New Zealand’s change of electoral system from majoritarian to proportional representation is often hailed as an example of how PR systems are beneficial for getting more women into parliament. Indeed, the number of women elected in the 1996 election increased more than in any other election in New Zealand, leading many to attribute this change to the change in electoral system.
However, when the time span under consideration is increased to include historical election results, the suggestion that the change in electoral system helped to increase the number of women in national parliament is no longer as clear as in a simple before-after comparison.
The proportion of women in the New Zealand parliament started to increase in the 1970s. Between 1978 and 1993, the proportion of women in parliament increased in a relatively linear fashion, increasing at a rate of 3% per election (red trend line).
In 1996, the first time the new electoral system was used (PR), the proportion of women elected into parliament increased noticeably more than in previous elections (8%). In the election that followed, the proportion of women in parliament could not be increased, and for the three most recent elections (results for 2011 are not included in the plot: 32.2% women; they actually corroborate my case), the proportion was around what one would have expected based on the historical trend between 1978 and 1993.
If there were an effect attributable to the electoral system on its own, it is unclear why we did not observe a linear increase in the proportion of women in New Zealand’s parliament since 1996 (green trend line). In this case, the large increase in 1996 would constitute the “penalty” of having had a non-PR system. However, if the majoritarian system was acting as an impediment to including more women in national parliament, the results of the past three elections seem inexplicably low.
I argue that we cannot attribute the change in the proportion of women in New Zealand’s parliament to a change of the system with any confidence. This does not mean that electoral systems are irrelevant, but a more cautious approach should be taken when making causal statements.
Note: This is an elaboration of a point I make in my monograph on political representation I’ve meant to post for a long time.