Who The Heck Is Kenneth Arrow?
(written on Monday Dec. 13 but not posted 'til Thurs. 12-16-04):
Who The Heck Is Kenneth Arrow?
A defense of the Electoral College
Okay, today, Monday, December 13, 2004 is Election Day. Election Day for the President Of The United States, that is--admittedly not for the hundreds and hundreds of other offices which had their Election Day several weeks ago, on Tuesday, November 2nd. Those offices are important, too, indeed collectively speaking even moreso than the Presidential election that so many here in the US and elsewhere around the world were fixated upon for the weeks, months, and even years leading up to that earlier date.
However, TODAY, Monday, December 13 2004, is the REAL date for the US Presidential election, the REAL date in a true and meaningful sense and not merely in a manner more apropos to the board game "Trivial Pursuit." To be sure, in most years the Electoral College is something of a formality, a final affirmation of the 51 separate decisions made in early November when the members of said EC were chosen. In some years, such as 2000, the super-slim margin of victory in terms of "pledged electors" that George Bush had over Al Gore (and over the "victory bar" of 270 EVs) made the actual voting process such a newsworthy event that it was one that was covered live by even network TV, but in most years the actual election of the US President is so uneventful that when it happens most Americans don't even know that it's Election Day.
Over the years, especially since the 2000 Presidential election, dozens and dozens of arguments have been made both for and against the Electoral College, most of them concentrating on (in the case of those opposed to the EC) the occasionally-anomalous discrepancy between the national popular-vote total in the Presidential race and the final winner as voted upon by the 538 real voters for President, or (in the case of those who support the Electoral College) the need to appeal to small-state populations, the disincentive to engage in fraud in states likely to produce landslide totals for one candidate or the other, etc. etc. etc. All of those are good and important arguments, but they miss out on THE MOST CRUCIAL ASPECT of the present method of electing the President, namely, the requirement that the winner of the race receive not only the most votes but RECEIVE A MAJORITY OF THE VOTES CAST.
This characteristic of the Electoral College, that is, the requirement that the winner receive a majority, is almost always overlooked in discussions of this system but is in fact the linchpin of the entire procedure. This is so crucial that it is enshrined not only in the US Constitution's specific procedures for Presidential (and Vice-Presidential) selection, it is even found in its procedures for what would happen were that proecedure not to produce an outright winner. That is to say, should the Electoral College's Electors (the real Presidential voters) not produce a majority for any one candidate, the US House of Representatives, voting not as individuals but as state delegations, shall elect the US President, with the similar "majority" requirement--that the winner receive a majority of the state delegations so voting (not simply a pluralty, but a majority). The US Vice-President would be similarly elected by the US Senate, with the same caveat--the winner has to receive a majority to be elected US Vice-President. Even further down the line of Presidential succession, at least as the laws are currently written, the same process applies--the Speaker Of The House and President Pro-Tempore of the Senate are (I believe) elected by majority-requirement votes in their respective bodies, and the Cabinet officials next in line need to be confirmed (as I understand it) by a majority vote of the US Senate. That is to say, (barring for the moment a discussion of a Presidential "recess appointment" to a Cabinet position which would be in the line of Presidential succession) {let's call this 'Yoest's observation'} "no person can become US President without the election by or approval of a (non-trivial) majority of 'substantial governmental persons' be those persons (in order) Presidential Electors, US House members, or Senators." (Readers, I'm on a roll here--let's leave any discussion of Alexander Haig out of this for the time being, okay?)
Alright, so we've established this whole majority thing, so what's the big deal about that? Or, more specifically, when is this post going to get around to answering the question "Who the heck is Kenneth Arrow????"
Who the heck is Kenneth Arrow, anyway????
Good question, and more people need to know who he is and why his work is so significant for the Electoral College and any serious discussions of its merits or lack thereof.
Kenneth Arrow was not a "Founding Father," nor was he in any direct or indirect way an influence on the US Constitution or any of its amendments, but his influence on our understanding of our Constitution and its 12th Amendment in a certain sense makes him an "honorary Founding Father" (after the fact, nonetheless). He was born in 1921 (and is still alive, which means he's really an honorary Founding Father and not a real one!), and his influence on discussions of the Electoral College are more the result of the content of his book "Social Choice And Individual Values" (published in 1951, for which he won the 1972 Nobel Prize), said content now more commonly known as "Arrow's Impossibility Theorem." This theorem, and our Presidential election system's (systems') ways of avoiding it, are the real reasons why we should defend the Electoral College, and serve as some sort of prolegomena for any future discussions of its relative strengths and weaknesses.
Great, good, ok, now we can move on, right? Well, sure, yes, from here we can, and that would be a good thing, but maybe it would be even better if we discussed briefly what this "Arrow Impossibility Theorem" actually is, and exactly how it justifies our Electoral College. Does anyone out there actually know what the Impossibility Theorem actually says? If so, please stop reading right now, because the description I'm about to give of it is very bad--even "wrong" in a way, but for purposes of Electoral College discussions it'll do. One can quickly go to http://en.wikipedia.org/wiki/Arrow%27s_impossibility_theorem and get a good description of it, the essence of which (copied and pasted from that Wikipedia entry so linked, with snippage of internal subdesriptions) goes like this:
(from Wikipedia):
"The theorem’s content, somewhat simplified, is as follows. A society needs to agree on a preference order among several different options. Each individual in the society has a particular personal preference order. The problem is to find a general mechanism, called a social choice function, which transforms the set of preference orders, one for each individual, into a global societal preference order. This social choice function should have several desirable ("fair") properties:
unrestricted domain or universality: ...{snippage}...
non-imposition or citizen sovereignty: ...{snippage}...
non-dictatorship: ...{snippage}...
positive association of social and individual values ...{snippage}...
independence of irrelevant alternatives :...{snippage}...
Arrow’s theorem says that if the decision-making body has at least two members and at least three options to decide among, then it is impossible to design a social choice function that satisfies all these conditions at once."
(highlight this description as I'm going to refer to it again in several more paragraphs)
OK, now, great so we know what it says, but what does that mean? What does that really mean, that is, in the "real world" what are the actual, real, meaningful implications of that statement? Let me explain by giving one simple example, an example that we've all likely experienced in some fashion similar to the one about to be described. Let's take a multi-candidate election, Adams, Baker, Curtis, Davis, and Evans, all running for the same office. How do you pick a winner, that is, who gets the office? More specifically, what type of election system do you use to pick that winner? How about "most votes," let's try that. Great, what if the winner gets 21%? That means 79% of the people(!) voted against the "winner." How about "runoffs," where either the top two enter a "second round" or one where in each round the lowest votegetter is eliminated? Cool...until you realize that everyone's favorite "compromise candidate" may end up getting eliminated in the runoff(s). Well, then, let's try "ranking" the candidates in each voter's preferential order. Fine, until the, say, Baker voters realize that their selection of their real "second choice" (say, Adams) may elect her instead their first choice...so they engage in "insincere voting" by making Davis their "second choice" and electing him instead. What system can you use????
Arrow's Impossibility Theorem proves that there are NO systems that you can use!!!!!! NO system can produce a result that can be unequivocally said to be "fair," can avoid all of the dilemmas that are readily observable in the above simple example (let's ignore Condorcet for the moment). What, then, IS the way out of this, if in fact there is one????
Well, there is a way out--sort of. Remember what I wrote (copied & pasted from Wikipedia above, actually) about when the dilemma actually rears its ugly head. Go back and refer to that, as I'm not going to re-paste it again here (upon closer examination it reveals the "flaw" in my argument and I don't really want to do that, okay gang out there in Blogosphere???) THE ELECTORAL COLLEGE GETS US OUT OF THE PROBLEMS IDENTIFIED BY ARROW!!!!!! That is so significant that it needs to be said again: THE ELECTORAL COLLEGE GETS US OUT OF THE PROBLEMS IDENTIFIED BY ARROW!!!!!! YES, it's an inelegant solution, YES it's "unfair" in a certain and admitted sense, and I do admit that and if you haven't figured it out yet good just ignore my "protesteth too much" verbiage of the last several sentences. HOWEVER, I didn't write all of this simply to diss(miss) my own argument here!!!!! Let me offer this defense--the "unfairness" of the Electoral College is "prior" to the whole election process, prior to the time when we actually vote for President (this year it was December 13th, although one can apply my defense to even the 51 Presidential semi-elections we had on November 2nd. That is the true beauty of the Electoral College, its avoidance of all of the problems associated with Arrow's Impossibility Theorem. Something not even the Founding Fathers likely even realized :)
Let me phrase the above a little more simply. Imagine all of the problems associated with the Adams/Baker/Curtis/Davis/Evans election listed above, all that you can think of--"insincere voting," runoff paradoxes, massively-minority-elected "winners," etc. Furthermore imagine that that Adams/etc. election were a US Presidential election, without an Electoral College-type of system. Now furthermore imagine that one of those candidates, say Baker, were to win, on the first ballot, an outright majority of all the votes cast. See how all of the paradoxes/dilemmas/"unfairnesses(!)" now magically disappear??? Isn't that neat!!!!!! Doesn't that solve all of our problems!!!!! Well, okay, sort of but no, not totally, not necessarily, BUT, hey, it's a GREAT START and is FAR, FAR SUPERIOR TO ANY AND ALL ALTERNATIVES THAT HAVE BEEN SUGGESTED!!!!! At least any that I've seen, anyhoo. So this Electoral College and, more importantly, its "majority requirement" are what makes our Presidential election method so good, so long-surviving, and is why we all need to know who the heck Kenneth Arrow is, to answer that question at long last :)
That's really the end of this essay, and I've made my point and can really stop here, but anyone who wants to think more about this issue should really put some thought into two recent US Presidential elections. There are two recent US Presidential elections which highlight the strength of the Electoral College system, and, believe it or not, neither one of them is the 2000 election!!!! They are in fact the 1968 and 1992 elections, the last two which had serious third-party contenders in them who almost certainly created "winners" out of "losers." Ironically, they also are similar in that they both gave us two of our country's worst Presidents--Richard Nixon in 1968 (when the third-party candidacy of George Wallace cost Hubert Humphrey the election) and Bill Clinton in 1992 (when the third-party candidacy of Ross Perot cost George H.W. Bush the election). With all of that said, I still will defend the Electoral College (and its "majority" requirement for election), even though I recognize that such problems can (and over time will) occur. (Adding to the irony, the only other election in US history when a third-party candidate altered the course of the election, at least since the Civil War, was in 1912, when Theodore Roosevelt's Bull Moose campaign gave the election to Woodrow Wilson, who is one of the few Presidents who can give Nixon or Clinton a serious challenge to the title "Worst President Ever." In my opinion Wilson was the worst, but I know that that point is debatable). So, if three times in 140+ years we get a clearly-and-demonstrably "bad" result from the Electoral College, well, to me maybe that's not too bad. At any rate it beats getting a "clearly-and-demonstrably-bad" result once every 4 years!!!!!!!! :-)
So, then, who the heck is Kenneth Arrow, and why should anyone care??? Now you know, and maybe you learned a little more about the Electoral College along the way :) All on a day when many people didn't even know there was a Presidential election going on!!!!!! :-)
Who The Heck Is Kenneth Arrow?
A defense of the Electoral College
Okay, today, Monday, December 13, 2004 is Election Day. Election Day for the President Of The United States, that is--admittedly not for the hundreds and hundreds of other offices which had their Election Day several weeks ago, on Tuesday, November 2nd. Those offices are important, too, indeed collectively speaking even moreso than the Presidential election that so many here in the US and elsewhere around the world were fixated upon for the weeks, months, and even years leading up to that earlier date.
However, TODAY, Monday, December 13 2004, is the REAL date for the US Presidential election, the REAL date in a true and meaningful sense and not merely in a manner more apropos to the board game "Trivial Pursuit." To be sure, in most years the Electoral College is something of a formality, a final affirmation of the 51 separate decisions made in early November when the members of said EC were chosen. In some years, such as 2000, the super-slim margin of victory in terms of "pledged electors" that George Bush had over Al Gore (and over the "victory bar" of 270 EVs) made the actual voting process such a newsworthy event that it was one that was covered live by even network TV, but in most years the actual election of the US President is so uneventful that when it happens most Americans don't even know that it's Election Day.
Over the years, especially since the 2000 Presidential election, dozens and dozens of arguments have been made both for and against the Electoral College, most of them concentrating on (in the case of those opposed to the EC) the occasionally-anomalous discrepancy between the national popular-vote total in the Presidential race and the final winner as voted upon by the 538 real voters for President, or (in the case of those who support the Electoral College) the need to appeal to small-state populations, the disincentive to engage in fraud in states likely to produce landslide totals for one candidate or the other, etc. etc. etc. All of those are good and important arguments, but they miss out on THE MOST CRUCIAL ASPECT of the present method of electing the President, namely, the requirement that the winner of the race receive not only the most votes but RECEIVE A MAJORITY OF THE VOTES CAST.
This characteristic of the Electoral College, that is, the requirement that the winner receive a majority, is almost always overlooked in discussions of this system but is in fact the linchpin of the entire procedure. This is so crucial that it is enshrined not only in the US Constitution's specific procedures for Presidential (and Vice-Presidential) selection, it is even found in its procedures for what would happen were that proecedure not to produce an outright winner. That is to say, should the Electoral College's Electors (the real Presidential voters) not produce a majority for any one candidate, the US House of Representatives, voting not as individuals but as state delegations, shall elect the US President, with the similar "majority" requirement--that the winner receive a majority of the state delegations so voting (not simply a pluralty, but a majority). The US Vice-President would be similarly elected by the US Senate, with the same caveat--the winner has to receive a majority to be elected US Vice-President. Even further down the line of Presidential succession, at least as the laws are currently written, the same process applies--the Speaker Of The House and President Pro-Tempore of the Senate are (I believe) elected by majority-requirement votes in their respective bodies, and the Cabinet officials next in line need to be confirmed (as I understand it) by a majority vote of the US Senate. That is to say, (barring for the moment a discussion of a Presidential "recess appointment" to a Cabinet position which would be in the line of Presidential succession) {let's call this 'Yoest's observation'} "no person can become US President without the election by or approval of a (non-trivial) majority of 'substantial governmental persons' be those persons (in order) Presidential Electors, US House members, or Senators." (Readers, I'm on a roll here--let's leave any discussion of Alexander Haig out of this for the time being, okay?)
Alright, so we've established this whole majority thing, so what's the big deal about that? Or, more specifically, when is this post going to get around to answering the question "Who the heck is Kenneth Arrow????"
Who the heck is Kenneth Arrow, anyway????
Good question, and more people need to know who he is and why his work is so significant for the Electoral College and any serious discussions of its merits or lack thereof.
Kenneth Arrow was not a "Founding Father," nor was he in any direct or indirect way an influence on the US Constitution or any of its amendments, but his influence on our understanding of our Constitution and its 12th Amendment in a certain sense makes him an "honorary Founding Father" (after the fact, nonetheless). He was born in 1921 (and is still alive, which means he's really an honorary Founding Father and not a real one!), and his influence on discussions of the Electoral College are more the result of the content of his book "Social Choice And Individual Values" (published in 1951, for which he won the 1972 Nobel Prize), said content now more commonly known as "Arrow's Impossibility Theorem." This theorem, and our Presidential election system's (systems') ways of avoiding it, are the real reasons why we should defend the Electoral College, and serve as some sort of prolegomena for any future discussions of its relative strengths and weaknesses.
Great, good, ok, now we can move on, right? Well, sure, yes, from here we can, and that would be a good thing, but maybe it would be even better if we discussed briefly what this "Arrow Impossibility Theorem" actually is, and exactly how it justifies our Electoral College. Does anyone out there actually know what the Impossibility Theorem actually says? If so, please stop reading right now, because the description I'm about to give of it is very bad--even "wrong" in a way, but for purposes of Electoral College discussions it'll do. One can quickly go to http://en.wikipedia.org/wiki/Arrow%27s_impossibility_theorem and get a good description of it, the essence of which (copied and pasted from that Wikipedia entry so linked, with snippage of internal subdesriptions) goes like this:
(from Wikipedia):
"The theorem’s content, somewhat simplified, is as follows. A society needs to agree on a preference order among several different options. Each individual in the society has a particular personal preference order. The problem is to find a general mechanism, called a social choice function, which transforms the set of preference orders, one for each individual, into a global societal preference order. This social choice function should have several desirable ("fair") properties:
unrestricted domain or universality: ...{snippage}...
non-imposition or citizen sovereignty: ...{snippage}...
non-dictatorship: ...{snippage}...
positive association of social and individual values ...{snippage}...
independence of irrelevant alternatives :...{snippage}...
Arrow’s theorem says that if the decision-making body has at least two members and at least three options to decide among, then it is impossible to design a social choice function that satisfies all these conditions at once."
(highlight this description as I'm going to refer to it again in several more paragraphs)
OK, now, great so we know what it says, but what does that mean? What does that really mean, that is, in the "real world" what are the actual, real, meaningful implications of that statement? Let me explain by giving one simple example, an example that we've all likely experienced in some fashion similar to the one about to be described. Let's take a multi-candidate election, Adams, Baker, Curtis, Davis, and Evans, all running for the same office. How do you pick a winner, that is, who gets the office? More specifically, what type of election system do you use to pick that winner? How about "most votes," let's try that. Great, what if the winner gets 21%? That means 79% of the people(!) voted against the "winner." How about "runoffs," where either the top two enter a "second round" or one where in each round the lowest votegetter is eliminated? Cool...until you realize that everyone's favorite "compromise candidate" may end up getting eliminated in the runoff(s). Well, then, let's try "ranking" the candidates in each voter's preferential order. Fine, until the, say, Baker voters realize that their selection of their real "second choice" (say, Adams) may elect her instead their first choice...so they engage in "insincere voting" by making Davis their "second choice" and electing him instead. What system can you use????
Arrow's Impossibility Theorem proves that there are NO systems that you can use!!!!!! NO system can produce a result that can be unequivocally said to be "fair," can avoid all of the dilemmas that are readily observable in the above simple example (let's ignore Condorcet for the moment). What, then, IS the way out of this, if in fact there is one????
Well, there is a way out--sort of. Remember what I wrote (copied & pasted from Wikipedia above, actually) about when the dilemma actually rears its ugly head. Go back and refer to that, as I'm not going to re-paste it again here (upon closer examination it reveals the "flaw" in my argument and I don't really want to do that, okay gang out there in Blogosphere???) THE ELECTORAL COLLEGE GETS US OUT OF THE PROBLEMS IDENTIFIED BY ARROW!!!!!! That is so significant that it needs to be said again: THE ELECTORAL COLLEGE GETS US OUT OF THE PROBLEMS IDENTIFIED BY ARROW!!!!!! YES, it's an inelegant solution, YES it's "unfair" in a certain and admitted sense, and I do admit that and if you haven't figured it out yet good just ignore my "protesteth too much" verbiage of the last several sentences. HOWEVER, I didn't write all of this simply to diss(miss) my own argument here!!!!! Let me offer this defense--the "unfairness" of the Electoral College is "prior" to the whole election process, prior to the time when we actually vote for President (this year it was December 13th, although one can apply my defense to even the 51 Presidential semi-elections we had on November 2nd. That is the true beauty of the Electoral College, its avoidance of all of the problems associated with Arrow's Impossibility Theorem. Something not even the Founding Fathers likely even realized :)
Let me phrase the above a little more simply. Imagine all of the problems associated with the Adams/Baker/Curtis/Davis/Evans election listed above, all that you can think of--"insincere voting," runoff paradoxes, massively-minority-elected "winners," etc. Furthermore imagine that that Adams/etc. election were a US Presidential election, without an Electoral College-type of system. Now furthermore imagine that one of those candidates, say Baker, were to win, on the first ballot, an outright majority of all the votes cast. See how all of the paradoxes/dilemmas/"unfairnesses(!)" now magically disappear??? Isn't that neat!!!!!! Doesn't that solve all of our problems!!!!! Well, okay, sort of but no, not totally, not necessarily, BUT, hey, it's a GREAT START and is FAR, FAR SUPERIOR TO ANY AND ALL ALTERNATIVES THAT HAVE BEEN SUGGESTED!!!!! At least any that I've seen, anyhoo. So this Electoral College and, more importantly, its "majority requirement" are what makes our Presidential election method so good, so long-surviving, and is why we all need to know who the heck Kenneth Arrow is, to answer that question at long last :)
That's really the end of this essay, and I've made my point and can really stop here, but anyone who wants to think more about this issue should really put some thought into two recent US Presidential elections. There are two recent US Presidential elections which highlight the strength of the Electoral College system, and, believe it or not, neither one of them is the 2000 election!!!! They are in fact the 1968 and 1992 elections, the last two which had serious third-party contenders in them who almost certainly created "winners" out of "losers." Ironically, they also are similar in that they both gave us two of our country's worst Presidents--Richard Nixon in 1968 (when the third-party candidacy of George Wallace cost Hubert Humphrey the election) and Bill Clinton in 1992 (when the third-party candidacy of Ross Perot cost George H.W. Bush the election). With all of that said, I still will defend the Electoral College (and its "majority" requirement for election), even though I recognize that such problems can (and over time will) occur. (Adding to the irony, the only other election in US history when a third-party candidate altered the course of the election, at least since the Civil War, was in 1912, when Theodore Roosevelt's Bull Moose campaign gave the election to Woodrow Wilson, who is one of the few Presidents who can give Nixon or Clinton a serious challenge to the title "Worst President Ever." In my opinion Wilson was the worst, but I know that that point is debatable). So, if three times in 140+ years we get a clearly-and-demonstrably "bad" result from the Electoral College, well, to me maybe that's not too bad. At any rate it beats getting a "clearly-and-demonstrably-bad" result once every 4 years!!!!!!!! :-)
So, then, who the heck is Kenneth Arrow, and why should anyone care??? Now you know, and maybe you learned a little more about the Electoral College along the way :) All on a day when many people didn't even know there was a Presidential election going on!!!!!! :-)
posted by Greg Yoest at 1:16 PM
0 comments
