In lieu of an abstract, here is a brief excerpt of the content:

Appendix 2. Changes in Civil Liberties Rankings Figures 6 and 7, in Chapter 6, chart changes in civil liberties over election periods. The outcomes matrices from the empirical analysis of the results displayed in those figures are presented below in Tables 19 and 20. The first matrix is based on an analysis of how many units of change in CL ranking are associated with each set of elections and the periods between them. Thus, if 10 first elections resulted in a positive change (more or better civil liberties) of two units on the CL scale each, that would make a total positive change of 20. If among first elections there were also 5 elections that each resulted in a negative unit change (less or worse civil liberties) of one unit, that would make a total negative change of 5. The balance (20 − 5) would accordingly be a total of 15. The second matrix is based on an analysis of how many of the elections (expressed as shares) in each electoral cycle are associated with changes in CL scores, either independently of the elections or dependent on them. table 19 Unit changes of ranking in civil liberties over electoral periods (data for Fig. 6) 1st Period 2nd Period 3rd Period 4th Period Pre- 1st Non- 2nd Non- 3rd Non- 4th election Election election Election election Election election Election (A) Effects (B) Effects (C) Effects (D) Effects Positive changes 23 62 8 17 1 5 1 2 Negative changes 10 3 11 4 1 1 1 4 Balance 13 59 3 13 0 4 0 2 table 20 Percentage of elections followed by changes in civil liberties (data for Fig. 7) 1st Period 2nd Period 3rd Period 4th Period Pre- 1st Non- 2nd Non- 3rd Non- 4th election Election election Election election Election election Election (A) Effects (B) Effects (C) Effects (D) Effects Positive change (%) 26 47 13 28 3 10 5 10 Negative change (%) 11 4 18 7 3 3 5 14 No change (%) 63 49 68 66 93 87 91 76 Total % 100 100 99 101 99 100 101 100 Total N 70 70 60 61 30 30 21 21 Table 21 gives the unit changes in CL scores for each country’s elections up to four electoral cycles, where that many have occurred. Selection of cases was the same as in Chapter 6. The countries have been coded as “Yes” when it fully corroborates the main hypothesis, as set out in Chapter 5, “(Yes)” when there was a slight deviation but the country ’s track record overall corroborates the hypothesis, or “No” when the country’s track record does not support the hypothesis, either contradicting it or providing inconclusive evidence. Changes in civil liberties occurring as a result of the election are measured as changes in civil liberties according to the operationalization in Chapter 6. A few deviations have been necessary, however, for various reasons. A few countries in Africa run presidential and parliamentary elections at different intervals. In these circumstances, the period of measurement has been adjusted in the data set to capture years unaffected by elections, as per the specification of election-related changes above. In some instances, coups and premature legislative elections have made the equation impossible. The full list of adjustments that had consequences for the coding is as follows: Burkina Faso, 2nd presidential election, 1998 Period: 1992–1996 reduced to 1993–1994 due to the parliamentary elections in 1992 and 1995. Change: 1993 record a ranking of “4” instead of “5” in 1992. Chad, 2nd presidential election, 2001 Period: 1997–1999 reduced to 1998–1999 due to the parliamentary election in 1997. Change: 1998 record a ranking of “4” instead of “5” in 1997. Djibouti, 3rd parliamentary election, 2003 Period: 1998–2001 reduced to 2000–2001 due to the presidential election in 1999. Change: 2000 record a ranking of “5” instead of “6” in 1998. Gabon, 1st presidential election, 1993 Period: 1989–1991 changed to 1991–1992 due to the parliamentary election in 1990. Change: 1991 record a ranking of “3” instead of “5” in 1989, and 1992 ranked “4” instead of “3” in 1991. Gambia, 1st presidential election, 1996 Period: 1992–1994 reduced to 1994–1995 due to coup in 1994. Change: 1992 ranked “2,” but after the coup in 1994 it was given a ranking of “6.” Mauritania, 3rd parliamentary election, 2001 Period: 1997–1999 reduced to 1998–1999 due to general elections 1997. Change: 1997 was ranked “6” instead of...

Share