What are the effects of election thresholds?

Research team

  • Emil OW Kirkegaard - Idea, data-gathering, data treatment, writing

Description and background

Many countries have an election threshold for their parliaments, but these vary a lot from country to country. I was curious about the effects and justifications for these thresholds. They are, after all, a conservative tool that helps keep smaller parties out of parliament. In that sense it is a blatant limitation of democracy. For that reason, there must be a strong justification for having them if they are to be had for good reasons. However, a literature search on both Google and Google Scholar gave about zero useful results. Strange that such an obvious and impactful part of democracy has not been researched. I therefore set out to study it myself. This page is about the findings.

Methods

Wikipedia has a list with election thresholds around the world. These were used, though their quality is somewhat unknown. Assuming that the errors are randomly distributed, the errors would tend to balance out and attenuate any correlation found, not make it strong. It is therefore an extra stringest test of such a relationship.

Tested relationships

Election threshold and ...

Results were as predicted by common sense hypothesis. The effect is significant in 2 of the 3 samples at the 5% level, but it has about the same size in all samples, which shows the effect is robust.

Data set n Correlation p value
1) counting parties in both houses and in unicameral systems 36 -0,485686 0,002675
2) counting parties only in the upper houses and unicameral systems 34 -0,456778 0,006615
3) counting parties only in unicameral systems 23 -0,400363 0,058350
  • Number of parties in parliament and Human Development Index

 

  • ... Human Development Index