Social Stratification in Eastern Europe After 1989:
General Population Survey

Codebook, Part III

CONSTRUCTED VARIABLES

In order to make the data easier to work with, we have constructed a number of variables. These are shown below, with explanations as necessary. In all cases, the original variables used to construct these variables have been retained. Thus, users who are dissatisfied with the decisions we have made in effecting the conversions are free to construct their own variables. To review a list of the constructed variables, click here.

Sub-sample identifiers
Although for purposes of sampling, we divided the 1988 nomenklatura elite into eight categories, we neglected to include an identifying variable in the questionnaire. Thus, we add it here. For convenience, we also distinguish the three subcategories of the current elite.

GROUP Elite subcategory

STRATA In Poland and in Russia, still more detailed subdivisions of the sample were made. This variable was not used in the remaining countries.

Poland


Russia

Size of place of interview

Note: RESSIZE1 and RESSIZE2 are not currently in the data set.

RESSIZE1

RESSIZE2

Children

TOTCHILD

Highest level of education attained, except for respondent (expanded CASMIN categories)

We have converted all education responses (for respondents, children, parents, spouse, and spouse's father--but not grandfathers, since the response categories are different) into an expanded version of the international education classification developed by Mller, et al. [Müller, Walter, Paul Lüttinger, Wolfgang König, and Wolfgang Karle. 1990. "Class and Education in Industrial Nations." Pp. 61-91 in Class Structure in Europe, edited by Max Haller. Armonk, NY: Sharpe.] for the CASMIN project. The conversion maps are shown in Appendix B and the SPSS command files effecting the conversions are shown in Appendix D. Because a different algorithm was used to create CASMIN codes for respondents, they are shown separately below.


The expanded CASMIN codes are:

IDADEDUC       CASMIN code for respondent's father
IMOMEDUC     CASMIN code for respondent's mother
ISPEDUC          CASMIN code for respondent's spouse
ISDADEDU       CASMIN code for respondent's spouse's father

Years of school completed, except for respondent

We have also converted all education responses to "virtual" years of school completed. We have not counted the actual number of years spent in school (for which we lack information for all but respondents) but rather have recoded the country-specific responses to the education variables to the typical number of years of schooliing implied by the category (e.g., in the U.S. completed university education would be taken as equivalent to 16 years of schooling even though it takes some individuals five or six years to complete university). The conversion maps are shown in Appendix B and the SPSS command files effecting the conversions are shown in Appendix D. Again, because a different algorithm was used to create a "years of school completed" variable for respondents, the variable is shown separately, below.

YDADEDUC       Years of schooling: respondent's father
YMOMEDUC     Years of schooling: respondent's mother
YSPEDUC           Years of schooling: respondent's spouse
YSDADEDU        Years of schooling: respondent's spouse's father

Recoded education variables for respondent

HIEDRC Expanded CASMIN code for respondent

HIEDYRS Years of schooling of respondent

Occupation recodes

All occupations were coded initially into an expanded version of the 1988 edition of the International Standard Classification of Occupations (ILO, 1989), referred to in this codebook as ISCO88. The expanded ISCO88 codes are shown in Appendix C. All ISCO codes were then recoded to four standard international occupational classifications: the International SocioEconomic Index of occupations [ISE] (Ganzeboom, DeGraaf, and Treiman, 1991); the Erikson, Goldthorpe, and Portocarero occupational class categories [EGP] (Erikson and Goldthorpe, 1992:35-47, as modified by Ganzeboom; a six category of the EGP [SIX] (Ganzeboom, Luijkx, and Treiman, 1989); and the Standard International Occupational Prestige Scale [PRE] (Treiman, 1977: Appendix A). The maps converting ISCO88 into ISE, EGP, SIX, and PRE are shown in Appendix C. For all four sets of variables, the following non-response categories were used:

ISE01               ISEI score for the occupation on line 1 of the Activity History
to
ISE28               ISEI score for the occupation on line 28 of the Activity History

DADISE14      ISEI score for father's occupation when respondent age 14
DADISE48      ISEI score for father's occupation in 1948
DADISE52      ISEI score for father's occupation in 1952
DADISE63      ISEI score for father's occupation in 1963
DADISE73      ISEI score for father's occupation in 1973
PGISE             ISEI score for paternal grandfather's main occupation
MOMISE14    ISEI score for mother's occupation when respondent age 14
MGISE            ISEI score for maternal grandfather's main occupation
SPISE              ISEI score for spouse's current occupation
SDADISE        ISEI score for spouse's father's occupation when spouse age 14
EGP01             EGP code for the occupation on line 1 of the Activity History
to
EGP28             EGP code for the occupation on line 28 of the Activity History
DADEGP14     EGP code for father's occupation when respondent age 14
DADEGP48     EGP code for father's occupation in 1948
DADEGP52     EGP code for father's occupation in 1952
DADEGP63     EGP code for father's occupation in 1963
DADEGP73     EGP code for father's occupation in 1973
PGEGP            EGP code for father's father's main occupation
MOMEGP14   EGP code for mother's occupation when respondent age 14
MGEGP           EGP code for mother's father's main occupation
SPEGP            EGP code for spouse's current occupation
SDADEGP      EGP code for spouse's father's occupation when spouse age 14
SIX01              Six category EGP code for the occupation on line 1 of the Activity History
to
SIX28              Six category EGP code for the occupation on line 28 of the Activity History
DADSIX14      Six category EGP code for father's occupation when respondent age 14
DADSIX48      Six category EGP code for father's occupation in 1948
DADSIX52      Six category EGP code for father's occupation in 1952
DADSIX63      Six category EGP code for father's occupation in 1963
DADSIX73      Six category EGP code for father's occupation in 1973
PGSIX             Six category EGP code for father's father's main occupation
MOMSIX14    Six category EGP code for mother's occupation when respondent age 14
MGSIX            Six category EGP code for mother's father's main occupation
SPSIX              Six category EGP code for spouse's current occupation
SDADSIX        Six category EGP code for spouse's father's occupation when spouse age 14
PRE01             Prestige score for the occupation on line 1 of the Activity History
to
PRE28             Prestige score for the occupation on line 28 of the Activity History
DADPRE14     Prestige score for father's occupation when respondent age 14
DADPRE48     Prestige score for father's occupation in 1948
DADPRE52     Prestige score for father's occupation in 1952
DADPRE63     Prestige score for father's occupation in 1963
DADPRE73     Prestige score for father's occupation in 1973
PGPRE            Prestige score for father's father's main occupation
MOMPRE14   Prestige score for mother's occupation when respondent age 14
MGPRE           Prestige score for mother's father's main occupation
SPPRE             Prestige score for spouse's current occupation
SDADPRE       Prestige score for spouse's father's occupation when spouse age 14

Respondent's recent activity history

For respondents we created four additional sets of variables from the entries in the activity history: (1) those associated with the job held by the respondent at the time of the interview [suffix c]; (2) those associated with the final job held by the respondent--the current job for those employed at the time of the interview and the most recent job for those not employed at the time of the interview [suffix f]; (3) those associated with the job held in January 1993 [suffix 93]; and those associated with the job held in January 1988 [suffix 88]. We also created activity variables at the time of the interview, in January 1993, and in January 1988. Finally, we created several variables concerned with tenure: the date (month and year) that the final job began, the date (month and year) that the final job ended, the tenure (number of months) in the activity, and the total number of months of employment since first leaving school.

For all variables in this section the following non-response codes were used, copied from the original variables from which these variables were constructed:

Activity: for codes see Appendix A

ACTC        Current activity
ACT93       Activity in January 1993
ACT88       Activity in January 1988

Occupation variables (OCCC to PRE88): for codes see Appendix C

OCCC       ISCO code for the current occupation [pertains only to those currently employed]
OCCF        ISCO code for the most recent occupation [includes last previous occupation for those not currently                         employed]
OCC93      ISCO code for the occupation held in January 1993
OCC88      ISCO code for the occupation held in January 1988
ISEC          ISEI score for the current occupation [pertains only to those currently employed]
ISEF          ISEI score for the most recent occupation [includes last previous occupation for those not currently employed]
ISE93        ISEI score for the occupation held in January 1993
ISE88        ISEI score for the occupation held in January 1988
EGPC       EGP code for the current occupation [pertains only to those currently employed]
EGPF        EGP code for the most recent occupation [includes last previous occupation for those not currently employed]
EGP93      EGP code for the occupation held in January 1993
EGP88      EGP code for the occupation held in January 1988
SIXC        Six category EGP code for the current occupation [pertains only to those currently employed]
SIXF         Six category EGP code for the most recent occupation [includes last previous occupation for those not                        currently employed]
SIX93       Six category EGP code for the occupation held in January 1993
SIX88       Six category EGP code for the occupation held in January 1988
PREC        Prestige score for the current occupation [pertains only to those currently employed]
PREF        Prestige score for the most recent occupation [includes last previous occupation for those not currently                        employed]
PRE93      Prestige score for the occupation held in January 1993
PRE88      Prestige score for the occupation held in January 1988

Industry and Organization: for codes see Appendix A

INDC        Industry code for the current occupation [pertains only to those currently employed]
INDF         Industry code for the most recent occupation [includes last previous occupation for those not currently                        employed]
IND93       Industry code for the occupation held in January 1993
IND88       Industry code for the occupation held in January 1988
ORGC       Organization type code for the current occupation [pertains only to those currently employed]
ORGF        Organization type code for the most recent occupation [includes last previous occupation for those not                         currently employed]
ORG93      Organization type code for the occupation held in January 1993
ORG88      Organization type code for the occupation held in January 1988

Codes for Self-employment

SEC       Self-employed? Current occupation [pertains only to those currently employed]
SEF       Self-employed? Most recent occupation [includes last previous occupation for those not currently employed]
SE93      Self-employed? Occupation held in January 1993
SE88      Self-employed? Occupation held in January 1988

Codes for number supervised

SUBC        Number supervised in the current occupation [pertains only to those currently employed]
SUBF        Number supervised in the most recent occupation [includes last previous occupation for those not currently                   employed]
SUB93       Number supervised in the occupation held in January 1993
SUB88       Number supervised in the occupation held in January 1988

Codes for full-time vs. part-time work

FULLC         Worked full time in the current occupation [pertains only to those currently employed]?
FULLF          Worked full time in the most recent occupation [includes last previous occupation for those not currently                       employed]?
FULL93        Worked full time in the occupation held in January 1993?
FULL88        Worked full time in the occupation held in January 1988?

Location of job: for codes see Appendix F (Vol. II of General Population Codebook)

LOCC         Location of the current occupation [pertains only to those currently employed]
LOCF         Location of the most recent occupation [includes last previous occupation for those not currently employed]
LOC93       Location of the occupation held in January 1993
LOC88       Location of the occupation held in January 1988

TENUREC        Number of months since beginning the activity of the respondent at the time of the survey.
TENUREF         Number of months in the last job [total number of months in the last job for those not currently employed                                and  number of months since beginning the current job for those employed at the time of the survey].
TENURE93       Number of months since beginning the activity of the respondent in January 1993.
TENURE88       Number of months since beginning the activity of the respondent in January 1988.
WKMON          Total number of months employed since first leaving school

YEARF            Year last job began
YEARENDF    Year last job ended

1992 job characteristics (for income analysis)

These variables were created using a slightly different logic from what was used to create the -c, -f, -93, and -88 variables. The difference is that the -92 characteristics refer to the last activity begun prior to January 1, 1992, whereas the -93 and -88 variables refer to the activity reported for January 1993 and January 1998 respectively. The distinction is unlikely to have any practical importance. For details on the construction of the -92 variables see ACT92.INC in Appendix D.

The -92 variables are coded in exactly the same way as the -c, -f, -93, and -88 variables above, so the codes are not repeated. Note, however, that the order in which the -92 variables are listed differs slightly from that for the -c, -f, -93, and -88 variables.

ACT92           Activity in January 1992
YRBEG92      Year began January 1992 activity
OCC92           ISCO code for the occupation held in January 1992
ORG92           Organization type code for the occupation held in January 1992
SELF92          Self-employed in the job held in January 1992?
MOBEG92      Month began January 1992 activity
IND92             Industry code for the job held in 1992
FULL92          Worked full time in the job held in January 1992?
SUB92            Number of subordinates in the job held in January 1992
EGP92            EGP code for the occupation held in January 1992
SIX92             Six category EGP code for the occupation held in January 1992
ISE92              ISEI score for the occupation held in January 1992
PRE92             Prestige score for the occupation held in January 1992
EXPERIEN      Number of years of work experience [=WKMON/12]. See WKMON above.
WKMON92    Number of months employed in January 1992

Income

Because respondents were permitted to record incomes either as annual or as monthly incomes, we have created several sets of standardized income variables, making various assumptions specified below. The first set simply standardizes all incomes to monthly amounts, but makes no imputations of missing income data. The second set imputes values for missing or implausible reported values; there is also a set of "flag" variables that identifies imputed cases. The third set standardizes incomes to U.S. dollar equivalents. Users who are unhappy with our manipulations of the data are of course free to make their own, since the original variables are retained in Part II.

The following variables constitute the first set--monthly incomes, with blanks and missing values set to zero. Thus, in this set of variables 0 is indistinguishable from missing values.

RMTOT       Respondent's total monthly income

These variables were created as follows--see the *N2.INC files in Appendix D for details: Income from main job reported in years--or judged to refer to yearly income--was converted to monthly income by dividing by the number of months worked in 1992 or, if this was not known, by 12. Income from social benefits was converted by dividing by the number of months in which the respondent was not working. The remaining income variables for respondents and all income variables for spouses were converted by dividing by 12, since information to make a more precise specification of months was not available. Incomes from the four sources were then summed to produce a total income variable for respondents and for spouses, respectively.

The following variables constitute the second set--monthly incomes, with missing values and implausible values (values that could not realistically occur in these societies given the other characteristics of respondents, and hence almost certainly constitute errors in reporting or recording) imputed. The imputation was made by assigning expected values from a regression of each income variable being imputed on a set of characteristics known to determine incomes; separate estimates of social welfare incomes were made for the retired and those on maternity leave (again, see the *N2.INC files in Appendix D for details). We are aware that using expected values rather than values that reflect the variance around the regression surface overstates the relation between income and its determinants. However, given the relatively small fraction of cases imputed, we do not believe much distortion is introduced. We have included "flag" variables that identify the imputed cases in order to enable users to study the consequences of our imputation method. We experimented with other forms of imputation, particularly procedures that replicated the distribution of observed values around each point on the regression surface, but rejected them on the ground that they produced too many negative values and that arbitrarily assigning a minimum value would distort the relationships even more than assigning point estimates. The preferred procedure, for those seriously interested in the analysis of incomes, might be to utilize multiple-imputation methods, which any user can do by starting with the original income variables in Part II.

Except where noted, cases that could not be imputed, because data were missing for the independent variables in the regression equations, were assigned a value of 0. The problem of how to decide in any particular instance whether a missing value for a type of income represents no income or represents no information is an extremely difficult one, which we do not believe we have completely solved. Moreover, decisions about how to handle missing income data depend--always, but with particular force for income--on the design of the analysis. We thus strongly urge users to begin any analysis of the income variables with a careful consideration of how to treat zeros and how to treat the missing values we have assigned.

CRMTOT    Corrected respondent's monthly income - total

Several additional variables were created, as follows: Monthly income from other household members and total monthly household income were estimated by dividing by 12 if the income was reported as annual. The houshold monthly per capita income was estimated by dividing the total monthly household income by the number of persons in the household.

TOTHMTOT Monthly income from other household members

HSHMTOT Total monthly household income

INCPCA Monthly household income per capita

The following variables indicate whether a value was imputed in the "corrected" income variables.

FLGCRTOT      Flag for CRMTOT
FLGHMTOT     Flag for HSHMTOT

Finally, we standardized incomes to U.S. dollars, on the basis of market exchange rates (Economist Intelligence Unit Country Reports). For all countries except Russia, we used the average exchange rates for 1992, since respondents were asked to report 1992 incomes. xx FIX: CHECK WHAT WAS ASKED OF ELITE SAMPLE. For Russia we used the average exchange rate for the period the survey was in the field (April-June 1993), since respondents were for the most part asked to report their current incomes. Since at the time of the survey the exchange rates for the Czech Republic and Slovakia were identical, we wed the Czech exchange rate for both countries.

Here are the exchange rates used: xx FIX TO GET EXCHANGE RATES AT TIME OF FIELDWORK.

Bulgaria                   21.2 leva = $1.00
Czech Republic       28.6 korun = $1.00
Poland                    18,732 zloti = $1.00 [average of July, August, Sept. 1993]
Hungary                  81.5 forints = $1.00
Russia                     867 rubles = $1.00
Slovakia                  xx
Slovenia                  xx

The list of standardized variables follows. These variables have the same missing data structure as the second set, the variables with imputed values.

S$CRMTOT           Standardized r's monthly income - total
S$HSHM                Standardized household total monthly income
S$INCPA               Standardized income per capita

Languages spoken

Questions T1 and T2 are converted into six variables, indicating the number of additional languages spoken and the level of competence at each of several languages.

NUMLANG           Total number of additional languages spoken (in addition to language of interview)

For each language mentioned, the respondent was asked: How well do you speak it? Only a little, fairly well, very well. This information was used to create the following variables.

ENGLISH              English language competence
FRENCH               French language competence
GERMAN              German language competence
RUSSIAN              Russian language competence
LANGCOMP        Highest level of competence in any language (except language of interview)

[Note: missing value codes not used for LANGCOMP; those mentioning a language but not responding positively to any of the "how well spoken" questions coded 0 on this variable.]

Communist Party history

CPEVER               Ever a member of the Communist Party?

CP88                      Member of the Communist Party in 1988?

YRLEFTCP           Year left Communist Party?

SPCPEVER           Spouse ever a member of the Communist Party?

FACPEVER          Father ever a member of the Communist Party?

MOCPEVER          Mother ever a member of the Communist Party?

CPOFFEVR           Ever held unpaid office in the Communist Party?

CPOFF88              Held unpaid office in the Communist Party in 1988?

YRLFTOFF          Year (last) left unpaid office in the Communist Party

MP88                    Member of Parliament in 1988?

MP93                    Memberof Parliament now?

Religion

RELIGB                 Religion in which the respondent was brought up

RELIGP                Respondent's current religion

COUNTRY-SPECIFIC VARIABLES

[Note: These variables are in separate files from the international variables.]

All countries

SURVEY                Sample interviewed

COUNTRY            Country in which survey conducted

RESPID1              Respondent identification number in the survey

xx THE COUNTRY-SPECIFIC VARIABLES FOR EACH COUNTRY NEED TO BE ADDED HERE.

Bulgaria

Czech Republic

Hungary

Poland

Russia

Parents' property

PHOUSE14          a Did your parents own or have personal use of a house or separate apartment?

PDACHA14           b Did your parents have a country house <dacha> of their own or state-owned                                   <vedomstvennaja>?

PCAR14                c What about a personal or service car? [that is, a car they owned or a state car available for their                                  business (and, in practice, personal) use.]

PHOUSE83          H5a Parents had house or apartment in 1983?
PHOUSE73          H6a Parents had house or apartment in 1973?
PHOUSE60          H7a Parents had house or apartment in 1960?
PHOUSE50          H8a Parents had house or apartment in 1950?
PHOUSE40          H9a Parents had house or apartment in 1940?
PDACHA83         H5b Parents had dacha or vedomstvennaja in 1983?
PDACHA73         H6b Parents had dacha or vedomstvennaja in 1973?
PDACHA60         H7b Parents had dacha or vedomstvennaja in 1960?
PDACHA50         H8b Parents had dacha or vedomstvennaja in 1950?
PDACHA40         H9b Parents had dacha or vedomstvennaja in 1940?
PCAR83               H5c Parents had personal or service car in 1983?
PCAR73               H6c Parents had personal or service car in 1973?
PCAR60               H7c Parents had personal or service car in 1960?
PCAR50               H8c Parents had personal or service car in 1950?
PCAR40               H9c Parents had personal or service car in 1940?

Attitudes about the current situation in Russia

UNIQUE          a The western countries cannot be an example for Russia; it has its own particular way of                                 development.
SOVIET            b Rejection of the [Soviet] Union was a major mistake of the politicians.
DICTATOR      c The only way out from Russia's current circumstances is to put all power in the hands of one                                  person.
FIXPRICE        d The state should fix stable prices for most goods.
EQUALINC      e The fairest way to distribute income is not to permit extreme differences in income.
FOREIGN         f Attracting foreign investments to Russia will result in exporting the land's wealth and the further                                  pauperization of the country.
BUSINESS       g Let business people get their profits since in the end this will be to everybody's benefit.
MARKET         h Transition to a market economy is necessary to overcome the current crisis and ensure the                                  prosperity of Russia.
PRIVATE         i The transfer of enterprises and land into private property is necessary to give people more                                incentive to work well.
BETTER            j It would be better if everything in the country remained as it was before 1985.

Getting ahead

RICH88                a1 Coming from a wealthy family [having rich parents]: 1988?
RICH93                a2 Coming from a wealthy family [having rich parents]: now?
EDUC88               b1 Having a good education: 1988?
EDUC93               b2 Having a good education: now?
AMBIT88             c1 Having ambition [being career-oriented]: 1988?
AMBIT93             c2 Having ambition [being career-oriented]: now?
WORK88             d1 Hard work: 1988?
WORK93             d2 Hard work: now?
NET88                  e1 Knowing the right people: 1988?
NET93                  e2 Knowing the right people: now?
POL88                  f1 Having political connections [having support from relatives, friends, etc.]: 1988?
POL93                  f2 Having political connections [having support from relatives, friends, etc.]: now?
RISK88                g1 Willingness to take a risk: 1988?
RISK93                g2 Willingness to take a risk: now?
REAR88               h1 Having a good upbringing: 1988?
REAR93               h2 Having a good upbringing: 1993?
LOYAL88             i1 Devotion to one's boss: 1988?
LOYAL93             i2 Devotion to one's boss: 1993?

Job attitudes [Questions W9 - W14 were asked only of those currently employed.]

xx CONTINUE FORMATTING FROM HERE.

Back to Top of Page

For problems with this page, please e-mail libbie@ucla.edu.

Return to SSEE Home.