Princeton Election Consortium

A first draft of electoral history. Since 2004

Symposium on the Law of Democracy – Stanford Law Review

February 7th, 2016, 6:03am by Sam Wang


This weekend I attended a symposium on the Law of Democracy at Stanford University. The schedule is here. The list of attendees and presenters included Stephen Ansolabehere, Nate Persily, Charles Stewart III, Jane Schacter, Bertrall Ross, Sam Isaacharoff, Heather Gerken, Rick Hasen, Robert Bauer, Ben Ginsberg, Richard Pildes, Bruce Cain, Maggie McKinley, Rabia Belt…this is quite a concentration of election-law practitioners and researchers.

Below I give live highlights of selected parts of the program, including:

  • Rick Hasen on voting rights under the current Supreme Court.
  • Federal Election Commissioner former chair Ann Ravel on the broken regulatory process.
  • The legal future of my statistical partisan-gerrymandering standards (article now accepted at SLR).
  • Ben Ginsberg on how this year’s Republican nomination process may play out.
  • The consequences of rising mistrust, not just in government institutions, but across many sectors of society.

First, a brief aside. The event is an unfamiliar area for me. I focus specifically on election statistics, and I am at this conference to talk about partisan gerrymandering. This is just one part of election law, and there is lots of new stuff to learn about. Instead of covering it all, I’ll highlight just a few presentations.

This symposium is not the biggest news in town. This week, Stanford University named Marc Tessier-Lavigne as its new President. (Coincidentally, my lab has learned a lot from a member of Marc’s group, about how to make brain tissue transparent for mapping its circuitry.) The law faculty seem to view him as another in a string of technically-oriented presidents; the outgoing president is John Hennessy, a computer scientist. However, Tessier-Lavigne is an interesting guy. He is the first in his family to attend college, and he has expressed strong support for the value of a broad undergraduate education, and for the importance of research. I think he’s going to surprise people around here – in a good way.

And then there is the Super Bowl, which is down the road in Santa Clara tomorrow. The Denver Broncos have been training at Stanford. My hotel has bestowed upon me football-shaped cookies. Wait, where are the chad-shaped mints? No ballot-themed placemats?

>>>
The first panel on Friday kicked off with a joint presentation by Nate Persily, Charles Stewart III, and Stephen Ansolabehere on voter-ID laws. They focused on the increasing restrictions on voter access – and on changes in public attitudes. These changes are highly partisan. Republicans are highly favorable to voter-ID laws, as they always have. Support among Democrats has fallen off quite a bit in the last few years, and see these laws as a means of suppressing minority voters.

However, along with this distrust comes a lack of knowledge (on both sides). Voter-ID laws vary dramatically by state, so they are hard to keep track of. The presenters showed that 30-40% of voters don’t know whether they are required to show an ID when they vote. Several in the audience questioned whether a survey is the right way to probe this information, since after all, the thing that matters is practical information. The authors took the position that whatever the case, they’re pretty sure legislators should not assume that people know what the law is. What a mess!

>>>
Saturday, 8:30am Pacific. As is the case at all conferences, some of the most interesting discussions happen during breaks. Over lunch on Friday, Nate Persily, Robert Bauer, and Ben Ginsberg got into a discussion of the weakening of the political parties. Bauer or Ginsberg pointed out (and the other agreed) that a principal consequence of the McCain-Feingold campaign finance reform was to weaken state-level parties. Consequently the national parties have shouldered them out as a main source of funding.

Persily brought into the mix the idea that membership in political parties is declining. All were interested in this. I pointed out that so-called independents still vote with a particular party. However, Ginsberg replied that this still weakened the parties.

There is a broader point: trust in nearly all public institutions is declining. Not just Congress, the Presidency, and the judiciary, and not voter-ID laws. But also organized medicine, universities, news media, corporations – everyone, except maybe the military. To me this suggests that declining trust in public institutions is a root cause of whatever ails U.S. politics – not necessarily a consequence. The cause of declining trust seems to be unknown, but it has occurred in parallel with increasing economic inequality. For example, declining trust weakens the legitimacy of government, which then leads to lower voter turnout, detachment of officials from citizens, and other ills. Of course, that can feed back to reduce trust further – but as I said, trust is declining across many elite sectors.

The topic of mistrust arose again in the last session, with Richard Pildes and Michael McConnell. Pildes was talking about mistrust in Congress and how legislators run scared (i.e. reflexively vote yes) whenever they encounter the word “transparency” in a proposed law. McConnell thought that mistrust arose from malfeasance by any kind of intrusive authority, whether governments or corporations. I offered an alternative interpretation. I said that people looking to understand mistrust in government should keep in mind that it is not specific to that sector. I cited research showing that societal mistrust is highly nonspecific, and includes institutions that include science, medicine, universities, and other individuals.

The challenge is to understand: what is the true source of this mistrust? The study linked above suggests rising inequality, though their findings are strictly correlative. To the panel I suggested that the increasing availability of bad information was a possible cause. Or perhaps too much information in general, since it is always easy to find a nugget of information to undermine something you are told by authority. So in the end, transparency laws could be paradoxically counterproductive.

 

Let me note the self-referential aspect of my proposed explanation: before raising my concerns, I searched Google Scholar for the study in question, which I had read about online. And then I challenged the assertion of a senior political figure, McConnell.

>>>
Partisan gerrymandering. On Friday afternoon, I presented on the question of how courts should deal with partisan gerrymandering. My paper, which I am pleased to announce is accepted as part of the issue of the Stanford Law Review associated with this weekend’s conference, is available in draft form here.

Most of the Supreme Court, on both left and right, think that partisan gerrymandering is wrong. Five justices have opined that “partisan asymmetry” – a concept espoused by Bernie Grofman, Gary King, and others – could be a criterion. I presented statistical analyses, some of which are very simple, that could provide a rigorous and easily-used tool for judges. I emphasized the point that my measures are so simple that they are disintermediating – they can remove the need for an expert who stands between the judge and the facts.

For example, in the current Harris v. Arizona Redistricting Commission case, the difference between the median and the mean vote share can be calculated in sixth-grade-level homework, and shows an absence of asymmetry in Arizona’s state legislative districts. In another example, Maryland (Shapiro. v. McManus), Democratic Congressional districts are far more evenly distributed than would be expected by chance. My central idea was a “zone of chance” which defines what could arise from natural variations in districts. This is the same idea as significance testing in statistics.

What followed was a very lively discussion. Heather Gerken and Nick Stephanopoulos pointed out that the difficulty of getting justices to accept a particular threshold. My own view is that statistics has a long tradition of setting cutoffs such as the 95% significance threshold, and the Supreme Court has done this in the domain of IQ thresholds for capital punishment (which, like partisan gerrymandering, is a Constitutional question) and housing discrimination (which is a statutory question, very different). My analysis sets a threshold in a way that is natural and obvious – to a scientist. Gerken thought that work was needed to turn the statistical ideas into a form that courts could accept. I expressed the point of view that if statistical reasoning is off the table, then the problem is probably dead. However, there was enthusiasm in the room for finding a court case where these ideas can be applied.

Bruce Cain weighed in that many have taken a crack at the question of defining partisan gerrymanders – and failed. The implication is that based on past performance, failure is probable. My own sense is that previous approaches placed a premium on the role of the expert. A simple enough approach removes the expert from the process, which judges might like. The gerrymandering problem has been around for over 200 years (even predating 1812, the original Gerry-mander). The tools I propose (t-tests, chi-square testing, and numerical simulation) have been around for 50-100 years, so they are about ripe for use. I think he was among those who agreed that something new was needed – either new approaches, or the right court, or both. I remain optimistic.

 

Stephanopoulos also cited the problem that one had to know what to do with uncontested races, and with the existence of bias arising from population clustering. These are questions which I deal with in my paper.

>>>
Ann Ravel (Federal Election Commission). The Friday after-dinner speaker was Ann Ravel, outgoing chair of the Federal Election Commission. She has stated publicly that the FEC is basically broken, because a bloc of three votes (there are six FEC members in total) votes reliably against any investigation of campaign finance violations – by either party. This includes coordination of “dark money” such as super PACs with individual candidates. For example, a candidate can go to a super-PAC fundraiser where only two people are present. It was left unsaid, but the anti-regulation bloc is composed of Republicans. A student asked what could be done about this. Ravel replied that currently, the tradition is for the President to vet all six nominees with the Majority Leader and the Minority Leader of the Senate. She suggested one option: the President could nominate six persons of his/her own choosing, people of such sterling character that the Senate could not deny them.

>>>

 

Some great panelists this morning: Robert Bauer, Richard Hasen, Bertrall Ross, Michael Kang, Maggie McKinley, and Richard Pildes. Off to hear them.

Rick Hasen.On Saturday morning, Rick Hasen spoke on election law’s path in the Roberts Court’s first decade. He said that his negative (from a liberal point of view) predictions from 2006 have mostly – but not entirely – come true. In the domain of campaign finance, cases such as Citizens United have opened the way for near-unlimited flows of money from individuals and corporations into politics. There is little room for improvement. In the domain of voting rights, any approaches to preclearance for election/districting plans face severe constraints. These changes constitute a sharp right turn.

However, the news is not all bad. In the domain of election administration, there is room to maneuver because post-Bush v. Gore, cases now descend to state-by-state battles, as the Supreme Court has withdrawn from voter protection. (To me that sounds like pretty weak good news.)

He asked the general question of why the Roberts Court hasn’t gone further in its rightward turn. He cited three reasons:

  1. They’re going slowly, achieving goals across multiple cases. He wonders if that will continue with a new President, and as justices approach or pass the age of 80.
  2. They have a bunch of near-mandatory cases, where they don’t always want to make new law. Whenever cases come up through a 3-judge appellate court, they either have to take the case or it becomes binding precedent. So they take a lot of cases. Since Shapiro v. McManus made it easier to bring these cases – including gerrymandering – this trend seems to continue.
  3. The conservatives on the Court aren’t all the same. For example, Scalia likes campaign finance disclosure, but Thomas doesn’t. And in the area of partisan gerrymandering, Kennedy is the deciding vote.

 

10:45am: In Q&A, Robert Bauer points out super PACs are starting to move into areas that the parties used to fill, such as maintaining voter lists. He thinks that one could channel this to increase the competition among different sources of money. This includes sources such as public financing of campaigns (which Rick Hasen says won’t happen).

Ben Ginsberg. The Saturday after-lunch speaker, Ben Ginsberg, is a central player in Republican politics. He was national counsel to the Bush-Cheney campaign in 2000 and 2004. He walked us through the Republican delegate selection process in detail, which he knows very well. As you can imagine, we had a lot to talk about afterward.

Major points:

  • On the Democratic side, he thinks the nomination process will be effectively finished in April, as the primaries start demonstrating a clear leader – in combination with superdelegates, who compose about 15% of all delegates.
  • On the Republican side, there are few superdelegates. In past years a presumptive nominee has emerged when about 68-71% of delegates have been chosen. This year, that date is April 19th. However, the presence of Trump and Cruz in the field, two renegades, complicates matters.
  • It is frequently said that many Republican states have proportional assignment of delegates, as mandated by the Republican National Committee. However, state parties have evaded that requirement by inserting many loopholes into the rules (as I have written). He indicated that he didn’t have a strong intuition for how those rules play out in real life. I described my simulations, in which I implement the rules, allowing for up-and-down variations from state to state in a divided field. He did not argue with my result, that 30% of the popular vote by Super Tuesday can garner 50% of the delegates, if the field remains divided. He thinks the Republican establishment will wait until after that to winnow the field. So a divided field may be the scenario. Personally, I think if they do that, they have waited too long
  • If I understood him right, he disagreed with my contention that Trump getting 50% of delegates by March 1 could create substantial pressure on the party to give in. Instead, he thinks the battle could still be protracted.
  • Candidates have to decide how to allocate resources efficiently. Some focus on TV time (Trump, Carson), some focus on ground game (Bush, Rubio, Cruz), and some seem to do neither (Christie, Kasich). Also, money has to be saved to take care of later states such as Florida and Ohio on March 15th, require a lot of investment. If the nomination process goes to late March, super PACs will start to be more important as candidates run out of money.
  • In his general observation, candidates stay in the race for one round of primaries past the time when it is clear that they should drop out.
  • If nobody goes to the national convention with over 50% of committed delegates, he anticipates either a big fight over rules and credentials, if there are multiple plausible candidates; or intense dealmaking if one candidate is close to 50%. However, there won’t be power brokers, since there aren’t any such people left. It will be a free-for-all.
  • Delegates are only obligated to vote for their assigned candidate on the first ballot. After that, they are free to vote as they like! They are not picked by the candidate’s campaign, but at state conventions. This means they have complex loyalties. (This is an obvious way in which an outsider like Trump or Cruz is at a disadvantage.)

(The End. I couldn’t summarize all the great presentations, unfortunately.)

Tags: 2016 Election

23 Comments so far ↓

  • Doctor Science

    I am eager to read what you have to say about the NC gerrymander situation. Could they get a new map in 2 weeks if they called you in?

    • Sam Wang

      No way! They have their own experts. Also, I have no experience drawing maps. Finally…the districting process is still Republican-controlled, the same party that drew the offending map in the first place. They will probably draw a map that is as similar as possible to the one that was overturned, making small changes to a few districts.

      In this case, the overall map is a highly partisan gerrymander, but the case was brought on racial grounds, for which the legal case is on far more solid ground. So the remedy will be fairly focused on a few majority-black districts, unlike what one could potentially achieve with a partisan-gerrymander argument. As readers around here know, partisan gerrymandering is a justiciable offense, but there is no clear standard in place.

  • Doctor Science

    Alas! What would it take for Federal law to force states to use nonpartisan redistricting commissions? Is this even a thing that could be done?

    • Sam Wang

      In the absence of major new legislation in this area from Congress, which won’t happen any time soon, the only route is state-level legislation. Since legislators don’t like to give up power, the initiative process is the easiest way to go. This has happened in Arizona, California, and recently Ohio.

  • Amitabh Lath

    If you have time, visit SLAC (Stanford Linear Accelerator). Site of 3 Nobel Prize winning discoveries (quarks, charm, and tau lepton). Now the accelerator is used as a light source to uncover structure of bio molecules.

  • bks

    Be sure to ask what effect Superbowl parties will have on the New Hampshire primary.

    • Olav Grinde

      Ah, the Superbowl!

      As a European, it seems to me that the term “football” is wildly inaccurate. I keep waiting for the teams to “kick the ball” — and at most that happens a couple of dozen times during an entire game.

      Shouldn’t it instead be called American Carry-Ball??
      (ducks to avoid tomatoes)

    • bks

      Hey! We do play it on foot. We’re not riding on camels, horses or elephants. The real question is not about foot, but rather about ball because the object we carry, throw and kick is not one.

    • Amitabh Lath

      You jest! But I did read an article that if the Patriots were in the Superbowl then for some reason it would affect Trump support (presumably because NH Trump supporters would be partying in Boston, more so than say, Cruz supporters who probably all watch curling or something…)

      Punditry is hard to parody.

    • bks

      I was not being facetious Amitabh. I went to a superbowl party and we discussed politics a lot. That’s Berkeley. Don’t know what it’s like in New Hampshire.

    • Matt McIrvin

      The real big variable for the NH primary is going to be the weather. The snow that is starting to fall here in Massachusetts may stretch into tomorrow and affect, especially, the southern part of New Hampshire. But precisely how much, how far north and how late is really uncertain. It might not be so bad.

  • Amitabh Lath

    Has there any more discussion about the loss of trust in institutions and expert opinions? I have encountered this when speaking on anthropogenic global warming, as well as discussing causes of autism.

    In both cases what I found distressing was the questioning was not at the edges of the science but the very root, things like blackbody radiation and infrared spectrum of CO2 molecules, or in the case of autism basic studies of heredity that even Gregor Mendel would have understood.

    • Sam Wang

      There was. I brought it up in the last session, with Richard Pildes and Michael McConnell. Pildes was talking about mistrust in Congress and how legislators run scared (i.e. reflexively vote yes) whenever they encounter the word “transparency” in a proposed law. McConnell thought that mistrust arose from malfeasance by any kind of intrusive authority, whether governments or corporations.

      I offered an alternative interpretation. I said that people looking to understand mistrust in government should keep in mind that it is not specific to that sector. I cited research showing that societal mistrust is highly nonspecific, and includes institutions that include science, medicine, universities, and other individuals.

      The challenge is to understand: what is the true source of this mistrust? The study linked above suggests rising inequality, though their findings are strictly correlative. To the panel I suggested that the increasing availability of bad information was a possible cause. Or perhaps too much information in general, since it is always easy to find a nugget of information to undermine something you are told by authority.

      Finally, let me note the self-referential aspect of my proposed explanation: before raising my concerns, I searched Google Scholar for the study in question, which I had read about online. And then I challenged the assertion of a senior political figure, McConnell.

    • Doctor Science

      One test of your “more info and bad info” theory, Sam, would be to see if nonspecific trust is declining in societies in which inequality has *not* grown — the Netherlands or Sweden, for instance. They’ve got the same info issues the US has, so that would separate the two variables.

    • Sam Wang

      That is a good idea. I wonder if it has been done. I should look at the references to, and from, that psychology paper I linked.

    • Matt McIrvin

      The Internet makes it easier to form ad-hoc communities, and communities devoted to advocating ridiculous things are among them. It’s much easier to find people who believe the same things you do.

      Over just the past year I’ve been noticing a lot of activity from flat-earthers. Literal flat-earthers, who insist the earth is flat. It got some higher-profile attention recently when rapper B.o.B revealed himself to be one of them, but the ones I’ve been seeing around online are more likely to be far-right white neo-Confederate types, and also the sort of people who subscribe to basically every conspiracy theory at the same time.

    • Some Body

      The issue here, though, is not decline in authority as such (which would probably have had a net positive effect), but the decentralization of authority. All those web communities, sects, and conspiracy theory believers have sources of information and ”opinion leaders” that they trust, often all too blindly. When someone tells you they don’t believe in anthropogenic global warming, with very few exceptions, that wouldn’t be because they examined your claims and decided that tbey are not supported by evidence, but because they trust someone else’s statements for the contrary. That someone else is a greater figure of authority for them. Or to take the other side of the coin, when I reject the official account by Israeli military and media of some event or other in the West Bank, it’s not because I generally distrust authority, but because I find accounts by human rights groups to be more authoritative (and more likely to be objective). Most of my compatriots choose to view human rights groups as malicious foreign agents, and trust the official account (despite immense growth in inequality, by the way).

    • Some Body

      And to add one more point as a PS: I think our baseline in this discussion is anomalous. Western nations in the second half of the 20th century, with major network news and all, are perhaps a case of the most centralized information flow in history. Even the Catholic Church in the high middle ages had to rely on local pastors to reach the masses, leading to a proliferation of heresies.

    • Olav Grinde

      SAM: “what is the true source of this mistrust? … To the panel I suggested that the increasing availability of bad information was a possible cause.”

      I believe there is an additional core problem. Fact-checking and the search for objective truth was once the paramount priority, or at least a high priority, of news organizations. That is hardly the case anymore!

      CNN is not the only organization that practices false equivalences, quoting pundits on one side and the other side of any given topic, without making any reasonable effort to identify and report the truth.

      There was a time when, if a presidential candidate were to lie or play loose with the facts during a speech or debate, that would dominate the day-after headlines.

      That is no longer the case.

      Today, fact-checking simply is not given appropriate headline space. At best it’s relegated to side articles, tertiary article and footnotes.

      The only exception seems to be when it’s cherry picking that is done for purposes of political or religious propaganda (in either case: preaching to the congregation).

    • Amitabh Lath

      Maybe ’twas always so, and the years between winning WW2 and the moon landings constitute an era of anomalously high trust in institutions and now we are reverting back to historical means?

    • Some Body

      Could well be, in one sense. But I’d also say ”trust in institutions” is probably not a coherent measure of anything. Church and royalty enjoyed far greater trust in medieval and early modern Europe, for instance (there were at least two peasant revolts I know of, one inEngland and one in France, where a single false promise from the mouth of the king was enough to break up a previously-succesful uprising at once; but then, this would probably not work with a coup by nobility). But if you look narrowly at what is publicly considered to be knowledge, I’d say that period was indeed anomalous.

  • Doctor Science

    the President could nominate six persons of his/her own choosing, people of such sterling character that the Senate could not deny them.

    — and you all laughed bitterly, right? Talk about counterfactuals …

    • Sam Wang

      People in the room were pretty down on her overall description. The question, from a student, was somewhere between “what can be done?” and “what can be done that is least likely to fail?”

Leave a Comment