By Zachary Hayward
Faceless men, backroom deals, leadership spills. The cliches of politics have been taken apart every election cycle and rearranged in some new dispiriting configuration. We’re left wondering why our leaders seem out of touch and bereft of new ideas. Even more puzzling is that this occurs under our preferential voting system. From Amartya Sen and Eric Maskin (Both Nobel Prize winners) to Hasan Minaj and John Oliver, preferential voting and impartial drawing of electoral boundaries are the holy grail of electoral efficiency. Yet, here in Australia, we have both of these things and yet our politicians continue to disappoint. They fail to handle a pandemic, to combat climate change, and deal with a toxic culture in the parliament. Why is this? Have we failed in some way, or does the simple economic model of government have more serious problems?
At this point you might be wondering if this is going to be yet another article claiming that democracy is irredeemably flawed. This is not the claim I make. Arrow’s impossibility theorem, the idea that there is no coherent method to aggregate the preferences in a society wide social welfare function, is an exceedingly fragile result and one that has been effectively challenged by those like Amartya Sen (who found other serious problems). There is also little evidence that the types of strategic voting that preferential voting systems are vulnerable to are prevalent. That is, our system of voting is distortionary but much less so than plurality (first-past-the-post) voting, or electoral college type systems.
Rather, it is that the preselection process systematically chooses inferior candidates that minimise the quality of Australian political decision making. There is some evidence that politicians’ voting behavior is more than 80% correlated with the party whip and that the margin of victory has little if any effect on voting behavior . The authors of this study note that they do not take into account the effects or preselection. Only that there may be pressures that force politicians to toe the party line.
It should come as no surprise that political decision making often poorly approximates the views of the public. The selection mechanism we rely on, public elections does not seem to affect policy or at least voting at all. Rather, I argue that preselection dictates policy because the candidates selected are already predisposed to represent the views of the party branch, rather than those of the electorate. This mechanism operates with little public scrutiny. An opaque selection process with questionable efficacy actively selects against diversity and as a result degrades the quality of our entire political system. How can we have a representative democracy if the choices are simply between two white university educated men?
The charge against this is of course that this is meritocracy at work. However, this is almost laughable when one even glances at the experiences of nearly all women in politics (Julia Banks comes to mind) and epistemically unsound. Helen Landemore, a political philosopher at Yale argues that cognitively diverse groups (even with lower average intelligences) outperform cognitively homogeneous ones with high average IQs. This is to say, even if we believed that our parliament contained the brightest minds available to us it is still inferior to one that actively selects for diversity. Thus, the meritocratic argument put forward by politicians is both unlikely to be true and still likely to be selecting candidates on the wrong set of characteristics.
However, there is a second reason that this lack of diversity is persistent. The preselection process gives too much power to individuals and groups with vested interests. In both major parties, power brokers have opaque powers with very little public scrutiny possible. Sometimes it is not even clear who is responsible for the decision to endorse a candidate. In all parties, there are power brokers, unions, and lobbyists. These groups are themselves often homogenous and without public oversight, and there is no way for this to be corrected. We occasionally hear about the evils of branch stacking, but this is the system working as intended . The problem is the preselection process itself.
The reason why this is so damaging is because of the selection effect on politicians. Even though a candidate may face public scrutiny at the election (presumably it requires some merit to win an election), the lack of diversity at the preselection level implies that the pool of candidates is ultimately restricted to those who have a reasonable chance of being elected. Ultimately those selected by the major parties are the most likely to win. The consequences of this are that institutional decision making is impaired. Helen Landemore argues that cognitive diversity introduces negative correlations between decision makers, as compared to a group of decision makers with no correlation between their views. In the zero-correlation case, attitudes of decision makers are vulnerable to large distortions due to random chance. However, the negative correlations ensure that on average the correlations will effectively cancel out.
It is easy to believe that there is no solution to this problem, that institutional capture has simply gone too far and that there is no way to correct this systematic selection bias. However, there is a simple selection method that ensures that in a sufficiently large enough sample (say the entire parliament) we can generate a representative sample of the population. At a branch level parties should make a credible commitment to choose candidates at random. This could be conditional on some light ideological screening requirements. However, after this significantly reduced hurdle, the actual selection from the pool of candidates must be random. One objection to this is the idea that people may self-select into applying, however, credible commitments to choose at random would certainly open the field to all candidates who are interested. The indirect advantage of this is that it reduces the chance of members serving multiple terms, which may be a good thing.
Sometimes, removing decision making power from individuals who “know what’s best” increases the chances of a successful outcome. Indeed, it is better that nobody chooses who is endorsed to run for parliament because it will increase the options available to the public. The past focus on “electability”, “likeability” and “image” has only served to entrench the interests of one group, white straight males.
The most serious threat to the quality of our democracy is the stranglehold of preselection on the candidates at each election. However, we cannot destroy the parties themselves. The ideological structure they provide is useful. Rather they must work to increase the cognitive diversity of their candidates. This will more adequately represent the Australian people at large, but also greatly improve institutional decision making by avoiding correlated errors and increasing the oversight of elected officials. Randomizing our politicians is ironically the least chaotic option.
 Lee, David S., Enrico Moretti, and Matthew J. Butler. “Do Voters Affect or Elect Policies? Evidence from the U. S. House.” The Quarterly Journal of Economics 119, no. 3 (2004): 807-59.
 Branch stacking is when the membership of a regional branch is artificially inflated with “fake” members who occasionally do not even know they are part of the branch to support a particular candidate. Examples are common. They often have the implicit support of the party leadership.