Consumers do NOT stop spending before federal elections

Charlie Nelson
April 2013

It seems to be a common belief amongst retailers that consumers reduce their spending before federal elections.  There has never been any evidence provided to support this belief, nor any has any logical reason been provided for pre-election austerity.  In fact, the evidence shows that there is no spending strike before federal elections.

I have analysed retail sales data before, during, and after the last ten federal elections.  This covers the period 1984 to 2010.

If it were true that consumers reduced spending before elections, the annual growth pattern would look like that shown in Chart 1.  In this chart the month of the election is labelled month zero and is shaded in red.  The month before the election is labelled -1 and the month after the election is labelled 1.

If the retailers’ claim is correct, we would expect a fall in the growth rate perhaps two months before the election, a bigger fall in the month before the election, and also lower growth in the month of the election – depending on the timing of the election within the month.  Also, we may expect some increase in growth immediately after the election due to deferred spending.

The actual growth pattern is shown in Chart 2.  There is no statistically significant drop in growth in the months before the election.  Some months are higher and some lower, but not by a statistically significant margin.  Individually and cumulatively, the seven months before an election experience the same growth as the month of the election.

Growth in the months after the election are individually identical to growth in the month of the election.  So there is no drop in growth before the election and no catch-up increase in growth after the election.

Retailers should be more concerned about cumulative growth after elections.  Over the past ten elections, cumulative growth in the months after the election is significantly lower than cumulative growth in the months before the election.

This finding may have nothing to do with elections.  For example, retail sales growth slumped from January 2008, two months after the 2007 election, because interest rates had been raised too high by the Reserve Bank of Australia.  They increased interest rates once before the election date was set, once during the campaign, and twice in early 2008.  Not smart decisions on the eve of the Global Financial Crisis, which they clearly thought had no chance of happening.

But a cynic might suggest that Prime Ministers call elections at the peak of the economic cycle, especially if they expect a slowdown is coming.  So why did Julia Gillard make an early call on an election for September 2013?

If pessimistic retailers cut ad spending before an election, they will just hand market share to the more optimistic retailers who keep advertising (or even increase advertising to lift share of voice).  The pessimistic retailer will suffer from a self-fulfilling prophecy – for themselves but not for the market.

It may have been the case, before the 1980’s, that consumers were cautious before elections.  In those distant past days, an incoming government could make big changes to interest rates and exchange rates.  But the dollar was floated in 1983 and is now determined by the market.  And the Reserve Bank has been setting interest rates independently since the 1990’s (much to the chagrin of the government in early November 2007).

Chart 1

Chart 2