Vote weighting

© Paul Cooijmans

Fine-tuning democracy

To improve the efficiency of elections of any kind I recommend vote weighting by intelligence.

Behind this idea is the assumption that intelligence is a good indicator of one's judgment. Also, intelligence is of all human properties that might be relevant in this respect the one that is the most reliably and accurately testable.

In practice this could be done by including a brief I.Q. test in the voting procedure, which especially in computerized voting is no problem at all. The weight to be assigned to each vote would be determined by one's score. To add more weight to votes from higher scorers and compensate for their rarity, I advise this formula (although even better ones may be devised later):

Weight = 1/(1 - proportion outscored) - 1

Test norms could be based directly on the voting population. Norming and vote counting would take place in one go. Universal peace would be around the corner.

An even better weighting system would use measures of conscientiousness and associative horizon, next to intelligence (or an aggregate verifiability/creativity/awareness score), but at this moment I can not offer such measures in a usable form. Also, a possible future society consisting of only or mainly highly intelligent, highly conscientious people, would not need vote weighting at all. Vote weighting should be seen as a temporary instrument in advancing toward a better world.

A few comments to this proposal are dealt with below.

Comments and replies

Comment: Vote weighting only deals with the voters and does not affect the actual political parties, politicians, party programs and policies.

Reply: It indirectly does affect all of those. When it becomes known there will be vote weighting, new parties will be formed with intelligent programs that would have been without chance in an unweighted system (and/or existing parties will adjust their programs accordingly). On the other hand, existing successful party programs may be without chance in the weighted system.

Comment: Some nazi leaders had very high intelligence, so that shows that intelligence does not guarantee good policies.

Reply: This comment is mistaken, as in my system it is the intelligence of the voters that is weighted, not that of the politicians or party leaders. So projected onto the German situation of the 1930s, it is the intelligence of the German people that is relevant, not that of the nazi leaders. Had the people's intelligence been weighted in the elections back then, it is thinkable the nazis would not have gained power. In this respect I point to the many German intellectuals who fled Germany in those days, many of them Jews and some of them great scientists, obviously well aware of the danger. Had their votes been given greater weight as I propose, Hitler might have been sent home early.

Second, that nazi leaders had high intelligence is not surprising, as leaders in any organization must have at least well above average intelligence; that is simply a practical requirement for leadership. What matters here is the intelligence of the masses who voted the nazis into power.

Comment: Will this weighting system not result in an advantage for the political left (liberalism, progressiveness, socialism, multiculturalism), considering that intellectuals typically promote leftism?

Reply: Again, it is the intelligence of the voters, not that of the leading figures, that is relevant in vote weighting. The proverbial leftist intellectuals are merely the leaders, the spokesmen of the left. It is well known that the voting population of left-wing parties consists largely of the lower social classes, such as the working class and various minorities, like immigrants. Those groups vote left-wing because they feel that left-wing parties will best humour their interests. Given what we know of the relation between social stratification and the distribution of intelligence in modern society (the first more or less reflects the second, and this is getting more and more so because of "social mobility", the phenomenon that the intelligent and able tend to rise to higher social classes than they were born in) it is not to be expected that vote weighting will give an advantage to the left.

As for the left-wing intellectuals themselves, it is not certain they actually vote left-wing. The leading figures of the left are usually not members of the lower social classes they represent, but are well-to-do and, once in the voting cabin, may well "vote with their feet" instead of putting their vote where their mouth is. This is reflected in the friendly Netherlandic maxim Links lullen, rechts vullen ("talk left, fill your pockets right")

Nevertheless, it must be admitted that "intelligence is not everything", and that eventually other personality aspects must be included in the weighting process too, which will further reduce the possible unfair advantage of leftist intellectuals.

A objection occasionally raised against the present reply is that it is not true that the lower social classes vote mainly left-wing, and that in actuality they vote more often right-wing. This strange notion is based on the current "far-right" parties with nationalist and anti-islamization agendas, which indeed draw many votes from the lower classes. However, upon studying the programs of such parties it quickly becomes obvious that they are, in the social-economical sense, left-wing rather than right-wing, and should properly be called "left-wing conservative". So, on the understanding that what is mistakenly called "far-right" is really left-wing conservative, it remains true that the lower social classes vote left-wing.

Comment: Everyone's interest is of equal importance so everyone deserves an equal vote weight.

Reply: The error that makes this a non sequitur is in the word "interest"; voting should not be done out of individual self-interest, but for what is best for the common good. Persons of higher quality will better be able to decide that. Who votes to humour one's own interest should not vote at all (and this comment reveals the commentator in question to be such a person). The matter whether everyone's interest is indeed of equal importance it interesting in itself, but not relevant in the current context for the aforementioned reason that voting should not be done out of interest to begin with.

Comment: Some think that I.Q. fluctuates throughout life; is it therefore not unwise to base one's vote weight on one random recording of one's I.Q.?

Reply: The article states the weighting is to be done by including a brief I.Q. test in the voting procedure; so that is not one random recording, but repeated every time one votes. The question therewith disappears, and the related question whether or not I.Q. fluctuates becomes irrelevant in the present context.

Comment: Making good decisions depends not only one intelligence, but also on one's momentary emotional state. How do you take that into account?

Reply: That is to some extent implicitly dealt with in the vote weighing method. Persons of higher intelligence are on the whole - though this correlation is not perfect - less likely to base their decisions on emotion, while those of lower intelligence are more emotion-driven. So, vote weighting by intelligence should reduce emotion-based decisions in voting to some extent at least. The assumption here is that reason-based decisions are better than emotion-based ones.

Comment: I fear that social peace is not promoted by factually excluding so many people from decision making. How will they (the lesser intelligent) feel about decisions made by the more intelligent?

Reply: People are not at all excluded, but included with their respective weights. Technically, one does not even know one's exact weight, in fact one can theoretically boost one's weight by doing well on the test, so one has some control over it, which is psychologically beneficial and healthy. No one needs to feel excluded from decision-making. Rather, one may feel empowered and in control.

Comment: With this weighting formula, in a country with 10 million voters the lowest scoring 7.9 million would have only 5 % of the total vote weight, and the top 1500 have over 50 %. Corruption and the like would make this system fail.

Reply: While those numbers may be correct in theory, and seem fair given the importance of intelligence as well as a tremendous improvement over unweighted voting, one should consider that a test that is practical for this purpose will not have a high ceiling like that. Such tests end up going no higher than about the level of 1 in 100 to 1 in 200 with any validity (actually, the validity of any I.Q. test at all beyond that is still doubtful) so the highest weight will be shared by 50 000 or so voters out of 10 million, who will together have a majority of the total voting weight. This will still be a great leap forward compared to unweighted voting. Also, corruption regarding the test, just as any form of corruption, must simply not be tolerated. To abstain from implementing a good system out of fear of corruption would mean to give in to the bad, which one should never do.


A "system of plural voting based upon intelligence indices" is also mentioned by Lewis M. Terman in "The Great Conspiracy or the Impulse Imperious of Intelligence Testers, Psychoanalyzed and Exposed by Mr. Lippmann", New Republic 33 (December 27, 1922): 116120. See .