CSES Module 5: 2016-2021 Errata

CSES Module 5 Second Advance Release: May 14, 2020 Version

The Second Advance Release of CSES Module 5 of May 14, 2020, contains additions, updates, and improvements to the First Advance Release of May 21, 2019. Please do not use files from the previous release with files from the current version. Variable names and attributes, as well as the data, have sometimes been changed.

Eight new election studies have been added to the file, bringing the number of election studies included in Module 5 to 21 in total. The studies added to the Second Advance Release Dataset are in alphabetical order: Australia (2019), Brazil (2018), France (2017), Iceland (2016), Iceland (2017), New Zealand (2017), Norway (2017), and Turkey (2018).

Below, we provide a list of some of the primary changes and corrections to previously released Module 5 data. Users are advised that this list is not comprehensive but instead highlights the main modifications both to the data and documentation since the previous release.

DATA CHANGES

  1. E5000A-.E5000_I: To allow for easy identification of CSES alphabetic party/coalition codes within the dataset, nine new variables (E5000_A to E5000_I inclusive) have been added to the dataset. They serve as identifiers in the dataset for parties/coalitions that are classified as Party A, Party B, Party C etc… for each polity.
  2. E3009 (Government Performance) now includes the additional category “6. [See Election Study Notes]”.
  3. E3023 (Satisfaction with Democracy) now includes the additional category “6. [See Election Study Notes]”.
  4. E5078 (Effective Number Of Electoral Parties), E5079 (Corrected Effective Number Of Electoral Parties), E5080 (Effective Number Of Parliamentary Parties), E5081 (Corrected Effective Number Of Parliamentary Parties): The new category “997. Presidential Election Only – Not Calculated” has been added to these variables, to reflect that they are not calculated for election studies exclusively covering Presidential elections.
  5. E5092 (Gini Coefficient Of Equalized Disposable Income: Year Closest to Elec) has been re-classified as “Gini Coefficient Of Equalized Disposable Income: Year Closest to Election”. Availability of GINI data in the given election year often lags behind available data. Hence, CSES now provides the most recent year of data available closest to the election in question at the time of data processing.
  6. E5102 (TI Corruption Perception Index) has been updated for Chile (2017), Hungary (2018) and Italy (2018). For these election studies, E5102 now refers to data in the said election year, as this data has become available. Previously, data referred to the closest available year.
  7. E5111 (Percentage of Individuals Using the Internet), E5112 (Mobile Phone Subscriptions per 100 Inhabitants), E5113 (Fixed Telephone Lines per 100 Inhabitants): These data have been updated for Hungary (2018) and Italy (2018) to refer to data in the said election year, as this data has become available. Previously, data referred to the closest available year.

CODEBOOK CHANGES

  1. CODEBOOK PART 1-3: A brief explanation of the CSES Module 5 convention on numerical party/coalition coding schemes have been added to Parts 1, 2, and 3 of the Codebook. The information can be found in each segment of the Codebook by searching for the following: “<<>> CSES NUMERICAL PARTY/COALITION CODING”.
  2. CODEBOOK PART 1: The CSES Bibliography has been updated and sorted alphabetically.
  3. CODEBOOK PART 2 – E2008 (Socio-Economic Status) and E2009 (Employment Type): Variable Notes have been updated to further clarify the CSES convention for cases in which these questions have been asked to respondents out of labor force. The CSES principle is not to recode these inconsistencies but to report them in the Codebook.
  4. CODEBOOK PART 2 – E3015_LH_PL (Previous Lower House Election: Vote Choice – Party List), E3015_LH_DC (Previous Lower House Election: Vote Choice – District Candidate), E3015_UH_DC_ (Previous Upper House Election: Vote Choice – District Candidate): Variable Notes have been updated to further clarify the CSES convention for cases in which these questions have been asked to respondents stating not to have turned out in the previous election. The CSES principle is not to recode these inconsistencies but to report them in the Codebook.
  5. CODEBOOK PART 2 – E5092 (GINI Coefficient Of Equalized Disposable Income – Year Closest To Election Year Available): Variable Notes have been updated and now include a table providing an overview of the year in which the GINI coefficient was calculated.
  6. CODEBOOK PART 2 – E5106_1 (Population By Citizenship: Percentage Of Population Who Are Citizens), E5106_2 (Population By Citizenship: Percentage Of Population Foreign Born/Not Citizen), and E5106_3 (Population By Citizenship: Percentage Of Population Unknown Citizenship Status): Variable Notes have been updated and now include a table listing election studies and the year to which the data included in E5106_ refer to, as data is not always available for the year the election is held.
  7. CODEBOOK PART 2 – E5107 (Linguistic Fractionalization Index: Alesina et al. 2003), E5108 (Religious Fractionalization Index: Alesina et al. 2003), and E5109 (Ethnic Fractionalization Index: Alesina et al. 2003):Variable Notes have been updated and now include a table listing election studies and the year for which the indices have been calculated.
  8. CODEBOOK PART 2 – E2017 (Country of Birth):  For Hong Kong (2016), an Election Study Note has been added, explaining the election-study specific code “900. Other”.
  9. CODEBOOK PART 2 – E5008_1 (Voting Operations: Early/Advance Voting), E5008_2 (Voting Operations: Vote By Mail/Postal): For Ireland (2016), Election Study Notes for E5008_1 and E5008_2 have been added, detailing the circumstances under which Irish voters are eligible for advance and postal voting.
  10. CODEBOOK PART 3: For Hong Kong (2016), Election Study Notes for the Parties and Leaders table have been updated, informing users that data for PARTY H (Labor Party) is unavailable for E3017_H (Like-Dislike – Party H), E3018_H (Like-Dislike – Leader H), E3019_H (Left-Right – Party H), and E3021_H (Optional Alternative Scale – Party H).

DATA INCONSISTENCIES: CSES MODULE 5 ADVANCE RELEASE 2

We wish to alert users to four minor inconsistencies which persist in CSES Module 5 Advance Release 2. These discrepancies will be addressed in future releases of CSES Module 5.

  1. E2010 (Household Income – Quintiles) is coded “9. Missing” for 187 respondents for New Zealand (2017). However, these data should be classified as “8. Volunteered: Don’t Know”, as indicated by E2011 (Household Income – Original Variable).
  2. E3013_UH_DC_ (Current Upper House Election: Vote Choice – District Candidate) and E3015_UH_DC_ (Previous Upper House Election: Vote Choice – District Candidate) are coded “999995. Not Applicable: Unicameral System” for Chile (2017). However, these data should be classified as “999996. Not Applicable: No District Candidate Vote”, as respondents’ vote choice for the Chilean Senate is provided in the party-list vote variables E3013_UH_PL (Current Upper House Election: Vote Choice -Party List) and E3015_UH_PL (Previous Upper House Election: Vote Choice – Party List).
  3. E3013_UH_PF (Current Upper House Election: Candidate Preference Vote) is coded “95. Not Applicable: Unicameral System” for Italy (2018). However, these data should be classified as “96. Not Applicable: No Preference Vote System”, as Italy’s legislature is bicameral, but the electoral system for upper house elections does not allow a preferential vote to be cast.
  4. E5201_G (Parliaments And Government Database (ParlGov) Identifier) is coded “2643” for PARTY G (People’s Party) for Iceland (2017). The correct ParlGov identifier for that party is “2632”.

 

CSES Module 5 First Advance Release: May 21, 2019 Version

(no errata listed)