Data Set Description for Chapter 6: The Parliaments and governments database (ParlGov)

Data Exercise Contributor: Jens Wäckerle

Dataset-Chapter-6.utf8

The Parliaments and governments database (ParlGov) project provides two key services: A website that serves as an encyclopedia of elections, parties and cabinets and corresponding ready-to-use datasets that can be easily integrated with other datasets for research projects. The main homepage can be found here and provides an interface for all elections, cabinets and parties in all EU countries and most OECD democracies. All in all the website has 37 countries, with around 1700 parties, 990 elections and 1600 cabinets. The database is maintained by Holger Döring and Philip Manow at the University of Bremen.

The dataset can be accessed here. We will present the dataset below. While reading, please keep in mind the questions you see below and answer them once you reached the end. In the end, we will provide a link to a platform with an interactive version of the dataset and additional tasks.

Table 1: General Tasks for the Dataset
Tasks
Which electoral systems tend to have a higher number of parties in parliament?
Within electoral systems, there is considerable variation between countries in terms of the number of parties in parliament. Why? Provide examples.

Dataset Description

ParlGov provides data in multiple datasets. The three main ones are cabinet-level data, election-level data and party data. We will only discuss election and party data in this chapter. We will discuss the cabinet level dataset in the online material for chapter 10.

Election-Level Data

In the election dataset, the unit of analysis is a party in an election. For each election, we have an election type (parliament or European parliament) and a date. For each party the dataset provides vote share, seats, name and left-right position. Similar to the cabinet level dataset, there are id variables to link this dataset to the others. Table 2 shows an example from the last two elections in Greece.

Table 2: Variables in the Election dataset
country_name election_date party_name_short vote_share seats seats_total
Greece 2015-09-20 SYRIZA 35.46 145 300
Greece 2015-09-20 ND 28.10 75 300
Greece 2015-09-20 LS-CA 6.99 18 300
Greece 2015-09-20 PASOK 6.28 17 300
Greece 2015-09-20 KKE 5.55 15 300
Greece 2015-09-20 TP 4.09 11 300
Greece 2015-09-20 ANEL 3.69 10 300
Greece 2015-09-20 EK 3.43 9 300
Greece 2015-09-20 LE|PE 2.86 0 300
Greece 2019-07-07 ND 39.85 158 300
Greece 2019-07-07 SYRIZA 31.53 86 300
Greece 2019-07-07 PASOK 8.10 22 300
Greece 2019-07-07 KKE 5.30 15 300
Greece 2019-07-07 EL 3.70 10 300
Greece 2019-07-07 MR25 3.44 9 300
Greece 2019-07-07 LS-CA 2.93 0 300
Greece 2019-07-07 LE|PE 1.46 0 300
Greece 2019-07-07 EK 1.24 0 300

Party-Level Data

The party level dataset provides information for each unique party in the ParlGov database. The variables include name of the party, name of the party family and some ideological scales such as left-right and libertarian-authoritarian. It also provides a host of id variables to link the dataset to the other ParlGov datasets, but also to a host of external dataset, such as the Comparative Manifesto Project or the Chapel Hill Expert Survey. Table 3 shows an example from the major parties in Greece.

Table 3: Variables in the Party dataset
country_name party_name_short family_name left_right liberty_authority
Greece DIKKI Social democracy 2.3000 5.5493
Greece KKE Communist/Socialist 1.2530 5.1503
Greece LAOS Right-wing 9.1100 9.6300
Greece ND Christian democracy 6.7365 6.6712
Greece OP Green/Ecologist 2.5000 1.8000
Greece PASOK Social democracy 4.4968 3.6025
Greece POLAN Conservative 7.5714 7.0000
Greece SYN Communist/Socialist 2.7683 2.0238
Greece SYRIZA Communist/Socialist 2.8947 2.0066

Additional Data

In addition to these three datasets, ParlGov provides additional data sets that can easily be merged with the election, party and cabinet datasets. These include election parameters, for example measures of effective number of parties in parliament, polarization and disproportionality. Table 4 shows an example from Swedish elections since 1970.

Table 4: Variables in the Election parameter dataset
country_name election_date turnout enp_seats disproportionality polarization_seats
Sweden 1970-09-20 88.28 3.315829 1.6112804 0.3661790
Sweden 1973-09-16 90.84 3.351206 1.5411035 0.3739609
Sweden 1976-09-19 91.76 3.447036 1.2331752 0.3722299
Sweden 1979-09-16 90.72 3.481521 1.2951917 0.3962212
Sweden 1982-09-19 91.45 3.128799 2.3915941 0.4053347
Sweden 1985-09-15 89.85 3.395718 1.3025761 0.3955113
Sweden 1988-09-18 85.96 3.665062 2.4327626 0.3926681
Sweden 1991-09-15 86.74 4.192661 2.8821918 0.4406164
Sweden 1994-09-18 86.83 3.504157 1.1277405 0.4354835
Sweden 1998-09-20 81.39 4.293757 1.2066181 0.4546626
Sweden 2002-09-15 80.11 4.225240 1.5017159 0.4302354
Sweden 2006-09-17 81.99 4.149242 2.8556280 0.4323410
Sweden 2010-09-19 84.63 4.537195 1.2137336 0.4519894
Sweden 2014-09-14 85.81 4.993277 2.6216632 0.4774061
Sweden 2018-09-09 87.18 5.631374 0.4585756 0.4881628

Example: Effective Number of Parties and Electoral Systems

Using the election parameters and election datasets, we can study the effective number of parties by electoral system. Figure 1 shows the most recent election in the European countries in the ParlGov dataset, classified according to Figure 6.1 in the book.

Figure 1: Effective Number of Parliamentary Parties

Figure 1: Effective Number of Parliamentary Parties

Interactive Activity

Here, you will find an interactive version of the ParlGov dataset and several questions to answer and discuss. We suggest you open this app on a laptop or tablet. Enjoy!

The book introduces students to the most current theoretical and empirical research on European politics, and it does so in a highly accessible way through examples and data visualizations.