When I hear very serious people talk about confidence I often smell a rat. It’s such an amorphous thing and impossible to observe directly. Clearly there are times when it matters a lot, but I suspect it matters most at points of extremity, not most of the time. We’ve had the RBA talking a lot about the need for confidence to return to stimulate investment. But an awful lot of businesses invest more when they see demand outstripping supply and not before.
Anyway, here’s an abstract that’s worth noting in that regard, though it looks at confidence in another area. Often you hear pundits claiming that confidence is affected by punters’ opinions of how the government is going. Unless the effect is pretty strong – ie you might be spooked by really bad government – I find this dubious. And the evidence – or at least this evidence – suggests likewise.
Government Economic Policy, Sentiments, and Consumption
by Atif Mian, Amir Sufi, Nasim Khoshkhou – #21316 (EFG ME PE POL)
We examine how consumption responds to changes in sentiment regarding
government economic policy using cross-sectional variation across
counties in the ideological predisposition of constituents. When the
incumbent party loses a presidential election, individuals in
counties more ideologically predisposed toward the losing party
experience a dramatic and discontinuous relative decrease in optimism
on government economic policy. Using the interaction of constituent
ideology in a county with election timing as an instrument, we
estimate the impact of government policy sentiment shocks on consumer
spending, and we find a very small effect that cannot be
statistically distinguished from zero. The small magnitude of the
effect is estimated precisely. For example, we can reject the
hypothesis that pessimism regarding government economic policy
effectiveness during the Great Recession had as large an effect on
consumption as the negative shock to household net worth coming from
the collapse in house prices.