On bleeding and leading

My AFR oped today asks: does it matter for economic policy which individual leads a particular political party? The conclusion: not in most cases.

Of course, a 750-word piece can only do justice to one strand of a vast literature. For example, my piece didn’t even begin to touch upon the fascinating research agenda of my political science colleague Paul ‘t Hart, to mention just one of the people doing interesting work on political leadership.

Leadership is Overrated, Australian Financial Review, 25 August 2009

According to the betting markets, NSW Premier Nathan Rees is currently the fourth-favourite to lead his party to the next state election. If the odds are to be believed, Frank Sartor, John Della Bosca and Carmel Tebbutt are all more likely to lead Labor than Rees.

Just as with prior leadership battles, the temptation for the commentariat will be to follow each twist and turn, analysing the strengths and weaknesses of every contender. There is little doubt that political parties are important. But rarely do we ask whether it matters who leads a particular party. Should we care which jockey rides the horse?

One of the strongest proponents of the leadership-is-overrated school of thought was Leo Tolstoy, who argued that ‘great men – so called – are but labels serving to give a name to the event’. Yet as a plethora of political biographies attests, the view that individuals matter remains powerful. Isn’t it tempting to believe that individuals matter more than mere political parties and the nameless forces of history?

Perhaps the most tantalising version of the leadership-is-everything perspective is counterfactual history. If Harold Holt had not drowned at Cheviot Beach on 17 December 1967, would he have pursued a different set of policies from his successors? If Queensland Senator Bertie Milliner had not suffered a heart attack on 30 June 1975, might Gough Whitlam have avoided the Dismissal, and been able to do more with his time in office? If Neville Wran had not been handicapped by a throat operation in June 1980, would he have entered federal politics before Bob Hawke, and been an even more successful leader? The more one reads about the players in these events, the easier it is to be seduced into believing that they fundamentally changed the course of Australia.

In an intriguing article, Ben Jones (Northwestern University) and Ben Olken (MIT) have set about trying to estimate the impact of leadership on national outcomes, such as economic growth and fiscal policy. Ideally, we would like some random variation in leadership turnover, but Jones and Olken use the next best thing: instances in which leaders die from accidents or natural causes. In effect, they ask: did the deaths of John Curtin, Gamal Abdel Nasser and Franklin D. Roosevelt change their nation’s trajectory?

The answer, according to Jones and Olken, is that Tolstoy was half-right. Leaders matter in autocracies, but not in democracies. The deaths of Ayatollah Khomeini and Mao Tse-Tung were followed by rapid improvements in living standards for ordinary Iranians and Chinese. But in a typical democracy, economic outcomes are unaffected when a leader passes away. Constrained by political parties, institutions and interest groups, democratic leaders have less scope to change the world than the typical biography (or autobiography) might have you think.

Admittedly, these results do not rule out any leader effects whatsoever. It could be that while most leaders are interchangeable with the other grey suits in their party, a handful stand out from the crowd. Or maybe leaders have little impact on economic growth, but nonetheless leave their mark on foreign policy, public infrastructure, or the configuration of the public service.

Yet when we flip the question around, it turns out that the economy has a big impact on how long leaders stay in office. Caucus members may think that they are judging the contenders on merit, but one of the best predictors of whether a party will turf out its leader is the national growth rate. Just as the early-1990s recession saw the UK Conservative Party oust Margaret Thatcher in favour of John Major, and the Australian Labor Party remove Bob Hawke in favour of Paul Keating, so too the latest slump has already caused plenty of heads to roll.

Indeed, empirical research by Australian National University PhD student Paul Burke suggests that the global downturn will lead to an additional 11 national leaders being removed from office. Conversely, as Peter Costello can attest, economic good times are bad news for a leadership pretender.

The impact of the economy on leadership turnover reflects what psychologists have termed ‘the fundamental attribution error’: our tendency to overestimate the role of individuals, and underestimate the importance of luck and context. Comparing Nathan Rees with other Australian Premiers in the last decade, it is difficult to escape the conclusion that if he had been elevated to the Premiership in a boom time, Rees might now be regarded as a decent performer. Leaders may not affect the economy, but the economy sure affects them.

Andrew Leigh is an economist in the Research School of Social Sciences at the Australian National University.

Update: Murray Goot emails me to suggest that the media’s focus on individual leadership may have been partly a result of the public reporting of leadership polls, initiated by Morgan in 1967.

Advertisements
This entry was posted in Economics of Elections. Bookmark the permalink.