Random Promotion at US Universities?

Perhaps this provides part of the explanation of random promotion policies (equivalent to random wages) used by some universities.

Is this true? Do some US universities really promote faculty staff at random?

I came across this quote in an old (1982) paper by the famous economist Joseph E Stiglitz with the intriguing title

Utilitarianism and horizontal equity: The case for random taxation,”

which can be downloaded for free from http://ideas.repec.org/p/nbr/nberwo/0694.html.

It is not, you’ll be pleased to hear, advocacy for the insane policy of random taxation — unless you consider the old Roman punishment of Decimation to be a Good Idea!

Rather it is a thought experiment, with some interesting implications.


12 Responses

  1. Ronald Coase died last week aged 102

    “The Theorem” could be read as an internal deconstruction of the law & econ school rather than a basis of it, one that points to a paradox at the heart of neo-classical economics.

    Makes me wonder what would “The Sortition Theorem” say?
    >>If we believe people are politically equal, then why have elections?<<


  2. >>If we believe people are politically equal, then why have elections?

    Because people have different preferences, beliefs, cognitions, ideologies, interests etc. and need to have some way of expressing these differences (hence the term “preference election”). Equality is not the same thing as homogeneity.


  3. *** At the question by AhmedR Teleb “If we believe people are politically equal, then why have elections?” Keith Sutherland answers: “Because people have different preferences, beliefs, cognitions, ideologies, interests etc. and need to have some way of expressing these differences (hence the term “preference election”)”.
    *** Elections of representatives and referendums are the two main ways of “expressing these differences” in our polyarchies. We find that the results may be different, and very different; in the French referendum about the European Constitution, 55% voted NO, whereas a very big majority of elected representatives favored YES. Therefore at least one of the two ways of “expression” must be strongly wrong. I think both are.
    *** There are many differences among the citizens, along many dimensions. How to vote for Obama or for Romney could express such a diversity? And even if there are more parties, the problem remains.
    *** I have discussed the question “modern representative system vs democracy-through-sortition” with different people with a university level training. Through the discussion many of the opponents to democracy-through-sortition ended by rejecting the idea of political equality itself (= the voice of any citizen must be given the same value in sovereign political decisions). Something as: “you want to give the power to morons”! Well, my own experience has not a great statistical value. Had my fellow kleroterians a different experience when discussing the subject?


  4. Ahmed

    What exactly do you mean by the phrase “we believe people are politically equal” — is this an empirical observation or a normative aspiration? The former claim is clearly false — ante-natal (genetic) inequalities are subsequently compounded or trumped by post-natal (environmental) inequalities. So much for Franklin’s “self-evident” amendment to the Declaration of Independence, but Jefferson’s first draft was equally mistaken as God is (supposedly) disinclined to afford equality between the sheep and the goats, consigning the latter to eternal damnation. So that leaves equality as a normative aspiration — if so then, in a political context, equality can ONLY be achieved via voting. In preference elections everyone is equally impotent to determine the outcome on account of the numbers involved, but a modestly-sized allotted assembly can implement substantive political equality iff it is limited to voting, as this equality will be destroyed by the inequality of individual speech acts. The resultant equality is infinitely more fine-grained than Andre’s stark choice between Romney or Obama — that’s why my preference is for policy generation via e-petition followed by public votation and allotted deliberation. Note that this requires THREE different stages of voting as this is the only way of achieving political equality. We must be careful not to conflate voting with elections for persons or political factions.


    I think the scepticism that you have uncovered in your discussions is a result of the assumption that an assembly created by sortition should have an open mandate. Many educated people are aware of the burgeoning literature on the aggregate wisdom of crowds, but would naturally baulk at the prospect of being subject to the rule of random individuals (at least I certainly would!).


  5. If a joke needs to be explained, it was probably not funny to begin with.
    My comment was about ECONOMIC arguments for sortition, à la Coase.

    Such an argument would be based on transaction costs, information asymmetries, moral hazards, waste, etc…

    Here we go then, for instance, “The Kleroterian Lemmas”:

    1.1 Elections have unacceptably high transaction costs.
    1.1.1 The costs stem form the conflict between the publicity needs of the candidates and the “investigative” needs of the electorate.
    1.2 Being periodic they create a moral hazard for the elected official to use the time till next election for self-enrichment, power consolidation.
    1.3 Party-politics are a situation of information asymmetry between the internal party politics and public politics. The asymmetry being greater the fewer the political parties in a state.
    1.4 The relatively long period of time between elections and the socialization of the cost of government creates a large incentive for lobbying and corruption. That is, each tax payer feels a relatively small burden from each port-barrel subsidy but the subsidy recipient receives a windfall compared to its tax/lobby payment.
    *1.5 An election-only form of government creates large amounts of economic waste or tend toward plutocracy.

    This is half said in play, but the economic justifications of selection by lot certainly seem worth a closer look.


  6. “pork barrel” was, of course, the meaning


  7. Ahmed

    No-one would dispute your economic critiique of electoralism, but this does not add up to an economic justification of selection by lot (entirely absent from your five “Kleroterian Lemmas”). More specifically you fail to address two points:

    1. People have different preferences, beliefs, cognitions, ideologies, interests etc. and need to have some way of expressing these differences (hence the term “preference election”)

    2. You haven’t explained what you meant by your claim that people are politically equal (or is this the joke that your referred to?).

    PS I do wish people would respond a little more promptly — it’s quite hard carrying on a conversation when there is a two-week gap in between postings!


  8. My bad Keith on the slow reaction time, been busy teaching statistics!

    As far as my “political equality” assumption goes, the exact meaning may be irrelevant because I meant to use it as a starting point to point out inconsistencies in elections based as economic reasoning or simplifying assumptions. Here is a nicely done Wash Post article on “The Coase Theorem” that points out how simplifying assumptions can be used by opposing ideologies to justify opposing policies!

    For the record, political equality, as I used it meant “equally valid opinions/values with respect to policy” also implying the lack of “a priori” criteria for which political preferences or values or opinions should “win.” But, again, this is besides the point.

    Coase used the simplifying assumption of “zero transaction costs” to point out contradictions in economic theory. It was later used both by “right wing” law and econ people to justify non-government interference; and also used by “left wing” regulation advocates to support more government intervention in economic matters.

    It seems you and Andre demonstrated that it could be used to justify or critique election or selection by lot.

    Andre, “giving the power to morons” has also been, in my experience, the first reaction here in the US. But people tend to change their mind one they examine the status quo. But this again illustrates the differences between “voting” based on “a priori” preferences/prejudices/opinions and discourse/deliberation. Which is perhaps why random selection in addition to, in palace of, only voting whether for representatives or particular policies, i.e. elections or referenda.


  9. Ahmed

    I like your definition of political equality — “equally valid opinions/values with respect to policy”. As this is clearly a normative desideratum rather than a statement of fact, the relevant challenge is how to turn the former into the latter. Sortition can only achieve this in the act of voting, where every opinion/value has a numerically equal value (1). Not so for deliberation where enormous inequalities are introduced, derived from differentials in perceived status, knowledge and persuasive powers. This is normatively unproblematic when each deliberator is only speaking (or not, as the case may be) for herself, but rules out deliberation in an allotted legislative assembly whose mandate is purely one of statistical representativity. To put the problem simply — what if the allotted persons whose opinions/values are closest to your own opt to remain silent, or perform unpersuasive speech acts? (for the reasons given above). Such an outcome would be entirely unequal and unrepresenative and would not command the support of the vast majority who have been disenfranchised by the replacement of preference elections by sortition (as advocated by radical activists on this forum).

    I don’t think it’s sufficient to point out comparative inequalities in existing electoral arrangements — the onus is on those who are proposing an alternative (sortition) to understand exactly what it can and cannot do. The inherent inequalities in speech acts indicate that democratic equality requires a combination of sortition and other methods (preference elections and/or direct democracy) in order to establish a truly representative legislature.

    As you’ve been teaching statistics I would very much appreciate your response on this point. It strikes me as self-evidently true, and I’m puzzled why I have to keep making the argument on this forum. Non-sortinistas tend to grasp it immediately, so it strikes me that sortinistas are confusing the epistemic case for deliberation with the need for democratic equality.


  10. I would not venture to call myself an authority on statistics by any means. Teaching it, at an introductory level, has only taught me that mathematicians and statisticians see the world entirely differently. For example:
    1/3, 100/300, 33.33..%, etc., are equivalent mathematically but to a statistician the DENOMINATOR is crucial. The size of the sample is crucial to the meaning/significance of the measure, and furthermore, grouping stats differently can give entirely different results!

    So, on statisticss I will defer to Yoram and other real statisticians.

    As for the rest, between deliberation and preference voting I see a QUALITATIVE difference that I want to think about more before adding anything meaningful to what’s already been debated here.

    About persuasive abilities in a deliberative body, I believe Hannah Arendt would say: Yes, and? So what? Isn’t that what being human is about, differences between persuader and persuaded? Isn’t freedom to change one’s own mind just as important as the freedom to change someone else’s?


  11. I love this discussion, but there’s still no answer to my original question: Stiglitz asserts that US universities promote at random — do they?

    (Or does he mean automatic promotion by length of service, and would that be considered as ‘random’? — you know how sloppy some people can be with the use of the word ‘random’!)


  12. Ahmed

    Arendt is certainly correct about the freedom to persuade and to be persuaded, but this is from the perspective of the participating individual. The problem is when you try to combine individual freedom with representativity — what you end up with is randomness in the pejorative sense, whereas statistical representativity presupposes a lawful relationship between the sample and the target population. Deliberative democrats do indeed privilege quality over quantity, but they are not concerned with representativity — the sine qua non for democratic decision making in a polis where it is not possible for everyone to participate directly.


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.

%d bloggers like this: