’.
Income: I used a finely graded scale ranging from ‘1’ (income is between DM 400 and 600) to 22 (respondents' income is more than DM 15,000). In the OLS analysis I replaced missing data for eastern respondents with the mean for the East (mean = 11) and missing data for western respondents with the mean for the West (mean = 3).
Institutional trust: ‘I am going to mention several institutions. Please tell me (p.261) how much you trust each of them?’ Citizens evaluated the constitutional court, the federal parliament, the local administration, the justice system, and the federal government by means of a seven-point scale which is identical to the one used by MPs. In Table 9.1 responses were recoded to reflect trust (7, 6, 5) ambivalence (4), and distrust (3, 2, 1). The combined indicator ranges from 5 (representing distrust) to 35.
Party preference: ‘If there were a federal election next Sunday, which party would you vote for?’ I created three dummy variables representing the CDU, FDP, and SPD supporters. The excluded reference groups are the voters who would support the Bündnis ‘90/Greens, PDS, Republikaner, non-voters, and voters of other parties. This variable is used in Chapter 9 because the theoretical argument requires a refined breakdown of respondents’ party preference.
PDS Party preference: ‘If there were a federal election next Sunday, which party would you vote for?’ I created a dummy variable where ‘1’ represents those who would support the PDS and ‘0’ for supporters of other parties CDU, FDP, SPD, Bündnis '90/Greens, and Republikaner. This variable was used in Chapter 8.
Pluralism: The statements for mass publics lack the neutral middle category, but the statements are identical to those used in the Berlin surveys.
Political Tolerance. Respondents in eastern and western Germany were asked in the Times/Mirror survey (1991) ‘Here are some statements on different topics. Please tell me how much you agree or disagree with each of these statements’: (1) ‘Freedom of speech should not be granted to fascists’; (2) ‘Books that contain ideas dangerous to society should be banned from public school libraries’; (3) ‘Homosexuals should not be permitted to teach school.’ Response categories are: (1) Strongly agree; (2) Agree; (3) Disagree; (4) Disagree strongly. Entries in Figure 6.3 are respondents who disagree (3, 4) with each statement. About undemocratic parties, respondents were asked: ‘Some people feel that in a democracy all political parties should be allowed, even those that don't believe in the democratic system. Others feel that even in a democracy certain political parties should be outlawed. Which comes closer to your view?’ Response categories are: (1) Allow all: (2) Outlaw some; (3) Can't say. Entries are respondents who would allow all parties. The common characteristic of these statements is that they ask respondents to consider civil liberties with reference to specific circumstances.
Respondents in the World Values survey 1995–7 were presented with a list of groups closely resembling the list in the Berlin parliamentarian surveys: ‘Now I want to know something about various groups in this society. On this card, there are various groups. Please select the groups which you like the least.’ Respondents evaluated the following statements: (1) Members of [GROUP] should not occupy a public office; (2) [GROUP] should not be allowed (p.262) to hold demonstrations; (3) Members of [GROUP] should not be allowed as teachers in schools. Response categories are dichotomous (allow or not allow). Percentages in Figures 6.3 represent respondents who would allow an activity. Although the mass survey includes the problematic category ‘criminals’ on the list of groups, this category was fortunately chosen by only about 10 per cent in the surveys in Germany. I did not exclude this group from the analyses, for lack of access to the original data set at the time of this writing.
Positive expectations about future personal finances: ‘How is your personal financial situation?’ (1) Very good; (2) Good; (3) Mixed; (4) Bad; (5) Very bad.
Positive Evaluations of the contemporary economy: ‘How would you evaluate the contemporary economic situation in Germany?’ (1) Very good; (2) Good; (3) Mixed; (4) Bad; (5) Very bad.
Positive expectations about the future economy: ‘How well do you think the economy will do in a year?’ (1) Much better than today; (2) A little better than today; (3) The same; (4) A little worse than today; (5) Much worse than today.
Postmaterialism: Inglehart's postmaterialism indicator with four categories.
Socialism as an Ideal: In the Allensbach surveys, upon which Figure 5.2 is based, respondents were asked to respond to the statement: ‘Socialism is a good idea that was poorly implemented.’ Response categories are ‘yes’ and ‘no’. In the Allbus 1994 survey, used in Figure 5.5, respondents were given the same statement, but then either agreed (completely or somewhat) or disagreed (completely or somewhat).