Report on New Jersey's electoral redistribution process and its relevance for South Australia

This paper reports on a study trip to New Jersey to investigate that state's electoral redistribution methodology, which includes a fair outcomes requirement. The author spoke with members of the New Jersey Apportionment Commission, representatives of both parties and academic commentators. The author found that an understanding of New Jersey's redistribution process puts South Australia's into perspective. It indicates that South Australia's Electoral Districts Boundaries Commission (EDBC) should continue to maintain that it is not responsible for a fair outcome, but only responsible for drawing a set of districts that will provide a level playing field at a subsequent election. Because parties must be allowed to campaign, tight election contests can be expected to show the effect of those campaigns, and it would be inappropriate for the Commission to attempt to compensate one party for the other's campaign. The methodology used by New Jersey's Apportionment Commission is remarkably similar to the methodology used by South Australia's EDBC, although the two jurisdictions appear to have developed their methodology separately. There is scope for change in South Australia in one area. Here, a set of proposed districts is assessed using the results of just one election. New Jersey uses several, to neutralize the effects of different campaigns in the data. Results from just one election---even if it is the most recent election---may be misleading because one set of results cannot be assumed to be representative of all results. South Australia's new set of House of Assembly districts should be assessed against more than the 2010 election results. Wrong winner results have occurred in both New Jersey and South Australia but because they are understood to have been caused by campaign effects---specifically turnout differences---they have not called the electoral system or the redistribution methodology into question. At New Jersey's redistribution in early 2011, fair outcome criteria were used for the fourth successive time. Other US jurisdictions share the EDBC's understanding that the task of designing a level playing field involves ensuring that the system will respond when voters change their allegiance, as well as ensuring that the system contains no advantage to either party. Apart from New Jersey, US jurisdictions do not address geographically-induced bias. However, if or when the US Supreme Court does formulate a standard that will enable it to invalidate gerrymanders, the standard seems likely to include a requirement that a fair plan must reward both parties with a similar share of seats for any given share of the vote, and that in particular one party must not be more likely to win a majority of the seats if there is a 50:50 result. That is also the intent of South Australia's fairness clause.