British electors, on the face of it, are an ungrateful bunch. Consumer confidence, as measured by the EU Commission / GfK monthly indicator, is at the top of its range over the last 40 years – in the 96th percentile, to be precise, up from the 44th just before the May 2010 general election. The governing Conservative and Liberal Democrat parties, however, command the support of only 42% of voters, according to the poll of polls compiled by ukpollingreport.com, down from a 59% share at the last election.

Mismatches between consumer confidence and government popularity, in fact, are nothing new. The party of government has changed in elections four times since the early 1970s: February 1974 (Conservative to Labour), May 1979 (Conservative win), May 1997 (Labour win) and May 2010 (Conservative- Liberal Democrat coalition). Consumer confidence was also at the top of its range in 1979 and 1997, standing in the 96th and 95th percentiles respectively – see first chart. (Confidence had slumped during the 1978-79 “winter of discontent” but rebounded strongly into the election.)


It would be wrong to conclude that economic / financial factors are of little importance to voters. Statistical research reported in previous posts (e.g. here) shows that the poll difference between the main governing and opposition parties depends positively on average earnings growth and house price inflation, and negatively on interest rates, consumer price inflation and the unemployment rate. A model based on these factors explains a significant portion of historical poll variation and continues to suggest a small Conservative lead over Labour on 7 May.

Consumer confidence is an inferior political guide probably for two reasons. First, the confidence indicator is a forward-looking measure: it combines consumers’ expectations of their finances, saving behaviour, the economy and unemployment over the next 12 months. Voters, however, appear to rate governments on what they have delivered to date, discounting promised improvement, even when this is believed.

Secondly, earnings growth is much more important for voting intentions than other influences on confidence. The statistical research indicates that a 1 percentage point (pp) rise in earnings growth moves the poll gap by 4 pp in favour of the governing party; a 1 pp fall in consumer price inflation or the unemployment rate, by contrast, delivers a boost of only 1 pp. (Changes in interest rates also have a significant impact, explaining why governments and their central bank appointees try so hard to hold rates down before elections.)
 
These considerations suggest that political analysts should focus attention on the component of the consumer survey tracking current personal finances rather than the broader and forward-looking confidence measure. This component has lagged the pick-up in confidence, currently standing in the 70th percentile of its historical range. It is, however, rising fast, regaining its 2008 pre-recession level in March – second chart.

The Bank of England forecasts a rise in earnings growth to 3.5% in the fourth quarter of 2015, from 1.7% in the three months to February. The polling model suggests that the Conservative lead over Labour would be 6 pp if such growth was occurring now. Tory-supporting business leaders are increasingly frustrated by the lack of traction of the party’s election campaign, according to today’s Financial Times. Have they considered the impact of their own wage policies on the electoral arithmetic?