Site MenuWhy is Turnout Higher in Some Countries than in Others?
3. Turnout and Macro-institutional Variables
Tables 2A and 2B show the correlation with the macro-institutional variables once the effects of the socio-economic and geographic variables mentioned in the previous paragraph have been neutralized.
3.1 Compulsory Voting
The most important variable is legislation imposing compulsory voting. Its influence has been seen in all the studies analyzing the effects of institutional factors on turnout. All other things being equal, turnout as a function of the number of registered electors is 13 percent higher in countries where voting is compulsory and penalties are imposed for failure to comply (Table 2A). However, turnout does not seem to be affected by the obligation to vote when there are no penalties for failure to comply.
The results are quite divergent when looking at turnout in comparison with the population of voting age (Table 2B). In this case, compulsory voting does not seem to have any effect, whether enforced by penalties or not. This result does not seem very credible to us, in view of the fact that all previous studies have found that compulsory voting increases turnout and the fact that the abolition of compulsory voting in the Netherlands in the early 1970s effectively reduced turnout there by about 10 percentage points. We have more confidence in the results when turnout is calculated on the basis of registered electors (Table 2A). What our study shows, and what has never been shown by any previous research, is that compulsory voting does not really have any effect unless penalties are stipulated for electors who decide to abstain. A merely symbolic obligation is not sufficient.
3.2 Voting System
Voter turnout is 5 or 6 percentage points higher in countries in which the voting system is proportional or mixed compensatory. This result is consistent with what was found in previous research. It appears that electors are more inclined to vote when the voting system seems fairer for all parties, including small ones (Blais and Dobrzynska, 1998).
| Independent Variables | Dependent Variable: Turnout by Number of People Registered on the Lists of Electors |
|
|---|---|---|
| Regression Coefficient |
(Error) |
|
| North America | -4.23 |
(2.93) |
| Eastern Europe | -6.69** |
(2.59) |
| Population (log) | -3.27*** |
(1.09) |
| GDP per capita (log) | 9.57*** |
(2.42) |
| Compulsory voting with penalties | 12.60*** |
(3.32) |
| Compulsory voting without penalties | -3.82 |
(2.90) |
| PR and mixed compensatory | 4.80** |
(2.19) |
| Mali | -39.33*** |
(10.65) |
| Constant | 56.84*** |
(10.80) |
Number of cases: 148
Adjusted R2: 0.40
* significant at 0.10 (two-tailed test)
** significant at 0.05 (two-tailed test)
*** significant at 0.01 (two-tailed test)
| Independent Variables | Dependent Variable: Turnout by Population of Voting Age |
|
|---|---|---|
| Regression Coefficient |
(Error) |
|
| North America | -5.62 |
(3.81) |
| Eastern Europe | -6.35* |
(3.37) |
| Population (log) | -2.93* |
(1.51) |
| GDP per capita (log) | 10.98*** |
(3.10) |
| Compulsory voting with penalties | 0.18 |
(4.41) |
| Compulsory voting without penalties | 1.59 |
(3.88) |
| PR and mixed compensatory | 6.11** |
(2.86) |
| Mali | -34.59*** |
(12.06) |
| Constant | 46.70*** |
(14.24) |
Number of cases: 108
Adjusted R2: 0.30
* significant at 0.10 (two-tailed test)
** significant at 0.05 (two-tailed test)
*** significant at 0.01 (two-tailed test)