← LSE: Public lectures and events
LSE: Public lectures and events
Crises in Democracy: constituency re-districting and gerrymandering in the UK and US
01:26:56
Loading summary
Transcript37 lines
- [00:00]
A
Ladies and gentlemen, I think we should start on time because we have to be out of here by 8 o', clock, not later then. My name is Moshe Macruvert, I'm a CO director of the Voting Power and Procedures Group at the LSE and I welcome you on behalf of of our group. This is the inaugural public event of our second three year project funded by the Liberium Trust, to which I.
- [00:45]
B
Forward.
- [00:45]
A
My thanks for enabling this project and this event to take place. The project this Time is not purely theoretical as was the first one that terminated a couple of years ago. The title of the project is Voting Power in Practice and I did hardly explain the practical importance of the whole business of voting. It is not by any means a purely theoretical question.
- [01:30]
B
And.
- [01:34]
A
What we are doing during this project is trying to tie up the theory of social choice decision making by vote with practice. And this act, this tie up, this interface is actually encapsulated and exemplified by our two speakers whom I would like to introduce very briefly. I mean to introduce them fully would take more than we have time for. The whole event on our right, on my right is Professor Ian McLean who is professor of Politics, Oxford University and has the other quaint title, fairly quaint title of official Fellow in Politics at Nuffield College. What this actually means is that he is the senior person in social choice in this country. He is the author of a couple of dozen or co author and author of a couple of dozen books and a couple of hundred papers on social choice and on the history and I would say archaeology of social choice because he's unearthed fascinating discoveries from the unknown prehistory. But he's anything but an ivory tower academic. He's also dirtied his hands in actual politics. He's been a member and group leader of Oxford City Council and a member of Tyne and Weir Metropolitan County Council and has been an expert witness and advisor at various levels of government, from local government to parliamentary committees. So he's got more than a foot in the, in the practical camp. On my left, when I will sit down is, well, a complimentary, I would say complimentary figure. That's Sam Hirsch, who has kindly agreed to come from over the pond. He is a partner in Washington D.C. law firm of General Block, is a civil rights lawyer and has, I mean mainly engaged in voting rights litigation on behalf mainly of the Democratic Party and has represented cases in more than half of the 50 states of the United States and notably has been involved in litigation in the recent and very notorious gerrymandering case or alleged gerrymandering case in Texas, where the Democratic Party was the aggrieved party. But he's not only a practicing lawyer, he's also lectured at very prestigious law schools, including his alma mater, Harvard University. So I would invite, first of all, our first speaker to give us a talk mainly centered on the British side of the sub.
- [05:56]
C
Moshe, thank you very much and thank you for giving me this occasion to talk. And I hope that the tech guys will be able to tell me where my slides are.
- [06:12]
B
Right.
- [06:17]
C
Would people like the lights to be dimmed? Probably a good idea, I think, but not too dim. So that's who I am and that's what I'll be talking about. And here is the outline of what I want to say. That a fair allocation of districts would be one which gives each citizen as near as possible an equal share of voting power. And during the course of the time that I have, I hope to touch on the distinction very dear to the sponsors of tonight's lecture, the distinction between voting weight and voting power. I'm going to talk about the practical difficulties of implementing this, mostly about the uk, I'm going to say a little bit about the usa, but my colleague Sam Hirsch is going to say more, much more. I'm going to try to establish that equal population districts are a necessary condition for equal citizen power, but not a sufficient one. And I should say that since I learnt that they don't have one of these little red pointers, I have a prop which I just booked in a joke shop for reasons which are too complicated to go into. So if I need to point, I will use the torch of liberty. Right, well, here is a manifesto, an egalitarian manifesto from the United Kingdom. It dates, as you see, to 1838. And it was the founding document of what were called the Chartists. And those, those which I don't need to read out were their six points, as they called them. And you see that number five is equal electoral districts. Suppose now that the Chartist Committee was still in being and we had to report on progress. Well, we got manhood suffrage. Indeed, we now have womanhood suffrage too. We got the ballot. It was actually the first of their six points to be conceded. The abolition of property qualifications and payment of MPs came at the same same time in 1911. Equal electoral districts. Actually, we don't yet have it. And as to annual elections, well, a phrase from John McEnroe would seem to apply. The US situation, about which Sam Hirsch will say much more than I do. Equal districts are in, in fact two senses as interpreted by the court in two senses mandated in the Constitution. That's the original US Constitution as ratified in 1787, and the original phrase in red reflects one of the compromises that was necessary to have the Constitution at all. In case it's not obvious who who Indians not taxed are is a question which if anybody asks, I hope Sam will answer in the question time. But who all other persons are, in case it's not obvious, is slaves. And the deal was that the populous Southern states And Virginia in 1787 was the most populous state, wanted to have as many representatives as their entire population warranted. But of course the slave population had no say in the election of those representatives. After the Civil War, the relevant article of the US Constitution was altered to that, and it's under that article that the work of equal districts is done in the United States. One difference Come on to this in a moment. One difference between the two versions then is, well, the essential difference is that there was a class of persons who were in the population but ineligible to be in the electorate and therefore whether you drew your equality rule from the population or from the electorate made a big difference to the number of seats you assigned to each state. In modern conditions, basing it on population and basing it on electorate makes no practical difference. We should hesitate, Isaac would say, to call a country a democracy unless it respects the principle of one vote, one value and a necessary although, as I will show not a sufficient condition for that, is that all electoral districts should as far as possible contain an equal population or an equal population per representative. This evening I'm only talking about the single member electoral districts of the UK and the usa, but perhaps in questions I can discuss how these principles apply to multi member districts. I am now going to show how the United Kingdom is in violation of one vote, one value, and what might be done about it. As I said a moment ago, I in the course of that want to briefly touch on the difference between voting weight and voting power. I'm then going to show that while in the USA formal equality is achieved under the Constitution, substantive equality is not. So in the uk The Principle of Equal Electoral Districts the principle of equal electoral districts was first conceded in the Third Reform act of 1884, but not for Ireland, which was all part of the United Kingdom at the time. Famine and emigration had ensured that Ireland's 105 seats in a parliament of 670 in a House of Commons of 670 were far too many. But both parties in the 1884 settlement was bipartisan accepted that to reduce Ireland's representation would be too politically sensitive. Irish politicians would have accused the British of first starving them and then depriving them of seats. The 1884 redistribution shifted seats massively as the previous one in 1867 had done more timidly, from areas of relatively declining population which were then rural and southern England to areas of relatively rising population, then northern and industrial England and London and suburbs. However, this settlement provided no means of keeping the apportionment of seats up to date as the population shifted around. Letting sleeping dogs lie by not redistricting when population shifts is nice for incumbents but horrible for citizens. Any redistribution to equalize constituency sizes makes enemies of MPs in the seats to be abolished. The people it benefits have by definition too little or perhaps no say at all in the decision. Even in 1867 one of the to be abolished MPs had described described himself and his fellows as dying swans as he explained why he would vote against Israeli's Reform bill. Redistribution therefore was referred to a Speaker's Conference that met during the First World War, by which time the 1884 constituencies were very unequal in population. This device Speakers conference was designed to ensure that electoral reform had all party support. The problems Both with this one 1917 and the succeeding one in the Second World War 1944 were two I think. One it lacked professional advice. Politicians have an endearing but misplaced confidence in their own ability to solve apparently straightforward issues of districting. 2 It did not overcome the problem of vested interest. Although all parties in Parliament were represented, all parties by definition, neither of those parties not yet in Parliament nor the disenfranchised population were at the speaker's table in 1917. That therefore excluded all women who did not yet have the vote and those men who did not satisfy the then enforced demanding residence qualification. Nevertheless, the 1917 conference did a pretty good job. It recommended the principle that each vote shall as far as possible command an equal share of representation in the House of Commons and set a target population for each seat in Britain 70,000. Ireland was as before too sensitive to handle and the succeeding Representation of The People Act 1918 left it over represented. However, most of Ireland seceded in 1921 leaving Northern Ireland in the United Kingdom. Northern Ireland from the Government of Ireland Act 1920 until 1979 was deliberately underrepresented in the House of Commons to about two thirds of its population share on the basis that it had a devolved assembly. When it lost its devolved assembly in the 1970s its share of parliamentary seats was raised to its population proportion. Now it's got its devolved assembly back at about the fourth try. Its share of seats, however, has not gone back to two thirds of its population share. In fact, all three non English territories are over represented in the House of Commons, although all three of them now have devolved the same assemblies. And I show a graph for Scotland and in a moment I will for Wales. So here the one line is proportionate representation, proportionate to population. So we see that Scotland entered the Union Parliament in 1707 at below proportional representation. It passed to proportional representation appropriately in 1885 when the principle was conceded over the first first time. But it is then gently, and in 2001 not so gently, crept up. The downtick for the most recent election is as a consequence of one of the clauses in the Scotland act which scared that Scotland's over representation would be reduced to proportionate representation because Scotland has a parliament. But note, proportionate representation actually still slightly more than proportionate, not the underrepresentation that applied in Northern Ireland from the 20s to the 70s. Here's the data for Wales. So Wales actually starts over, goes to Approximately proportionate in 1885. And then the same story as Scotland, it starts getting over represented because the relative populations of both places are declining. And unlike in Scotland, there's no downtick in the most recent general election. So as I'll explain in a moment, the rather small redistribution community of academics heavily represented in this room, and if a bomb fell on this room, there would probably only be two or three left, have been complaining for at least 20 years that the rules for redistribution are hopeless. And one way that we were able to make our point was in a report which my predecessor and post, David Boss Butler and I submitted to the Committee on Standards on Public Life, which was reviewing the Electoral Commission government has just replied to that report within the last week and it has now discovered a new constitutional principle, unknown until a week ago, which is in red on that slide. The deliberate over representation of Wales and Northern Ireland in the UK Parliament. And as you see, it goes on on over representation. If that was going to be altered, this would require consideration of fundamental pillars of the devolution settlement. A cynic may believe that it's a fundamental pillar of the devolution settlement, that the Labour Party should retain the power to govern the uk, which depends heavily on its seats in Scotland and in Wales. Could it be argued that the over representation of Wales and Scotland was deliberate? Well, I did some research on this about 10 years ago when I got, and I believe I was the first outside person to get the minutes of the 1944, the Second World War speakers Committee which set up the present regime. And the key things happened in the eighth and ninth meetings of that committee. That committee, like the 1917 one, had representatives of all parties in the House of commons. Unlike the 1917 one, Scotland and Wales were over represented proportionate to population on the committee. And of course the Scots and the Welsh were well aware that their relative population had declined, that they were therefore at risk of losing seats. So in the eighth meeting you get that statement minuted. It would be very desirable to state from the outset quite clearly that the number of Scottish and Welsh seats would not be diminished. The absence of such feeling assurance gives rise to a good deal of political feeling, would lend support to separatist movements in both countries. Now you may ask, where were the Scottish and Welsh nationalists in 1944? The answer is nowhere. But it's clear that the political parties, the politicians in Scotland, used them as bogeymen in order to guarantee their floor level of representation. At the next meeting of the committee, somebody said, hey, if you have equal sized constituencies and a floor for Scotland and Wales, and if Scottish and Welsh population continues to decline, you've got a contradiction quite clearly. I don't need to spell out to this audience, you do have a contradiction between those two rules. And this is the comforting minute written by the secretary of the committee. England's position would be adequately covered by the formula for calculating the quota. The conference agreed that it was unnecessary to amend the rules to meet the line of argument that I've just mentioned. And I think Garrison Keillor has the best comment on that situation, which was then embodied in the legislation which we still have. So, as in Lake Wobegon, where all the children are above average, so in the parliamentary districting rules for the United Kingdom, on average, all constituencies in England must be above average. I'm now going to say something about the second source of inequality in UK districts, and this is incumbent protection and party protection between them. Those two principles have the effect of slowing down the process of redistricting to an unconscionably slow pace. But their effects in detail are slightly different. First, let me talk about the slow pace. The current redistricting which is currently in progress for England, depends on relative electorates in the year 2000, 2000 already seven years ago. For constituencies which are not yet in force, they will come into force at the general election, which we're expecting in 2000 and 22,009 or 2010 and will be in force for probably at least one more general election after that. So by the time they in turn are superseded by the next redistricting, the districting will be 15 years old and therefore there will be extreme inequality of relative electorates. People who live in growing parts of the country are systematically underrepresented. Those who living in declining parts, relatively declining, are systematically over represented. This imparts political as well as spatial bias. Declining parts of the country are poor and disproportionately likely to vote labor if they vote at all. Growing parts of the country have the opposite characteristics. One would expect therefore, that Conservative governments would hurry up redistributions and Labour governments would slow them down, and that both of these have in fact happened. A Labour government slowed down a redistribution in 1969 by a parliamentary trick. A Conservative government passed primary legislation in 1992 to speed it up in both cases for utterly naked partisan advantage. I think the moral of this is that redistricting should be put beyond the sticky hands of politicians to deal with. But even the Conservatives, whose vested interest is in speed, failed to speed up the process to the point that every other democracy with single member districts, with the sole exception of India, can manage. India is worse than the uk, but every other single member district regime is better. Relative populations are known with at most only a few months to why then does UK districting lag by anything between 9 and 15 years? I would say because of incumbent protection as opposed to party protection. It's just that the existing legislators rather like the set of constituencies which they have because they are the MPs for them. Now, when it comes to a redistricting, the fact that incumbent protection is not the same as party protection comes into play as it also does in the U.S. incumbent protection, of course means minimum change. Party protection means, in the jargon, packing and cracking. In other words, if a party can control a districting, and I expect Sam will have much more to say about this, then it wants to either pack the opposition voters into seats which are very safe for them and therefore.
- [25:32]
A
Out.
- [25:32]
C
Of harm's way from the point of view of the party which is doing the redistricting, or to crack them into groups which are just less than a majority in each district. Now if all sides did that equally, the result would be neutral and therefore perhaps democratic theorists would have not very much to worry about. It happens that these skills have not been entirely equal in the British system, and I am a veteran of the last two redistrictings. To give an example of Incumbent protection and party protection. I'll speak of West Yorkshire and the classic nightmare faced by the current Secretary of State for School Education, Ed Balls mp, and the current Housing Minister, Yvette Cooper mp. They sit for adjacent seats in West Yorkshire and they're also husband and wife with a fine sense of humour. The Boundary Commission for England proposed to abolish Mr. Bowles's constituency and pass a chunk of it to his wife. This gave rise to the longest inquiry that I know of in the history of parliamentary inquiries. I have an interest to declare, as I was an expert witness, although not for the Balls Cooper household. Under equal electorate rules, the Boundary Commission ought to have given West Yorkshire 23 seats. Its exact entitlement was to 22.49. How that should round up to 23 I'll come to in a moment, but it should. It, however, assigned 22 seats. Mr. Bowles lost his seat and I think the MP for a neighbouring seat has since been rewarded with an ambassadorship or something. Very tempting. So Mr. Bowles will stay in Parliament but for a different seat. In that case, Ed Ball's interest coincided with his party. There was a declining part of the country and the party had a common interest in resisting the reduction of seats. In other cases, the interest of incumbents may vary from the interest of the parties. An MP for a safe seat wants to keep his seat safe. The party wants to distribute the electorate as efficiently as possible by having a lot of seats which are just marginal in its favour. And these principles have been known for a long time. Elbridge Gerry was Governor of Massachusetts in 1812, and that's what he did. Many of you may have seen this famous cartoon, but a cartoonist in Boston in 1812, when he saw the pattern of districts that Governor Gerry was proposed, said, it's like a salamander drew that. And that's how we have the word. So that is Elbridge Gerry's packing and cracking of part of the state of Massachusetts in 1812. I'll skip now to the principle of one vote, one value. Recall that in 1917 the Speakers Conference recommended each vote shall, as far as possible, command an equal share of representatives in Commons. The current act, which is the product, as I say, of the Second Speaker's Conference 1944, says the electorate of any constituency shall be as near the electoral quota, where electoral quota is simply electorate of the entire country divided by the existing number of constituencies as practicable. They are the same, are they not? Well, actually, they're not. In the technical appendix to this lecture, which I'M not going to attempt to deliver today, but which should appear online in a few weeks. I will do the proof because the proof is not difficult. Nevertheless, it's not something that I'm comfortable with standing here and talking to a general audience. I simply assert that. That it is the case that with the rule written as it now is in the legislation, that as near as possible, each constituency will be the same size that entails and is entailed by splitting entitlements at not at a half, but at the harmonic mean, the formula for which is up there, which is always below a half, and it's further below a half, the smaller the unit. Whereas if you had gone with the 1917 formula saying each citizen should, as far as possible, have an equal share of an mp, you would have splitting at the arithmetic mean, which seems intuitively right. So if the rule says give each elector an equal share of an mp, that entails and is entailed by one unique rule, which is split entitlements at the arithmetic mean. The Isle of Wight, which is the smallest UK unit, has a precise entitlement of 1.45 seats. You can't easily divide the Isle of Wight and make a constituency which is part island, part mainland. So what do you do? Well, under that rule, you give it one seat under the rule which is actually on the statute book, make each constituency equal. That entails and is entailed by splitting entitlements of the harmonic mean. And the Isle of Wight, with about 1.45 standard constituencies, should be given two in rough numbers. It's possibly easier to think of it in the numbers. Standard Constituency is about 70,000 electorates. The Isle of Wight has 100,000. If you give it two seats, they're 50,000 each, 20,000 away from the target. If you give it one seat, it's 100,000 and it's 30,000 away from target. So the rule says give the Isle of Wight two seats. The Boundary Commission, however, has always given it one. So much for obeying the rules. Here is a graph showing the cumulative effect of these various distortions. And the zero line near the bottom is equal. Equal size constituencies below zero. Sorry, equal partisan effect. I should have said below zero means a bias to the Conservatives, above zero means a bias to Labor. Labor always loses when there is a review, which is what you would expect because it to some extent adjusts for population shift. However, it's since the 1964 general election, the partisan bias has always favoured Labour because of the effect of the delays in population shift. However, the huge partisan bias in the system in favor of Labour is not all due to these mechanical effects. And I said I would say something about voting weight and voting for power. And although I'm running out of time, I want to do that in a moment. The authors of the Decomposition of Bias have calculated that there are six components and an interaction effect. So seven in total and the largest single one is differential abstention. Why does Labour win far more seats on a given vote share than the conservatives in each case, plus means a bias in favor of Labour, minus in favor of the conservatives. Very large source of bias from differential abstentions. Which means which has an effect that the effect that very safe Labour seats, or what would be very safe Labour seats are safer than very safe conservative seats. Voters are rational, they know it won't make any difference and therefore they don't turn out. Turnout is very low and therefore Labour wins more seats in total with fewer votes than do the Conservatives. That's an illustration of the fact that voting weight, even if we had equal population districts, we'd still have this effect is not the same as voting power. Those in very safe seats have as near as makes no difference, zero voting power. I will be very brief because I mustn't trench on Sam's time, but there are two issues in the us. There's the apportionment issue, which is between states, and there's districting within states. As we're about to hear, formal equality is ensured in the US by the Constitution and the courts. The reality is very different. And I will tell one story about incumbent protection and then stop in the Old South. The justice department. The U.S. federal justice department is deeply suspicious of state politicians and state legislatures. It believes, not without justification, that they are white racists determined to oppress their black population. There's therefore a power for the Justice Department to intervene if a state legislature draws districts for the U.S. congress and the Justice Department thinks it doesn't like them. And in a number of southern states there's has taken place. And the case that a student of mine has studied was Georgia. After the 1990 population census, state legislature produced a districting. The Justice Department didn't like it. Justice Department rode like freedom Riders into the state of Georgia and said, you must make sure you have majority minority districts and that is to protect the existing African American incumbents in the state of Georgia. And how they. I haven't got a map, but they produced some districts of very extraordinary shapes. We may see some shape districts in the next talk. What did this do. It did two things. One, it made those districts very safe for African American incumbents. Two, it handed the state to the Republican Party since it was the exact opposite of what a rational state Democrat, if such person had been in charge of the state district would have done, they would have spread the Democratic voters, African American voters notoriously being the most heavily pro Democratic group in the entire population. They would have put them into as many districts as they had a hope of winning. And therefore there was a bizarre coalition involving three parties, African American incumbents, the federal Department of Justice, and the state Republican Party to ensure that formal equality did not lead to real equality. So although that's back to the uk, although US achieves a much better result than that, and I won't pause over that table because it will be available in a few weeks time, that it merely shows that even immediately after a redistribution, UK doesn't have very equal constituencies. Nevertheless, I end on two notes. That's one of them and that's the other. Thank you very much.
- [36:50]
A
We have questions later, but I now call on Sam Hersch to give our second talk of the evening.
- [36:59]
B
Thank you. Thanks to lse, to Professors Macova and Farah, and the entire Voting Power and Procedures team for inviting me to come and be on the same stage with Professor McLean. And thanks to Professor McLean for tolerating an enumerate American lawyer like myself. The idea of inviting an American election law litigator to lecture on how to draw fair districts is a bit ironic. Sort of like asking Donald Rumsfeld to lecture on how to manage a foreign occupation. But I do have some good news from the states and I'd like to mix it in with, unfortunately, way too much bad news. So first I'm going to address five problems that are perceived to be and mostly are real problems with American redistribution. I'm going to give five specific examples from different states and then talk about a couple of solutions at the end. The first problem that a lot of people talk about is that the redistricting process in the United States is too litigious. I make a good chunk of my living off redistricting litigation, so I'm really the wrong person to answer that one. But I would say there's basically two ways that things can become too litigious. One is if the standards for challenging the status quo are too plaintiff friendly and too difficult on the defendants. I really don't think that's true, as you'll see, because I think there are so many problems that need to be addressed and some of them courts could be addressing more actively than they are. The other problem is sometimes things get too litigious because the legal standards applicable to a given set of facts are too unclear and murky. And I do think that's a problem here, and more on that later as well. A second problem often referred to is the idea that districts are way too bizarrely misshapen. The good news here is this is actually now improving substantially in the United States. The heyday of truly bizarrely misshapen districts in the US was in the early 90s and after the 2000 census when districts were redrawn. We saw a substantial improvement in large part because the US Supreme Court came up with a doctrine for clamping down on the most bizarrely shaped districts in 1993 and refined over the course of the 1990s. The more important point about compactness, however, is I don't think it really matters all that much at the end of the day. First of all, no one really knows what the district they live in looks like on a map. And the people who really need to know it the candidates who campaign, the pollsters who take public opinion polls, the journalists who interview voters, they usually can find their targets. And what really matters about redistricting is how it translates public opinion, public sentiment into actual elected legislators, and eventually into public policy. Compactness also doesn't matter that much because with modern technologies and modern databases, it's possible to generate a rip roaring gerrymander that looks pretty on a map. So it's not a very good diagnostic tool. And we'll talk about that more later as well. The third of the five complaints you hear a lot about redistricting in the states is that it generates too little racial and language minority representation. There is definitely some truth to this, but perhaps less than some people think. For example, if we look at the African American population, which is 13% of the U.S. population, how does it fare in both U.S. senate elections, which are statewide and therefore don't involve redistricting or gerrymandering at all, and US House elections, which are district by district? Well, we have a U.S. senate that is 1% African American. We have a U.S. house that is 9% African American. Now that's partly just due to the nature of territorial single member districts, but it's also partly due to the fact that we have in the courts a fairly aggressive approach to enforcing the Voting Rights act of 1965, which is one of the great civil rights statutes of American history. And in particular the Voting Rights act has been interpreted basically to look past the question of whether there is in line drawing, discriminatory, racially discriminatory animus, discriminatory intent or purpose, and just focus on the effects and to focus on the issue that Professor Maclean talked about, is there packing and cracking? In other words, is there cramming of minority voters at very high levels of concentration into a tiny number of voters districts to disempower them in the neighboring districts, that's packing, whereas there cracking, where you sprinkle the minority population so broadly among many districts that they and their allies are an ineffective minority in every one of them. Those are basically illegal procedures in most instances in the US Due to the Voting Rights act and the way the courts have interpreted it. And I think the reason that we see 9% of Congress being African American, the House being African American today, while 1% of the Senate is, has a lot to do with the fact that that's been very effective in counteracting racially polarized voting, which is a very widespread and deep phenomenon in the US the fourth issue you hear a lot about is that redistricting generates too few competitive elections and too little responsiveness in our political system. That, in my view, is a genuine and huge part. As you know, we have a decennial process where the census is taken in the year ending in zero. Usually districts are redrawn in the year ending in one, and then the first elections are held under them in the year ending in two. So we had a 1990 census, a 91 redrawing, a 92 election. Same with 2000, 2001 and 2002. What you typically see is in the first election after a redistricting, there are a lot more incumbents who retire. There's a lot more who are defeated if they don't retire. And you get a lot of fresh blood in the House of Representatives and in state legislatures because members are having to take on new constituents and new challenges, and sometimes they're just not up to the task. What we saw in the 2001 redistricting and the 2002 elections, however, was the exact opposite. The number of incumbents retiring and being defeated went down rather than up. Turnover went down rather than rather than up. In fact, out of 435 districts, only four challengers in the entire United States won against incumbents in the November 2002 election, which shows how well the incumbents were protected by the redrawing of districts. In California where there were 50 challengers, none of them even reached the 40% mark, much less the 50% mark. There was a California Consultant who held up the California map feeling very proud, proud about his work on it, and told a reporter, this is a quote, this new plan basically does away with the need for elections. That means it's gone too far. We also saw out of 435 districts, 80 of them, one of the major parties didn't even bother to contest. And this is not just a phenomenon that is widespread across all American elections. It is peculiar to single member districted elections. Again, the Senate provides a useful comparison because we elect our senators statewide and governors likewise are elected statewide. And we saw in 2002, about half the senators and half the governors had races with a winning margin of 10 percentage points or less. So one out of two in the U.S. house we had districts. It was one out of 12. Dramatic difference. The fifth and final problem that you hear a lot about is that redistricting generates too much partisan bias that it unfairly favor one political party over the other. And I speak often here about two parties because we obviously far, far more than you have a fundamentally two party system where there's very tiny fraction of the vote that goes to any third party or independent candidates. And as Professor McLean mentioned, when one party controls the governorship and both state legislative chambers in an American state, they have the potential to generate enormous amounts bias in their favor and against their opposing party, again through these tools of packing and cracking. It's not atypical in a very competitive state where the electorate is almost perfectly evenly divided, but there is one party control of the redistricting process for whatever reason, to generate a plan that has no more than a third of the districts available to the out party and 2/3 of the districts will be controlled by the in point party, the party controlling the pen, so to speak. It used to be thought that this could only happen at great expense to other redistricting criteria. For example, it was thought if you really generate a great partisan gerrymander, the districts will be really ugly. And that is absolutely not true today. And I'll give an example of that momentarily. So the trade off allegedly between partisanship and compactness doesn't really exist except except at the extreme margins today. The other thing is, it used to be thought that if you really tried to hammer the opposing party, you would end up spreading your vote thin enough that your own districts would actually become competitive and potentially could be taken over, as in the Georgia example that Professor McLean spoke of. But again, based on really good software and more importantly really good databases about political voting behavior, that trade off is no longer so true today. Either, as I'll give some examples. We had a very odd circumstance in 2001 in that in the four largest, what we call purple states, red being Republican, excuse me, blue being Democratic, purple being highly competitive, these are Florida, Pennsylvania, Ohio and Michigan. We happen to have in all four states in 2001, a Republican governor, Republican state Senate and Republican state House. So we had a real experiment in how bad can it get. And it got pretty bad. And the redistributing plans that came out of those four states in the 2002 elections generated a two thirds Republican House delegation for the four states, 51 Republicans, 26 Democrats. And that was not spreading them so thin as to create a problem. In 2006 when we had a huge nationwide pro Democratic wave in reaction to the failures of the Bush administration. What we saw there is, if you look at the US Senate races in those four states, Republican popular support was down around 41%. It was clearly the minority party in all four states last year, in 2006, yet they still hung on to 57% of the House seats because the partisan gerrymander in those four states was so effective. So what you see here is, is a good. I don't mean that in a normative sense, but an effective partisan gerrymander can invert a popular minority and turn them into a governing majority. And in fact, for much of the Bush years, the Republican advantage in the House has been wholly or nearly wholly attributable to the partisan bias in these four states as well as a handful of others. So now we can see why our former Republican speaker of the House, Duke Gingrich said, said, and this is about as succinct as it gets, quote, redistricting is everything. Let's give some examples. I'm going to talk about five states. The first is Michigan. We're in federal elections. We have a very competitive state, slightly Democratic leaning, but again, it was one of those states that had Republican state political control in 2001. They drew beautiful districts. They weren't ugly at all. If I could project them up here behind me, you'd say they're really quite obviously not gerrymanders. But in fact, what they had done is they drew five districts that were absolutely packed with Democratic voters. And they contained, in addition to many, many excessive Democratic voters, they contained eight Democratic incumbents, setting up three interdemocratic primaries where incumbent had to run against incumbent. Then in the other 10 districts of the state, they spread the Republican vote vote out quite evenly. They gave each of the Republican incumbents a very comfortable district without any pairings. And they also created two new open seats. One for the state's chief elections officer, another for a member of the state Senate redistricting committee. By coincidence, both were Republicans, of course. Perhaps more disturbing than the output, however, was the process that led to it, which I unfortunately had to live through personally for months and months. We, we heard that there was a great Republican gerrymander afoot in Michigan, but no one had seen it, at least no one in the public, no one in the press. It was put out at the very last moment. It was subjected to one House committee hearing at 11 o' clock at night and then it was passed on a perfect party line vote that night at 2:35 in the morning with not a single Republican or Democratic dissenter in the House. And the complete lack of deliberation over this plan actually resulted in two substantive areas errors. One district was non contiguous and had two separate pieces. Another bigger problem was they left 4,578 people out of the map entirely. So those people had no one to vote for for Congress and this was a big goof. So what were they going to do? The worst thing is they planned to do all this on the very last day of the session before they all went off to summer recess. And therefore there was no easy way to fix it. So, so they all left town before they realized they'd left 4,578 people out of their map. And then they were too embarrassed to send it on to the Governor at that point. So they either had to bring it back to the next session of the legislature in the fall, which would have subjected it to the kind of searching evaluation that they were trying to avoid in the first place by rushing it, or they could do something more devious and just try to fix it by having the Secretary of the Senate and Administrative officer fix the bill and insert some language, give it to the governor, sign it and hope it didn't get struck down. And even that they didn't want to do at a time when it might be criticized in the press more than they were already getting hammered in the press up in Michigan for this. So they waited until a good day when there was something really busy happening in the news so they could do it without getting much press attention. And if it just so happens in Michigan, they record the precise time when the Governor receives the bill and the precise time when he signs it. Governor John Engler, the Republican Governor of Michigan, signed this bill at 4:54pm on Tuesday, September 11, 2001 eight hours after the World Trade center and the Pentagon were destroyed. Needless to say, this did not generate much press the next morning. It did, however, do its job. Michigan had in the 2002 elections a Republican Republican super majority, even though the majority of the people are Democratic. Not a single seat change in 2004 and not a single seat change in 2006, even when we had a huge Democratic tidal wave. This was a very effective gerrymander and it was, remember, full of pretty districts. The Supreme Court did not take this case, although two justices said they were interested in it. The other seven were not. We then turned to Pennsylvania, which is a case that we did get the circumstances Supreme Court to take. I think they took it probably because, unlike Michigan, the districts were really ugly. One was known as the supine Seahorse. One was called the upside down Chinese Dragon. They all got these animal names, but none of them look like a rectangle or any other regular form, I can guarantee you. It also turned out to be what our friend Bernie Groffman, a professor in the U.S. calls a dummy mander. Because come 2006, when Democrats got more popular, it actually wasn't quite a strong enough gerrymander. And the dumb Democrats actually did take over a majority of the House delegation there. But that was all after the Court ruled on the case. We asked the Court to say that if a plan is so biased that even a narrow majority of the electorate, even if the opposing party captures a narrow majority of the electorate, 50, 51, 52%, it's virtually impossible to carry half or more of the seats, that that should be a standard that is so severe or so excessive that the Court would strike it down. And we lost that fight. All nine justices interestingly, said that the excessive use of partisanship and redistricting was unconstitutional. Four said it should be struck down. Four said it should not be a matter within the power of the courts to deal with. And one Justice Kennedy kind of waffled in the middle and he said, well, the problem here is you have to figure out how much partisan effect is too much Paris, in effect, effect. And I don't have a test for drawing the line, so I'm going to uphold this map and let it be. Then we got into a new phenomenon which is going to be the story of the next three states. I talk about briefly. Colorado in 2003 for the first time had unified. First time in many years had unified Republican control. They didn't have it in 2001, when they first could have redrawn the map, but they got it through the 2002 elections. They came back to the legislature in 2005 and for the first time in modern American history, they redrew a map in the middle of the decade, seizing advantage the fact that they now have unified Republican control of the governorship in both state legislative chambers. The Republican state Senate president, in what I think is one of the clearest lines ever written about this phenomenon, told a reporter, quote, we didn't have to play nicey nicey, so they didn't play nicey. I don't know what that means, but I like it. This plan was struck down by a state supreme court on state constitutional grounds. And then we saw though this phenomenon continuing in other states, most notably in Texas, which was the second state in modern American history to do mid decade redistricting. There in 2001, when the plan should have been drawn, there was split control between Democrats and Republicans of the legislature. The Republican governor refused repeatedly to call a special session to do a redistricting. The federal court ended up drawing a map that had seven competitive districts, 14 Republican districts, 11 Democratic leaning districts. It was a reasonably fair map. We had better candidates in the 2002 election and took six of the seven competitive districts. But the Republicans at the state level were taking over the Texas House, giving Republicans in Texas unilateral control of the government for the first time ever. And it turned out later that that was because the state, the US Majority leader, Tom delay had raised a lot of illegal money in order to boost state legislative candidates for the Republican party. And he's of course under indictment for that now. But the newly Republican, the new Republican majority came back to Austin and made repeated attempts to pass a gerrymander in Texas. Attempts takes a 2/3 majority to have a quorum. So although the Democrats didn't control a majority, they could prevent a quorum from legislating and they walked out of the state house. The Republicans then tried to get them arrested and dragged back to the state Capitol to vote against the gerrymander because they needed the quorum. The Democrats got out in the middle of the night, loaded up in school buses and drove over the state line out of the jurisdiction of the police to Ardmore, Oklahoma, where they all checked into the Holiday Inn, which is one unpleasant place I can tell you, and thereby they avoided arrest. So then Tom delay, our US Majority Leader, called the Department of Homeland Security, our anti terrorist department, and said, can we track all the local planes, try to figure out where these guys flew to, never thinking they would actually, being Democrats, take a school bus and eventually the Department of Transportation Homeland Security turned him down. But this stymied the Republicans for a while. There was a similar walkout later of the Democratic senators who went to New Mexico. Much better place than Ardmore, Oklahoma. They had better taste. The Democrats took a real beating on this, however, in that the Republicans who were left behind in Austin started doing all sorts of things. They took away all sorts of staff privileges, everything as petty as parking privileges, up to things as sort of nasty as. There's a tradition that you can, if one of your constituents dies, particularly if he or she was a veteran, you can get a flag that flew over the Capitol and give it to the family. They will sometimes drape the coffin with it at their funeral and they refuse to give any of the flags out to any of the Democrats. Stuff like that. It's pretty cheesy. The map at the end of the day moved 8 million Texans into new districts. It targeted seven Democratic members of Congress and they got six of them in the very first election. And in the very first election with this map, there was only one close contest out of 32. So we brought a legal challenge. Unlike Michigan and Pennsylvania, where our challenges said at the time when they had to be redistricting, those states put too much emphasis on party among all the other things they were doing, including equalizing the populations according to the new census data. Texas wasn't like that because the map already in existence was perfectly lawful. The entire project in Texas was unnecessary and was purely partisan. Not just in effects, but purely partisan in its intent and in its motive. And we lost that 5 to 4 in the US Supreme Court as well. They said that it was okay because the map being replaced was drawn by a court and redistricting maps should be drawn by legislatures. And they also thought that the old map was a little pro Democratic, which is not actually true. But there are argument was you could replace a pro Democratic map with a pro Republican map. And I think that is referred to in kindergarten as two wrongs make a right. But that was the theory. They did, however, strike down part of the map on the grounds that was unfair to Latinos under the Voting Rights Act. And some commentators say that actually was a way of ameliorating some of the partisan impact of the delayed gerrymander. My last example comes from Georgia. Not the decade that Professor Maclean spoke of, but this decade. And this is perhaps the most pernicious form of this mid decade redistricting phenomenon. This does not involve Congress, this involves the state Senate. It does not involve the entire State. It involves one little patch, three state senate districts. One of them was a competitive district centered on the city of Athens, which is where the University of Georgia is located. And that, by Georgia standards, is a liberal bastion. So it created, because it's a very liberal area, the entire district was quite competitive. And in 2004, a Republican candidate won there very narrowly by a couple of percentage points. In 2005, he was already bored with his job, decided he wanted to run for a statewide office. So there was going to be a vacancy. And Jane Kidd, who was a state representative who had a smaller district inside the senate district, decided she would seek a promotion to the state senate. So she immediately announced that she. She was going to run for this open Senate seat in this very competitive district. And then the main Republican candidate who was going to run against her was the same guy that she had beaten for her own seat previously. So it was going to be a rematch. It was going to be a tight race. She had a good chance of winning it. This guy who was running against her, however, was also the brother in law of the retiring state senator who was trying to run for statewide office. So it was a little incestuous. So what did Georgia do for this fellow and his brother in law? They redrew these districts and they cut Athens right down the middle. They cut the University of Georgia campus right down the middle. And they basically shoved her part where she lived and her best supporters lived into another district that had very conservative Republican base out in the rural areas. And they basically made it impossible for her to win. The line, surprisingly, left, not surprisingly, left the brother in law on the other side. And he could then easily win after his most liberal constituents and his challenger had been removed from the district physically. That we would say in Colorado is not playing nicey nicey. This, I think, is the paradigm of a very dangerous trend. However, if redistricting ceases to be a decennial function, starts to be a biennial one, a piece of a campaign, a way to ruin your competition, not through better policy statements, not through better speeches, not through better door knocking or phone banking or television or radio ads, but rather just by changing the district lines. That is very dangerous. And it's not only dangerous in states like Georgia and Texas and Colorado, where one party controls. It's also dangerous in states with split control, because you could imagine you would fix one competitive district in one corner of the state to make it clearly Republican and then trade that off by fixing another competitive district elsewhere in the state to make it clearly Democratic so you could get bipartisan deals along these lines. And what would happen is the very small number of competitive districts we already have would shrink even further. So let me turn briefly, and I don't want to run too far over length, to the question of what we should do to fix all this. Clearly we got some problems that need fixing. I see two main avenues of reform, one being judicial review and the other being state constitutional amendment. Courts, as Professor McLean explained, have been pretty effective actually in ensuring one person, one vote at least in equalizing, let me phrase that more carefully, at least in equalizing the district populations within each state. They've also been fairly effective, as I pointed out in the Voting Rights act, in averting minority vote dilution. I think they could be equally effective if they wanted to be in averting partisan vote delay solution, because they would have to use the same basic techniques, which is outlawing packing and cracking. The court now has looked at this three times in the last 20 years. The U.S. supreme Court, they rejected our focus on thwarting majority will in the effects of a district, which we raised in the Pennsylvania case. They rejected our focus on having purely partisan motivation in the Texas case. So at this point, this issue of partisan gerrymandering is justiciable by the federal courts, but it has never yet been justiced. So we really don't know what the test is. All this could change if one seat on the Supreme Court changes, particularly after the next presidential election. But in the meantime, the real danger is that when a court see a terrible gerrymander, rather than being up front and validating it for what it is, they look for other things. I already mentioned the fact that there was a Latino issue that actually resulted in the invalidation of part of the Texas map. But here's a weirder example by far, because that was a very legitimate voting rights claim. I actually got the Pennsylvania map initially struck down because the biggest district had 646,380 people in it and the smallest district had 646,300. So there was a 19 person deviation. It was a ridiculous partisan gerrymander. But they struck it down not as a partisan gerrymander, but rather as a one person, one vote violation, which is pretty bizarre. We were happy with the result, but then they came back, they fixed the deviation from 19 people down to one person and then we had no claim left. And then it picks up where I described it earlier. So you get this problem where part partisan claims are reconfigured to be one person, one vote claims, or worse yet, race based claims. And you racialize the whole discussion of this issue rather than being up front in dealing with partisanship. The problem of competitiveness, however, is much harder for courts because when you ask what is the right answer on partisan bias, the answer is to minimize it, to treat the two parties symmetrically. And that flows naturally from the equal protection clause of our Constitution, which requires or guarantees the equal protection of laws when it comes to competitiveness, what's the right amount? You can have too much, you can have too little. If you have a state like Ohio that's competitive and has 18 districts and nine are locked down for Democrats and nine are locked down for Republicans, then no reasonable shift in public opinion will matter. That's no good. On the other hand, if you make all of the microcosms of the state, you may oscillate between 180 Republican delegations and 180 Democratic delegations. The minority voice will go unheard in the halls of Congress. There will be no stability, and you'll lose any experience you have among your legislators. So the ideal is somewhere in the middle, but the Constitution tells us nothing about where. So finally, if courts can't get the job done to fix this competitiveness problem, and they're not willing at this point to get it done to fix the partisan bias problem, let's think about institutional design and changing the way we the way our state constitutions require the process to be done. One focus is who does it? You can shift power from the state legislatures to independent or nonpartisan or bipartisan commissions. I think that's a good idea, but I don't think it's as important as how they do it, no matter who's doing it. I think we need to subject redistrictors to very tough, specific criteria. And there are two different approaches to this. Some would say blindfold. Give the job to commissioners and blindfold them. Don't give them information about electoral data. Don't tell them where the incumbents live. Don't tell them where registered Democrats and registered voters, registered Republicans live. Tell them that they are prohibited from looking at that data and that they've got to draw pretty districts. And let's hope it all works out. I think that is a terrible idea for five reasons. One, you can end up with a huge partisan bias inadvertently. We didn't solve racial vote dilution by ignoring race. We're not going to solve partisan vote dilution by avoiding any consideration of partisan data. Two, how do you keep these commissioners ignorant? How do you bar this information from seeping into their brains? Three, what if they get it? Anyway, an Arizona commission tried this and a guy stood up at a public hearing and started reading into the record the addresses of all the incumbents, so they had to drag him out with the police. Four, if the commissioners get infected and learn about some political data, then what? What do you do? You're stuck. And then fifth, I actually think it prevents reforms from getting enacted because everyone's too scared that pretty districts will hurt them. Whereas if you had a registering reform that made minimizing partisan bias a central consideration and described in detail what that means and how it should be operationalized, then neither political party would have to fear that it would get gutted through inadvertence or through some sort of trickier shenanigans. And that basically is what the New Jersey system has been since 1966 for state legislative districting. And it's been quite successful. And I think that we should think hard about broadening that out so that not only can the members of the country commission participate in the process, but the members of the public can. They can submit maps through the Internet, have them judge based on some clear cut formula for determining partisan bias or for that matter competitiveness or whatever the criteria are that you care about, and forcing basically a kind of competition which I'm happy to describe in more detail in the question and answer session, if you like. The central point here is I think the problems of gerrymandering are not entirely insoluble. I think the politics of always will be and probably should be at the center of this project. But just as the gerrymanders have figured out how to harness the power of actual electoral data for their benefit, now we reformers must figure out how to harness the power of actual electoral data for the benefit of some common good in the public interest. I think this can be done. I'm optimistic that we will make progress in the next 10 or 20 years on this in the United States states and perhaps then Democrats and Republicans will at least begin to, as the Colorado State Senate President put it, play nicey nicey. Thank you very much.
- [67:44]
A
Thank you very much. I would like to thank the speakers. We have some time for questions, but I urge you to be brief and ask real questions, not make long statements as sometimes people tend to do. Yes, I'll take a few questions and then bunch them. Could you please say who you are?
- [68:13]
D
Sure. I'm Paul Ingram. I'm co executive director of the British American Security Information Council. So quite appropriate to this talk. I am hesitant at bringing in another example to support my case because We've already had lots of examples, but the European Parliament votes. I bring this up for two reasons. The European Parliament votes. Vote for the European Parliament and the apportionment to different countries is not proportional. Proportional to the population. I wondered if Professor Maclean in particular could outline what that system is and why that could be seen as democratic under the criteria you're talking about. And if we look at how Britain in particular has divided its members for the European Parliament, we have a case at the moment where there is an additional member being allocated.
- [69:08]
B
And it.
- [69:08]
D
It could go to London or it could go to Scotland.
- [69:11]
B
I believe.
- [69:12]
D
Now, again, I believe there isn't a proportional system here, or not strictly. And then we come up to my second point, which is the modern equivalent of the slaves that you were talking about. We have people who are not on the electoral roll and we have people who are migrants who may be on the electoral roll role. They may be European electors, but they don't get a vote in the general election. This too can strongly distort the number of the relationship between the populations and the electors.
- [69:49]
A
Next question. I'd like to take three or four. Yeah, please. Can you please introduce yourself?
- [69:57]
C
Yeah.
- [69:57]
D
My name is Andrew Johnson.
- [69:59]
C
Two questions, one on American, one on the British side. Is there evidence of packing and cracking in the UK or is that more a US phenomenon? And on the American side, do you. What's the answer to dealing with the gerrymandering? Is it just to wait for one of the Supreme Court judges to die and just expect a change there, or do you have a legislation ready for implementation if something happened in the style of the post 911 when they brought out that legislation?
- [70:38]
A
Next question, please.
- [70:43]
E
Ken Ritchie from the Electoral Reform Society. My question, I think, is more towards Sam Hirsch, but Ian McLean might want to comment as well. Samher, you concluded by suggesting that you thought that, yes, there were ways of solving the gerrymander problem. But, you know, there's a sense in which even a gerrymander might result in a situation in which there are more people who have got, you know, who have got a representative of their preferred party. Just as gerrymanders take place in the States to make sure that there are people from minority communities that secure representation. I mean, it strikes me it is a complete muddle of arguments that are applied to how you actually draw the boundaries. Is it not that the problem is actually in the boundaries and the use of single member districts themselves? But until you actually move beyond that, do you move into an area where you can actually have, you know, politics that is going to be competitive, where there is going to be a real scope for debate for change that will take place between one election and another, you know, in different areas. And I just don't see how you can actually do it, you know, with this mix. I mean, sure, I accept that it is, you know, it has been so badly used by the parties in recent years that it is in total disrespute. But are there not places in the states where there is at least a debate around about experimentation that should be encouraged?
- [72:18]
C
Yes.
- [72:19]
A
Sam, would you like to go first?
- [72:22]
B
Sure. Brief response really to all. I just want to say briefly in response to the first questioner. In the United States, what we're equalizing is not the number of registered voters. It's the number total population, which includes non citizens, includes illegal noncitizens, includes children, includes people who could be registered but aren't. And the result of that is there are enormous disparities in the number of votes cast in some districts and in other districts, which is a problem. On the other hand, what the Supreme Court has made very clear is they're trying to equalize the number of constituents per representative, not the number of voters. And it's solving a different problem. And it's a controversial issue, but an important one. And it's really a big difference. You get literally three or four times as many votes in the most heavily voted district in Texas, say, versus the least heavily voted on the second questioner. As far as legislation, I think the chance that you would actually get federal legislation to fix this problem is very slim. First of all, I don't think the Congress has power to fix this problem for state legislatures. I actually think it's beyond their power constitutionally. They could fix the problem any way they want. As to congressional districts, in theory, in fact, I just don't see it as being terribly likely. I think the best hope is to try to get state constitutional amendments passed piecemeal in a few states, particularly those where you can put something on the ballot for public initiative and thereby change the state constitution. And what I would like to see put on the public ballot would be something roughly akin to New Jersey's system, where there are certain criteria you have to satisfy. Obviously, compliance with federal law is one, a certain degree of equal population is another. And then among maps that all satisfy those minimal criteria, basically there's a competition to see who can minimize partisan bias bias, who can maximize competitiveness among districts. And this is done by having a sort of Tripart Commission with Democrats, Republicans and a neutral Chair. I think it only works if the chair is very responsible or is bound by very specific rules about what it means to minimize partisan bias, to maximize competition, et cetera. I do think it's solvable. I don't expect Congress to do it to itself, frankly, in my lifetime, but I think if some states do it and have success with it, it may spread to other states. As for the last question, certainly what a lot of what I'm discussing is trying to suffuse a fundamentally non proportionate system with a certain degree of proportionality. We've been able to do that to some extent for our larger racial and language minority groups. We've to been able able to do it in states like New Jersey to some extent for political parties. I think it's correct to say that single member districts inherently tend to over represent large groups, underrepresent smaller groups and in general not be as responsive to small or medium sized shifts in public opinion. However, they have some benefits as well. You end up having a local member of Congress, a local member of the senior state senate and state House who is your member, who you can visit and who knows your area well, and you end up more likely with a two party system that generates majority governments. Now these are trade offs and I suppose a lot of what I'm talking about is really trying to get some of the benefit of a portion representation system pasted on top of what is fundamentally a territory based single member districting system, which I think we are not too likely to change. However, there are some county and city governments and school boards that have gone to semi proportional systems like limited voting and cumulative voting, sometimes in response to being held liable for Voting Rights act violations throughout the United States. They're mostly in smaller places, although we have some history. For example, the whole state of Illinois used to have proportional semi proportional representation system for the Illinois house up until 1970, I believe. 1980. So there's some history of this in the United States, there's some of it now, but basically the dominant system is the territorial one. And I'm trying to get some of the advantages of a proportional system pasted on top of that.
- [76:40]
A
Ian, thank you.
- [76:42]
C
I'll take the first and the third questions together because Ken says surely the problems are unavoidable with single member districts. And Paul says even with the multi member district European Parliament there are problems and those perspectives are both correct. And of course the issues of local representation and so on are very well known. I won't go into them. The advantage of for the purposes of tonight's agenda of moving to something like a large district European Parliament style system for say the House of Commons would certainly be that the issues districting are diminished. They wouldn't however go away. You would have to decide the boundaries of your multi member seats and you also have to decide how to assign seats within those to the parties which contest them. Now the European Parliament problem therefore breaks down into a between state problem and two within state problems. Between state problem is that the size of state delegations in the Parliament and more importantly in the Council of Ministers is, as Paul rightly said, not proportionate to population in the Council of Ministers. In fact, in both cases it's it happens to be. And whether this is purely chance or design, I don't know. I suspect purely chance. But the world's greatest experts on the following subject are in the room. The ratio is approximately a square root ratio and the mathematical reasons why that is a good ratio were approved in 1946. The European Union dates to the early 1950s. Had they read Penrose's seminal paper on the subject when they first wrote the constitution of the European Union? Almost certainly not. So the fact that the between state proportions are roughly as to square root of population I think is just a happy accident. I would say that for the European Parliament it should be something more like proportionate to population. For the Council of Ministers it should certainly be something like square root. But others in the room have strong opinions on that. The within state problem for the within member state problem for the European Parliament consists, as I just said, number one of how do you assign boundaries of seats? And having done that, of multi member districts, having done that, how do you assign the number that each district gets? Number two, how do you assign seats to parties within these.
- [79:29]
E
Units?
- [79:30]
C
Now here the social choice political science community scored a small victory and I suspect that Paul the questioner actually knows that and he's planted me a soft question so that I can boast. What happened for the within state problem was that the Electoral Commission, which is is charged with this, produced four methods in a consultation paper for assigning seats to each of the UK's 12 European Parliament constituencies, which are its 12 standard regions. And I got together a bunch of academic specialists in this, including the American colleague who is the co author of the standard war work on apportionment in the us. And we sent a memo into the Electoral Commission saying they're all rubbish. None of the four methods that you propose meets the statutory criteria that you have to meet and you should go for what real enthusiasts and I recognize at least two or three in the room know as the Saint Lague system of apportionment of seats to the 12 regions and they've done that. And furthermore, to answer your specific question, they've confirmed that they're doing that again in the upcoming reapportionment where the total number of European Parliament seats to each region is determined. So my answer on that point is they've done the right thing. Whether it goes to Scotland or London, I don't know, but it should be going according to good numbers. For the within district, they're not using the Saint Legau apportionment rule, they're using the Daunt apportionment rule which favors large parties. What a surprise since it was large parties who wrote the rule. I had a little bit of parliamentary fun some years ago with Jack Straw, who was then the minister responsible for introducing the system, a well known opponent of proportional representation, and he had a riff in the House of commons about 10, 15 famous Belgians and that Victor Dont was one of them, and said studies that have been done in the Home Office prove that the DANT system is the fairest. And I said innocently, well, actually it didn't say this in what I wrote, but since I knew that mathematically that could not be true, I asked innocently, for the Home Office studies, it was a little bit like the case of people coming to prove some theorem which you know to be false. I knew that the studies must be wrong and it was a question of working through them until I found the mistake and which I did. And although there's no substantive triumph at all, I did force Jack Straw to apologize to the House of Commons. However, they stuck with Dhont. And finally the second question, question addressed to me, is there evidence of packing and cracking in the uk? Not very much, because in equilibrium the parties are equally good at it. And what they do is the Conservatives produce one set of bogus local ties and the Labour Party produces another set of bogus local ties and a great deal of public money is wasted at inquiries which go on for weeks and weeks and in equilibrium it's a standoff. And in a way contrary to something that Sam implied. We do have non partisan districtors in the shape of what are called Assistant commissioners who really don't know the area and who don't know which local ties claims are more bogus than which others. In equilibrium that would be okay. It so happened that in the last two, let's say not the current one, but the one before and the one before that, the Labour Party was much better organized than the conservatives and so they net did more packing and cracking than the conservatives. My impression of the one which is currently going on is that the main parties were roughly equally competent and so the packing and cracking cancelled out.
- [83:34]
A
So some good news. I think we have time for a couple more questions. Yes.
- [83:44]
D
Hello, my name is Daniel.
- [83:46]
B
My question is for Mr. Hirsch and it's really simple actually. I just wonder about the effect of felon disenfranchisement on redistricting in the U.S. especially as concerns those areas of the U.S. those states which get more population from prison movement. Yeah, that's a good question.
- [84:08]
A
Is there another question if not? Yes, please go ahead. Yes, go ahead.
- [84:17]
B
In almost all states in the United States, felons and in some states ex felons are not allowed to vote. And that has a huge political impact because it's thanks largely, I think, to many years accumulation of drug laws and drug convictions actually affects a very sizable number of people. So that's one effect. Another effect is that when district populations are being equalized. For the reason I gave earlier to the other gentleman's. In answer to the other gentleman's question, we're looking at total population rather than registered voters. So if you have a 10,000 voter prisoner facility in a state legislative district, it can be actually a very underpopulated district because none of them are voting yet they count towards the population total. And you'll see this for example, in upstate New York and then in downstate New York, meaning New York City is actually where most of those folks are from and their districts have no prison population and therefore they have a lot more voters than they otherwise would have. So they are underrepresented. And then some rural legislator upstate who has a prison in this district, his constituents are very overrepresented at the congressional level. Whether or not this has a large effect or not, I'm not sure. But certainly when you're talking about state legislative districts which are smaller, it can have a pretty substantial effect, at least on some districts. The biggest, bigger problem though is these rules that don't allow people who have completely served their time, are done with parole and done with probation, and are not allowed back into the polity and not allowed to vote and participate as equal citizens. I think that's really unjust and I think that there is a strong movement now which will ultimately succeed to get rid of those laws.
- [86:09]
A
Well, all good things must come to an end. This has been a very good thing. I enjoyed myself and learned a few things I would like. Oh, we are hoping to have a podcast of these proceedings eventually, which will be available. I would like to thank you, the audience, for being here and for your participation and of course, our two speakers, speakers which have really contributed to a very successful evening. Thank you very much.