Joint Programming Initiative

More Years, Better Lives

The Potential and Challenges of Demographic Change

Survey of Household Finances, 2008 (EFF)
Encuesta Financiera de las Familias, 2008 (EFF)

Topic
Social Systems and Welfare
Work and Productivity
Wellbeing
Housing, Urban Development and Mobility
Relevance for this Topic
Country Spain
URL
More Topics

Governance

Contact information

Olimpia Bover
Banco de España, Departamento de Estudios Monetarios y Financieros (Department of Monetary and Financial Studies, Bank of Spain)
Alcalá, 48

28014 Madrid
Spain
Phone: 91 338 5000
Email: bover(at)bde.es
Url: www.bde.es/.../EFF_2008.html

Timeliness, transparency

No information is provided in this matter

Type of data


Survey

Type of Study

Longitude survey: long-term study of random or different samples

Cross-section, regular

Data gathering method

Face-to-face interview (CAPI, PAPI)

Type of data


Survey

Type of Study

Longitude survey: long-term study of random or different samples

Cross-section, regular

Data gathering method

Face-to-face interview (CAPI, PAPI)

Type of data


Survey

Type of Study

Longitude survey: long-term study of random or different samples

Cross-section, regular

Data gathering method

Face-to-face interview (CAPI, PAPI)

Type of data


Survey

Type of Study

Longitude survey: long-term study of random or different samples

Cross-section, regular

Data gathering method

Face-to-face interview (CAPI, PAPI)


Access to data


Ad-hoc tables (cross-sectional an longitudinal) and microdata can be accessed online in Spanish (http://app.bde.es/eff_www/microdatos). Tables can be controlled by income percentiles, age of the main breadwinner, occupation, educational level, housing tenure, household members that work, and household size.

Conditions of access


Agreements with the Spanish Statistical Office (INE) and the Tax Office are necessary in order to build and carry out the sample.


No information


Aggregated tables according to the main areas of research and five sets of imputed anonymised microdata files. Data can be accessed for free once permission is granted.


Data are compiled in STATA and Excel (csv) formats


Essential 2008 EFF documents that are available in English and can be freely accessed (http://www.bde.es/bde/es/areas/estadis/Otras_estadistic/Encuesta_Financi/Contenedor_encue/EFF_2008/EFF_2008.htm) are: - The Spanish Survey of Household Finances (EFF): description and methods of the 2008 wave (Occasional Paper N.º 1103) - The User Guide - English Questionnaire Some other documents are only available in Spanish (http://app.bde.es/eff_www/microdatos): - Main results - Variables definitions to built tables - Hidden variables and replicated weights files

Access to data


Ad-hoc tables (cross-sectional an longitudinal) and microdata can be accessed online in Spanish (http://app.bde.es/eff_www/microdatos). Tables can be controlled by income percentiles, age of the main breadwinner, occupation, educational level, housing tenure, household members that work, and household size.

Conditions of access


Agreements with the Spanish Statistical Office (INE) and the Tax Office are necessary in order to build and carry out the sample.


No information


Aggregated tables according to the main areas of research and five sets of imputed anonymised microdata files. Data can be accessed for free once permission is granted.


Data are compiled in STATA and Excel (csv) formats


Essential 2008 EFF documents that are available in English and can be freely accessed (http://www.bde.es/bde/es/areas/estadis/Otras_estadistic/Encuesta_Financi/Contenedor_encue/EFF_2008/EFF_2008.htm) are: - The Spanish Survey of Household Finances (EFF): description and methods of the 2008 wave (Occasional Paper N.º 1103) - The User Guide - English Questionnaire Some other documents are only available in Spanish (http://app.bde.es/eff_www/microdatos): - Main results - Variables definitions to built tables - Hidden variables and replicated weights files

Access to data


Ad-hoc tables (cross-sectional an longitudinal) and microdata can be accessed online in Spanish (http://app.bde.es/eff_www/microdatos). Tables can be controlled by income percentiles, age of the main breadwinner, occupation, educational level, housing tenure, household members that work, and household size.

Conditions of access


Agreements with the Spanish Statistical Office (INE) and the Tax Office are necessary in order to build and carry out the sample.


No information


Aggregated tables according to the main areas of research and five sets of imputed anonymised microdata files. Data can be accessed for free once permission is granted.


Data are compiled in STATA and Excel (csv) formats


Essential 2008 EFF documents that are available in English and can be freely accessed (http://www.bde.es/bde/es/areas/estadis/Otras_estadistic/Encuesta_Financi/Contenedor_encue/EFF_2008/EFF_2008.htm) are: - The Spanish Survey of Household Finances (EFF): description and methods of the 2008 wave (Occasional Paper N.º 1103) - The User Guide - English Questionnaire Some other documents are only available in Spanish (http://app.bde.es/eff_www/microdatos): - Main results - Variables definitions to built tables - Hidden variables and replicated weights files

Access to data


Ad-hoc tables (cross-sectional an longitudinal) and microdata can be accessed online in Spanish (http://app.bde.es/eff_www/microdatos). Tables can be controlled by income percentiles, age of the main breadwinner, occupation, educational level, housing tenure, household members that work, and household size.

Conditions of access


Agreements with the Spanish Statistical Office (INE) and the Tax Office are necessary in order to build and carry out the sample.


No information


Aggregated tables according to the main areas of research and five sets of imputed anonymised microdata files. Data can be accessed for free once permission is granted.


Data are compiled in STATA and Excel (csv) formats


Essential 2008 EFF documents that are available in English and can be freely accessed (http://www.bde.es/bde/es/areas/estadis/Otras_estadistic/Encuesta_Financi/Contenedor_encue/EFF_2008/EFF_2008.htm) are: - The Spanish Survey of Household Finances (EFF): description and methods of the 2008 wave (Occasional Paper N.º 1103) - The User Guide - English Questionnaire Some other documents are only available in Spanish (http://app.bde.es/eff_www/microdatos): - Main results - Variables definitions to built tables - Hidden variables and replicated weights files


Coverage


Wave 1: Data collected in 2002. Wave 2: Data collected in 2005 with a sample size of 5,962 households. Wave 3: Data collected in 2008.The 2008 EFF wave is a partial panel by means of the re-interview of the 5,962 households that took part in the 2005 wave in order to obtain statistical information on transitions between states and individual changes in magnitudes. The total sample size was 6,197 interviewees to achieve representativeness and oversampling requirements. To preserve as much of the oversampling scheme devised for large municipalities and all provincial capitals as possible, tightly controlled replacements were chosen. For the wealthier 1%, the number of observations is over nine times what would be expected with random sampling. The total number of valid, completed interviews is 6,197. There are 3,967 households in the EFF2008 sample that had also participated in the EFF2005. Among the 3,967 panel households, 1,925 have participated since 2002 and 2,042 households have participated since 2005. The fieldwork lasted just over 8 months, from approximately the end of November 2008 to July 2009, although the reference period was March 2009. The fieldwork for the wealth and income surveys was particularly demanding because of the nature and difficulty of the questions asked.


2002


Households and individuals


Municipal Register (January 1, 2008) was the population frame to be sent out to the Tax Office as the basis to build the sample size with demographic, income and wealth data.


Spain


All ages of household members


The main EFF contents are as follows: - demographics (household size, composition, members occupation, civil status, educational level, health status) - real assets and their associated debts (main residence – ownership, purchase, value, means of payment, loans taken, refurbishment -, other residences, household furnishings, fittings and appliances, household means of transport) - other debts (loans, refinancing, credit lines) - businesses and financial assets (self-employment, shares in companies, accounts and deposits, mutual funds and portfolio investments, fixed income securities, other) - pension plans and insurances (pension schemes, life insurances, other) - labour market situation and labour income (employment an related income, working situations - employed, unemployed, self-employed, retired and disabled persons -, non-employment related incomes, income from employment in 2007, employment history) - income from real or financial assets in the previous calendar year (2007) - means of payment (instruments, cash machines, cheques, transfers and direct debits, telephone and Internet banking) - consumption and savings The classification variables to build tables are income percentiles, age of the main breadwinner, occupation, educational level, housing tenure, household members as workers and household size).


• Azpitarte, F. ”Measurement and identification of asset-poor households: a cross-national comparison of Spain and the United Kingdom”. The Journal of Economic Inequality 9(1) (2011): 87-110. DOI 10.1007/s10888-010-9135-2 • El-Attar, M., & Poschke, M. “Trust and the Choice Between Housing and Financial Assets: Evidence from Spanish Households”. Review of Finance 15(4) (2011): 727-756. DOI: 10.1093/rof/rfq030 • Fisher, P.J. ” Is There Evidence of Loss Aversion in Saving Behaviors in Spain?”. Journal of Family and Economic Issues 34(1) (2013): 41-51. Available at: link.springer.com/.../s10888-010-9135-2
• Pijoan-Mas, J., & Sánchez-Marcos, V. ” Spain is different: Falling trends of inequality”. Review of Economic Dynamics 13(1) (2010): 154–178. DOI:10.1016/j.red.2009.10.002

Coverage


Wave 1: Data collected in 2002. Wave 2: Data collected in 2005 with a sample size of 5,962 households. Wave 3: Data collected in 2008.The 2008 EFF wave is a partial panel by means of the re-interview of the 5,962 households that took part in the 2005 wave in order to obtain statistical information on transitions between states and individual changes in magnitudes. The total sample size was 6,197 interviewees to achieve representativeness and oversampling requirements. To preserve as much of the oversampling scheme devised for large municipalities and all provincial capitals as possible, tightly controlled replacements were chosen. For the wealthier 1%, the number of observations is over nine times what would be expected with random sampling. The total number of valid, completed interviews is 6,197. There are 3,967 households in the EFF2008 sample that had also participated in the EFF2005. Among the 3,967 panel households, 1,925 have participated since 2002 and 2,042 households have participated since 2005.


2002


Households and individuals


Municipal Register (January 1, 2008) was the population frame to be sent out to the Tax Office as the basis to build the sample size with demographic, income and wealth data.


Spain


All ages of household members


The main EFF contents are as follows: - demographics (household size, composition, members occupation, civil status, educational level, health status) - real assets and their associated debts (main residence – ownership, purchase, value, means of payment, loans taken, refurbishment , other residences, household furnishings, fittings and appliances, household means of transport) - other debts (loans, refinancing, credit lines) - businesses and financial assets (self-employment, shares in companies, accounts and deposits, mutual funds and portfolio investments, fixed income securities, other) - pension plans and insurances (pension schemes, life insurances, other) - labour market situation and labour income (employment and related income, working situations - employed, unemployed, self-employed, retired and disabled persons -, non-employment related incomes, income from employment in 2007, employment history) - income from real or financial assets in the previous calendar year (2007) - means of payment (instruments, cash machines, cheques, transfers and direct debits, telephone and Internet banking) - consumption and savings The classification variables to build tables are income percentiles, age of the main breadwinner, occupation, educational level, housing tenure, household members as workers and household size).


• Azpitarte, F. ”Measurement and identification of asset-poor households: a cross-national comparison of Spain and the United Kingdom”. The Journal of Economic Inequality 9(1) (2011): 87-110. DOI 10.1007/s10888-010-9135-2 • El-Attar, M., & Poschke, M. “Trust and the Choice Between Housing and Financial Assets: Evidence from Spanish Households”. Review of Finance 15(4) (2011): 727-756. DOI: 10.1093/rof/rfq030 • Fisher, P.J. ” Is There Evidence of Loss Aversion in Saving Behaviors in Spain?”. Journal of Family and Economic Issues 34(1) (2013): 41-51. Available at: link.springer.com/.../s10888-010-9135-2
• Pijoan-Mas, J., & Sánchez-Marcos, V. ” Spain is different: Falling trends of inequality”. Review of Economic Dynamics 13(1) (2010): 154–178. DOI:10.1016/j.red.2009.10.002

Coverage


Wave 1: Data collected in 2002. Wave 2: Data collected in 2005 with a sample size of 5,962 households. Wave 3: Data collected in 2008.The 2008 EFF wave is a partial panel by means of the re-interview of the 5,962 households that took part in the 2005 wave in order to obtain statistical information on transitions between states and individual changes in magnitudes. The total sample size was 6,197 interviewees to achieve representativeness and oversampling requirements. To preserve as much of the oversampling scheme devised for large municipalities and all provincial capitals as possible, tightly controlled replacements were chosen. For the wealthier 1%, the number of observations is over nine times what would be expected with random sampling. The total number of valid, completed interviews is 6,197. There are 3,967 households in the EFF2008 sample that had also participated in the EFF2005. Among the 3,967 panel households, 1,925 have participated since 2002 and 2,042 households have participated since 2005.


2002


Households and individuals


Municipal Register (January 1, 2008) was the population frame to be sent out to the Tax Office as the basis to build the sample size with demographic, income and wealth data.


Spain


All ages of household members


The main EFF contents are as follows: - demographics (household size, composition, members occupation, civil status, educational level, health status) - real assets and their associated debts (main residence – ownership, purchase, value, means of payment, loans taken, refurbishment , other residences, household furnishings, fittings and appliances, household means of transport) - other debts (loans, refinancing, credit lines) - businesses and financial assets (self-employment, shares in companies, accounts and deposits, mutual funds and portfolio investments, fixed income securities, other) - pension plans and insurances (pension schemes, life insurances, other) - labour market situation and labour income (employment and related income, working situations - employed, unemployed, self-employed, retired and disabled persons -, non-employment related incomes, income from employment in 2007, employment history) - income from real or financial assets in the previous calendar year (2007) - means of payment (instruments, cash machines, cheques, transfers and direct debits, telephone and Internet banking) - consumption and savings The classification variables to build tables are income percentiles, age of the main breadwinner, occupation, educational level, housing tenure, household members as workers and household size).


• Azpitarte, F. ”Measurement and identification of asset-poor households: a cross-national comparison of Spain and the United Kingdom”. The Journal of Economic Inequality 9(1) (2011): 87-110. DOI 10.1007/s10888-010-9135-2 • El-Attar, M., & Poschke, M. “Trust and the Choice Between Housing and Financial Assets: Evidence from Spanish Households”. Review of Finance 15(4) (2011): 727-756. DOI: 10.1093/rof/rfq030 • Fisher, P.J. ” Is There Evidence of Loss Aversion in Saving Behaviors in Spain?”. Journal of Family and Economic Issues 34(1) (2013): 41-51. Available at: link.springer.com/.../s10888-010-9135-2
• Pijoan-Mas, J., & Sánchez-Marcos, V. ” Spain is different: Falling trends of inequality”. Review of Economic Dynamics 13(1) (2010): 154–178. DOI:10.1016/j.red.2009.10.002

Coverage


Wave 1: Data collected in 2002. Wave 2: Data collected in 2005 with a sample size of 5,962 households. Wave 3: Data collected in 2008.The 2008 EFF wave is a partial panel by means of the re-interview of the 5,962 households that took part in the 2005 wave in order to obtain statistical information on transitions between states and individual changes in magnitudes. The total sample size was 6,197 interviewees to achieve representativeness and oversampling requirements. To preserve as much of the oversampling scheme devised for large municipalities and all provincial capitals as possible, tightly controlled replacements were chosen. For the wealthier 1%, the number of observations is over nine times what would be expected with random sampling. The total number of valid, completed interviews is 6,197. There are 3,967 households in the EFF2008 sample that had also participated in the EFF2005. Among the 3,967 panel households, 1,925 have participated since 2002 and 2,042 households have participated since 2005. The fieldwork lasted just over 8 months, from approximately the end of November 2008 to July 2009, although the reference period was March 2009. The fieldwork for the wealth and income surveys was particularly demanding because of the nature and difficulty of the questions asked.


2002


Households and individuals


Municipal Register (January 1, 2008) was the population frame to be sent out to the Tax Office as the basis to build the sample size with demographic, income and wealth data


Spain


All ages of household members


The main EFF contents are as follows: - demographics (household size, composition, members occupation, civil status, educational level, health status) - real assets and their associated debts (main residence – ownership, purchase, value, means of payment, loans taken, refurbishment -, other residences, household furnishings, fittings and appliances, household means of transport) - other debts (loans, refinancing, credit lines) - businesses and financial assets (self-employment, shares in companies, accounts and deposits, mutual funds and portfolio investments, fixed income securities, other) - pension plans and insurances (pension schemes, life insurances, other) - labour market situation and labour income (employment an related income, working situations - employed, unemployed, self-employed, retired and disabled persons -, non-employment related incomes, income from employment in 2007, employment history) - income from real or financial assets in the previous calendar year (2007) - means of payment (instruments, cash machines, cheques, transfers and direct debits, telephone and Internet banking) - consumption and savings The classification variables to build tables are income percentiles, age of the main breadwinner, occupation, educational level, housing tenure, household members as workers and household size).


• Azpitarte, F. ”Measurement and identification of asset-poor households: a cross-national comparison of Spain and the United Kingdom”. The Journal of Economic Inequality 9(1) (2011): 87-110. DOI 10.1007/s10888-010-9135-2 • El-Attar, M., & Poschke, M. “Trust and the Choice Between Housing and Financial Assets: Evidence from Spanish Households”. Review of Finance 15(4) (2011): 727-756. DOI: 10.1093/rof/rfq030 • Fisher, P.J. ” Is There Evidence of Loss Aversion in Saving Behaviors in Spain?”. Journal of Family and Economic Issues 34(1) (2013): 41-51. Available at: link.springer.com/.../s10888-010-9135-2
• Pijoan-Mas, J., & Sánchez-Marcos, V. ” Spain is different: Falling trends of inequality”. Review of Economic Dynamics 13(1) (2010): 154–178. DOI:10.1016/j.red.2009.10.002


Linkage


EFF2008 will allow comparisons with the new harmonised European wealth surveys, known as “Household Finance and Consumption Survey” (HFCS). The demographic questions were phrased so they would be comparable with similar questions in other household surveys carried out by the National Statistics Institute.


The three EEF waves can be partially accessed through the ID household variables to match all rotating interviewees.

Linkage


EFF2008 will allow comparisons with the new harmonised European wealth surveys, known as “Household Finance and Consumption Survey” (HFCS). The demographic questions were phrased so they would be comparable with similar questions in other household surveys carried out by the National Statistics Institute.


The three EEF waves can be partially accessed through the ID household variables to match all rotating interviewees.

Linkage


EFF2008 will allow comparisons with the new harmonised European wealth surveys, known as “Household Finance and Consumption Survey” (HFCS). The demographic questions were phrased so they would be comparable with similar questions in other household surveys carried out by the National Statistics Institute.


The three EEF waves can be partially accessed through the ID household variables to match all rotating interviewees.

Linkage


EFF2008 will allow comparisons with the new harmonised European wealth surveys, known as “Household Finance and Consumption Survey” (HFCS). The demographic questions were phrased so they would be comparable with similar questions in other household surveys carried out by the National Statistics Institute.


The three EEF waves can be partially accessed through the ID household variables to match all rotating interviewees


Data quality


The data from the completed interviews were revised in detail by the team at the Banco de España to uncover potential inconsistencies and implausible values in all stages of fieldwork with some discarded interviews (no income information, no information in Euros). In the EFF2008, both cross-sectional and longitudinal weights are provided. In line with the confidentiality restrictions mentioned above, design and non-response weights were calculated by the Tax Office following detailed instructions from the National Statistics Institute. More specific information on technical procedures is offered in Spanish and English (www.bde.es/.../do1103e.pdf
).


In order to allow for full comparability with the new HFCS, some small changes, with respect to the 2005 version, have been introduced to the 2008 questionnaire: a) a more explicit grid of questions determining the members of the sample, b) equity in privately held businesses is now determined independently from self-employment status, c) questions about ownership of managed accounts regarding assets not included in the specific asset categories, and d) some additional questions on inheritances received.


Efforts to homogenize definitions and wordings have been done since 2002, and especially between 2005 and 2008 (see Questionnaire).

Data quality


The data from the completed interviews were revised in detail by the team at the Banco de España to uncover potential inconsistencies and implausible values in all stages of fieldwork with some discarded interviews (no income information, no information in Euros). In the EFF2008, both cross-sectional and longitudinal weights are provided. In line with the confidentiality restrictions mentioned above, design and non-response weights were calculated by the Tax Office following detailed instructions from the National Statistics Institute. More specific information on technical procedures is offered in Spanish and English (www.bde.es/.../do1103e.pdf
).


In order to allow for full comparability with the new HFCS, some small changes, with respect to the 2005 version, have been introduced to the 2008 questionnaire: a) a more explicit grid of questions determining the members of the sample, b) equity in privately held businesses is now determined independently from self-employment status, c) questions about ownership of managed accounts regarding assets not included in the specific asset categories, and d) some additional questions on inheritances received.


Efforts to homogenize definitions and wordings have been done since 2002, and especially between 2005 and 2008 (see Questionnaire).

Data quality


The data from the completed interviews were revised in detail by the team at the Banco de España to uncover potential inconsistencies and implausible values in all stages of fieldwork with some discarded interviews (no income information, no information in Euros). In the EFF2008, both cross-sectional and longitudinal weights are provided. In line with the confidentiality restrictions mentioned above, design and non-response weights were calculated by the Tax Office following detailed instructions from the National Statistics Institute. More specific information on technical procedures is offered in Spanish and English (www.bde.es/.../do1103e.pdf
).


In order to allow for full comparability with the new HFCS, some small changes, with respect to the 2005 version, have been introduced to the 2008 questionnaire: a) a more explicit grid of questions determining the members of the sample, b) equity in privately held businesses is now determined independently from self-employment status, c) questions about ownership of managed accounts regarding assets not included in the specific asset categories, and d) some additional questions on inheritances received.


Efforts to homogenize definitions and wordings have been done since 2002, and especially between 2005 and 2008 (see Questionnaire).

Data quality


The data from the completed interviews were revised in detail by the team at the Banco de España to uncover potential inconsistencies and implausible values in all stages of fieldwork with some discarded interviews (no income information, no information in Euros). In the EFF2008, both cross-sectional and longitudinal weights are provided. In line with the confidentiality restrictions mentioned above, design and non-response weights were calculated by the Tax Office following detailed instructions from the National Statistics Institute. More specific information on technical procedures is offered in Spanish and English (www.bde.es/.../do1103e.pdf
).


In order to allow for full comparability with the new HFCS, some small changes, with respect to the 2005 version, have been introduced to the 2008 questionnaire: a) a more explicit grid of questions determining the members of the sample, b) equity in privately held businesses is now determined independently from self-employment status, c) questions about ownership of managed accounts regarding assets not included in the specific asset categories, and d) some additional questions on inheritances received.


Efforts to homogenize definitions and wordings have been done since 2002, and especially between 2005 and 2008 (see Questionnaire).


Applicability


The Spanish Survey of Household Finances (EFF) is a survey conducted by the Banco de España that collects information about income, assets, debts, and consumption at the household level. A desirable characteristic of the EFF present in all waves to date is the oversampling of wealthy households to achieve representativeness of the population and aggregate wealth. An additional important characteristic of the EFF is that the second and third waves have a full panel component to study transitions and to account for heterogeneity among households. However, a complete fresh cross-section collection would be better for capturing the structure of the population at the time of the survey. The compromise adopted was to try to re-interview all the previous wave respondents and, additionally, to incorporate a refreshment sample to preserve cross-sectional representativeness and overall sample size.

Applicability


The Spanish Survey of Household Finances (EFF) is a survey conducted by the Banco de España that collects information about income, assets, debts, and consumption at the household level. A desirable characteristic of the EFF present in all waves to date is the oversampling of wealthy households to achieve representativeness of the population and aggregate wealth. An additional important characteristic of the EFF is that the second and third waves have a full panel component to study transitions and to account for heterogeneity among households. However, a complete fresh cross-section collection would be better for capturing the structure of the population at the time of the survey. The compromise adopted was to try to re-interview all the previous wave respondents and, additionally, to incorporate a refreshment sample to preserve cross-sectional representativeness and overall sample size.

Applicability


The Spanish Survey of Household Finances (EFF) is a survey conducted by the Banco de España that collects information about income, assets, debts, and consumption at the household level. A desirable characteristic of the EFF present in all waves to date is the oversampling of wealthy households to achieve representativeness of the population and aggregate wealth. An additional important characteristic of the EFF is that the second and third waves have a full panel component to study transitions and to account for heterogeneity among households. However, a complete fresh cross-section collection would be better for capturing the structure of the population at the time of the survey. The compromise adopted was to try to re-interview all the previous wave respondents and, additionally, to incorporate a refreshment sample to preserve cross-sectional representativeness and overall sample size.

Applicability


The Spanish Survey of Household Finances (EFF) is a survey conducted by the Banco de España that collects information about income, assets, debts, and consumption at the household level. A desirable characteristic of the EFF present in all waves to date is the oversampling of wealthy households to achieve representativeness of the population and aggregate wealth. An additional important characteristic of the EFF is that the second and third waves have a full panel component to study transitions and to account for heterogeneity among households. However, a complete fresh cross-section collection would be better for capturing the structure of the population at the time of the survey. The compromise adopted was to try to re-interview all the previous wave respondents and, additionally, to incorporate a refreshment sample to preserve cross-sectional representativeness and overall sample size.


  • The information about this dataset was compiled by the author:
  • Vicente Rodríguez
  • (see Partners)