Objectives
This first panel survey to be conducted in the context of the German Life History Study focused on the implications of institutional change during the process of societal transformation for the lives of East German men and women born in 1929-31, 1939-41, 1951-53, and 1959-61. The primary objective was to describe and explain change in the social inequality gap during this process of transformation. Which social positions disappeared, which were new? How did the allocation criteria and mechanisms for matching persons to these jobs change? Because reunification triggered dramatic changes on the East German labor market, the project focused on exploring who still had good labor market opportunities after reunification and who was likely to encounter difficulties. Issues of particular interest were, first, the impact of sector-specific job cuts and growth on labor market opportunities; second, the marketability of credentials attained in the GDR (in terms of both formal qualifications and experience); third, the labor market opportunities of different groups (in particular, women, the "older" age group of 40- to 50-year-olds, former members of the Socialist Unity Party and of the GDR elites, and people in specific occupations).

Questionnaire Design
Based on the instrument implemented in the "Life Courses and Historical Change in the German Democratic Republic" study (LV-DDR), a standardized, computer-assisted interview procedure that could be conducted either by telephone or in face-to-face interviews (combined CATI/CAPI strategy) was developed. In a pretest, the comprehensibility and practicability of the instrument as well as its acceptance among respondents was tested. The instrument focused particularly on the respondents' occupational history since 1989. Members of the 1929-31 cohort were questioned about their current employment status; if already retired, they were asked about their last job and how they now spent their days. Members of the 1939-41, 1951-53, and 1959-61 cohorts reported on their career history including (initial and continuing) vocational and professional training. Data on changes in the residential situation since 1989, current membership of organizations, social contacts and friendships before and after 1989, and changes in marital status or partnerships since 1989 were also collected (see Codebuch, pdf format, 355 KB). Participants in the non-response study were additionally asked about their family of origin and siblings, all employment spells up to 1989 (occupational history module) and all spouses/partners up to 1989 (partner module). Moreover, they were asked to give details of their highest educational level and highest vocational or professional training qualification up to 1989 and their membership of political parties up to 1989.

Sample Design
The addresses of the 2,331 participants in the LV-DDR study constituted the starting point for this follow-up survey. Owing to panel mortality, the updated data bank contained a gross sample of 2,159 addresses. Before and during the follow-up survey, great efforts were made to minimize neutral non-response. In preparation for the interviews, target persons were informed in detail about the topic and methods of the study, and notified that they would receive a lottery ticket for participating in the survey. In November 1996, the incentive for a complete interview was increased to a cash payment of 50 DM.

All LV-DDR target persons whose contact protocols indicated one of the following statuses were selected for the non-response study: "not at home," "could not be contacted," "absent for extended period," "indisposed for health reasons," "refused due to lack of time," "refused to give information on principle," "household member pretended that target person was not at home," "interview prevented by third party," or "target person refused to give interview." Note that only those refusals whose contact protocols indicated that it would be reasonable to contact them again were approached as candidates for the non-response study (see Methodenbericht, p. 26, pdf format, 169 KB).

Data Collection
The LV-Ost Panel study ran from March 1996 to June 1997. A total of 1,407 persons provided usable interviews, with approx. 330 interviews per cohort and virtually equal proportions of men and women. The coverage rate was 74% (see Table 1).

Table 1: Coverage and Reasons for Non-Participation in the LV-Ost Panel Study

	Birth Cohort			Total
	1929-31	1939-41	1949-51	N	%
Gross sample	548	540	534	2,159	100.0
Neutral non-response	79	47	63	261	12.1
Adjusted sample	469	493	471	1,898	100.0
No contact made to household or target person	21	23	22	92	4.8
Ill	27	9	4	40	2.1
Refusals	89	85	89	328	17.3
Realized interviews	332	376	356	1,438	75.8
Unusable interviews	3	10	13	31	1.6
Unable interviews/ coverage rate	329	366	343	1,407	74.1

It was thus possible to "update" the life histories of about 60 percent of the respondents surveyed in 1991/92 in 1996/97. Interviews were conducted by telephone (CATI) or – if the target persons did not have a phone or specifically requested a face-to-face interview – in person (CAPI). For the 1929-31 cohort, interviews took approx. 50 minutes with the CATI method and approx. 60 minutes with the CAPI method; in the other cohorts, they took approx. 65 minutes with the CATI method and approx. 75 minutes with the CAPI method (see Methodenbericht, p.26, pdf format, 169 KB).

In total, the addresses of 1,246 of the 2,131 systematic non-respondents in the LV-DDR study were included in the non-response study (48% of the addresses in the gross sample). A total of 188 interviews were ultimately realized, corresponding to a coverage rate of 19.5% (see Table 2).

Table 2: Coverage Rate and Reasons for Non-Participation in the LV-Ost Non-Response Study

	N	%
Gross sample	1,246	100.0
Neutral non-response	282	22.6
Adjusted sample	964	100.0
No contact made to household or target person	61	6.3
Ill	29	3
Refusals	670	69.5
Realized interviews	204	21.2
Unusable interviews	16	1.7
Unable interviews/ coverage rate	188	19.5

The high proportion of refusals in the non-response study shows that the success of our attempts to contact neutral non-respondents and encourage them to participate in a new study was very limited.

Representativeness
To test the representativeness of the panel study, differences between those who did and did not participate in the panel study were analyzed, and selected marginal distributions of the panel study compared with those of the 1995 microcensus (see Wehner 2003, pdf format, 135 KB). Apart from an education effect frequently observed in life history research and higher regional mobility among young single men, in particular, few differences were discerned between those who did and did not participate in the panel study or in the comparison with the 1995 microcensus data. When analyzing the panel survey data, however, it is important to bear in mind that (single) men and skilled workers are underrepresented in the 1950 and 1960 birth cohorts. The higher proportion of white-collar workers in the panel study reflects the fact that people with higher educational levels are generally more willing to participate in surveys of this kind. However, it may also reflect individual difficulties in classifying the respondents' occupational positions. This is a recurrent problem in surveys on the former East Germany, where the distinction between blue- and white-collar workers was less clear-cut, and where those in occupations requiring formal training generally had the job title "Facharbeiter für ..." ("skilled worker for …), irrespective of the occupational position associated with this activity.

Data Editing
Once the data had been collected, they underwent thorough casewise editing. The fact that the data collected on the 1989-1991/92 period in the LV-DDR study and the LV-Ost panel study were not always identical, but had to be combined to create a consistent structure, posed particular difficulties for data editing. Data checking and correction lasted from spring 1997 to summer 1999 and focused on the domains of occupational history, (initial and continuing) vocational and professional training, and second jobs, as well as partnership, children, and membership of organizations (see Editionsregeln (Langfassung), pdf format, 81 KB). Respondents were re-contacted to clarify any ambiguities and contradictions that remained after the second and third rounds of data editing (see Nachrecherchebericht, pdf format, 92 KB). Once the data had been cleaned and entered, the (semi-)open responses were coded (see Bericht über die Bericht über die Codierung, pdf format, 34 KB) and a comprehensive data checking process began, with a particular focus on filtering procedures and data domains, and on re-checking the validity of temporal overlaps between episodes in different domains (see Bericht über die Prüfung der Daten, pdf format, 196 KB).

The data from the non-response study also underwent thorough casewise editing. This process, which lasted from spring 1997 to summer 1999, essentially followed the editing rules of the LV-Ost panel study, with the exception that the responses did not have to be aligned with data from the LV-DDR study (see Editionsregeln, pdf format, 139 KB). Between June and July 1999, participants in the non-response study were re-contacted to clarify any ambiguities and contradictions that remained after data editing, and to gather information on their schooling and vocational or professional training prior to December 1989 (see Nachrecherchebericht, pdf format, 76 KB). During and after this data editing and checking stage, responses to the open questions were coded. As far as responses to the semi-open questions, in particular, could be coded using the available category system, this was also done at this point (see Bericht über die Codierung, pdf format, 98 KB). Once the data had been cleaned and entered, a comprehensive data checking process began. This was conducted by Ralf Künster with the support of two student research assistants (see Bericht über die Prüfung der Daten, pdf format, 124 KB).

Notes on Data Analysis
The public-use file for the "East German Life Courses After Unification" study (LV-Ost Panel) available from the Central Archive for Empirical Social Research (ZA) in Cologne comprises individual SPSS data files for each domain surveyed. For reasons of data protection, the public-use files were factually anonymized in the same way as for the predecessor study, the LV-DDR study. Any direct references to places and all open-ended responses were removed. The original "questionnaire number" was replaced by a new ID number produced by a random generator. The ordering of cases was also changed, such that no direct links can be made between the public-use files and the questionnaires themselves.

The data from the non-response study can be accessed by scholars with guest scientist status or through other special arrangements with the project team.

Documentation (Downloads, in German)