Abstract copyright UK Data Service and data collection copyright owner.
The data arose from an inquiry into liberal democratic performance, understood as the delivery of liberal democratic values, or how far liberal democratic governments achieve in practice the values to which they subscribe in principle. Normative inquiry established that liberal democracy is founded on the principles of liberty and equality, which are achieved in practice through the operation of eight values: the legal values of civil rights, property rights, political rights and minority rights; and the institutional values of accountability, representation, constraint and participation. Twenty-one quantitative indicators of liberal democratic performance from a variety of sources were selected as proxy measures for these eight values. Data were collected for 40 countries (both new and old liberal democracies) for the years 1970 to 1998. The disaggregated structure of the database aimed to provide a range of scores for each country across different liberal democratic values, rather than subsuming them into a single 'democracy index'. These data are being used to investigate four main issues: (1) the extent to which there has been a 'third wave' of democratisation; (2) the comparative performance of old, established liberal democracies; (3) the relationship between constitutional design and liberal democratic performance; (4) and the relationship between economic development and liberal democratic performance. The data were primarily used descriptively, rather than for causal analysis. In order to 'ground' the data and check its plausibility, the data were supplemented with more detailed qualitative material on particular countries and issues. Overall the research aimed to provide a global comparative depiction of liberal democratic performance.
Main Topics:
The data file contains the Database of Liberal Democratic Performance (in SPSS format). The data cover the years 1970 to 1998, inclusive. There are 21 variables and forty country cases. Since scores do not exist for each variable for each year, there are 8,958 observations out of a potential total of 24,360. The 21 variables are organised according to eight liberal democratic values: the legal values of civil rights, property rights, political rights and minority rights; and the institutional values of accountability, representation, constraint and participation. The forty countries are divided between 'old' (pre-1970) liberal democracies (17 countries) and 'new' (1970 and after) liberal democracies (23 countries). The majority of variables have been taken from existing datasets, although in a few cases variables were created. Where necessary the dataset also contains inverted versions of the variables (so that each indicator runs from low performance to high performance) and scaled versions of the variables (so that each indicator runs from zero to one). Additionally, there are a number of dummy variables which divide the country cases into different regions and/or other analytical categories. Country cases were selected on both methodological and practical grounds. The initial population was the 118 countries described as either 'liberal' or 'electoral' democracies by Diamond (1997), and as reaching a minimum threshold of procedural democracy by Freedom House standards. Countries with less than 1.5 million inhabitants were eliminated (leaving 82 cases), as were countries formed or reformed as nation-states since 1989, including both Germany and the Czech Republic (leaving 67 cases). Countries were also eliminated if they did not appear in either the Minorities at Risk database (Haxton and Gurr, 1997) or the Cingranelli and Richards Human Rights database or the Political Risk Services database on property rights (Knack and Keefer, 1995) - leaving 56 cases of which 17 were 'established' democracies. Further countries were eliminated in order to balance the number of 'established' and 'new' democracies, and to achieve an appropriate geographical spread. This left 40 countries, distributed into selected clusters on historical, geographical, economic or institutional grounds. Clusters such as the three former Central American dictatorships (Guatemala, El Salvador and Nicaragua) were included, whereas cases like Papua New Guinea or Mali were excluded (see Diamond, 1997). Please see documentation for full details of references quoted in text above.
Purposive selection/case studies
Transcription of existing materials
variables were taken from existing datasets which were available both electronically and in hard co
