Beating the Cambridge Analyticas
27 March 2018
Robin Cohen is Senior Research Fellow at Kellogg College and Emeritus Professor of Development Studies at the University of Oxford. In this blog he writes as a concerned citizen rather than someone with a particular expertise in electoral systems. This blog post was originally published on openDemocracyUK, you can read the original post here.
Beating the Cambridge Analyticas: change the way we (s)elect our representatives
Whatever it’s called legally, this is electoral fraud and we need to stop it. Here’s my suggestion – let’s adopt sortition (selection by lot) together with what I will call ‘filtration’ as our preferred electoral system.
I use ‘Cambridge Analyticas’ in the plural because the company has been active in 200 elections in 40 countries. It spins off subsidiary companies and hides under parent companies. A particularly indiscreet executive may be ‘fired’, only (this is a safe bet) to surface somewhere else doing much the same thing.
Besides there are other companies engaged in, or capable of, data mining for the purposes of subverting our electoral processes. The law is miles behind the curve and no doubt there are plenty of regulations that can be passed limiting electoral expenditure on data mining and forcing open disclosure. These will help, but such interventions are too late.
Huge volumes of data have already been captured and (this is a safe bet) secretly stored. But the key problem is that so long as there is a known and limited list of candidates for election, there will always be a nimble computer scientist able to outwit and outrun the clunky reach of the electoral law.
We need to change the game to outsmart the Cambridge Analyticas. The best way to do this is vastly to increase the pool of candidates, then select our representatives by lot. This is technically known at sortition, or demarchy.
The system has many advantages, but in this context it would make digital electoral fraud uneconomic and, indeed, pointless. Sortition was, as classicists know, widely used in ancient Greece. As Aristotle put it: ‘It is accepted as democratic when public offices are allocated by lot, and as oligarchic when they are filled by election’ (Aristotle, Politics IV. 9, 1294b8). Allocation of offices by lot now gets an increasingly respectful hearing from political theorists. There is even a Sortition Foundation set up to promote the idea worldwide.
Here are just some of the advantages:
- – Corrupt, power-hungry or narcissistic politicians will be stopped in their tracks.
- – A public service ethic will be enhanced.
- – As the pool of talent will be massively enlarged there is a good chance that we will get better public servants compared with those who are self-selected or supported by special interest groups.
- – Party loyalties will be diminished, limiting block votes (whipping) and encouraging individual judgement.
- – Given random selection, our representatives will be more representative of the population. There will be no need for women-only shortlists or positive discrimination to include hitherto excluded minorities.
There are two obvious objections to sortition. First, unsuitable people might be drawn by lot. Second, it is generally acknowledged in the literature that the system works best in small communities or assemblies. How damaging are these objections?
The simplest way to respond is to sketch a short scenario as to how sortition might work, taking a big unit such as England as an example. Regarding ‘unsuitability’, the Greek system survives in jury service where ordinary citizens, chosen by lot, make important and complex decisions on crime and punishment. (In earlier times these were life and death decisions.) Yet, the UK’s jury system is widely admired and emulated.
The issue of scale can be addressed through successive rounds of filtration. Using a rough illustration from England, there are about 36 million people on the electoral roll. This is the biggest sample of potential candidates. Our first form of percolation draws on the Juries Act of 1974, which provides a few sensible exclusions (those with recent criminal convictions and those who are mentally incapable). This list will be rebadged the Citizen and Jury Roll, potentially allowing selection either to political office or to a jury.
The second and third form of filtration will be residence and length of office. Take selection to one of the civil parish councils in England. The population of civic parishes and urban wards ranges from a few hundred to tens of thousands but citizens eligible for selection by lot will be limited to those who have lived in the parish for (say) five years, cutting the 36 million pool to (about) half.
Their terms of office will be for one year. For illustrative purposes I will simplify England’s local government structure by assuming 5 tiers – described as parish/ward, town, district/borough, county/city and national. The terms of office will be 1, 2, 3, 4 and 5 years respectively and sequentially. To clarify, Citizen Politicians will be randomly selected for a town council only after they have served a year in a parish council and will only be eligible for selection to a district council having served two years in a town council, and so on along the system. The table below provides a rough guide to how the pattern of filtration will work.
|Tier||Term of Office||National Pool||No of Councils||No. of Councillors|
|Parish Council/Ward||1 year||18,000,000||12,000||80,000|
|Town Council||2 years||80,000||4000||40,000|
|District Council/Borough||3 years||40,000||2000||20,000|
|County/City Council||4 years||20,000||82||10,000|
|National Assembly||5 years||10,000||1||500|
When Citizen Politicians reach the National Assembly (the new House of Commons) they will have had ten years of experience dealing with problems and opportunities at all layers of society. However, the pool will be sufficiently large, even at the final level of filtration, to eliminate malicious data mining.
Naturally, there are many unanswered questions. Will there be recusals for medical or emergency staff or for those who are full-time carers? Yes, why not? Will Citizen Politicians be paid? Yes, at different rates at the different tiers. How will executives and cabinets be formed? Probably through slates of candidates elected by Citizen Politicians at each tier. There are other possibilities too. Will there be a Senate? Yes, either through further sortition or perhaps another form of (s)election.
My main point is not to answer every possible question about sortition, but to find a way to defeat the malign intent of those who seek to manipulate and undermine our democratic mechanisms and traditions. No electoral system is perfect and this is true of sortition. However, it is a whole lot better than what we have at the moment – an open goal for hoaxers and rascals.
 See, for example, Peter Stone ‘Sortition, voting, and democratic equality’, Critical Review of International Social and Political Philosophy, 19 (3), 2016, 339-56, DOI: 10.1080/13698230.2016.1144858.