Parpolity and Indirect Elections
[Contribution to the Reimagining Society Project hosted by ZCommunications]
What sorts of decision-making institutions and practices would be appropriate for a good society?
In several essays, I have argued that that the main traditional answers to this question are seriously deficient in one respect or another, and that we need an alternative, one which I called parpolity.
The most widely endorsed political system is representative democracy, a system whereby people vote for other people -- representatives -- who rule in their names. Representative democracy has several important flaws.
First, it treats politics as strictly instrumental -- that is, as a means to an end, instead of a value in its own right. But political participation is intrinsically worthwhile: it gives people the experience of controlling their own lives. The more that the task of thinking about how we can collectively manage our lives is delegated to others, the less knowledgeable we become regarding our society, the less we determine our own destinies, and the weaker become our ties of solidarity to our fellow citizens.
A second problem with representative democracy is that representatives for many reasons don't in fact represent their constituents. Representatives say one thing to get elected and then change their positions once in office. They have no real connection to the hundreds of thousands of people they represent. Their different life circumstances lead them to develop different interests from those of their constituents.
We could, of course, "mandate" representatives -- that is, require them by law to to keep their campaign promises. But what happens when circumstances change? Surely, we don't want representatives to be compelled to carry out policies that new developments have made inappropriate or even harmful? Alternatively, we could mandate all representatives to follow the evolving wishes of their constituents as reflected in public opinion polls. But if we do this, then the representatives are rendered technically irrelevant. There is no need for representatives to study or debate the issues because it doesn't matter what they think. All that matters is that they vote according to their constituents' stated wishes. Mandated representatives could simply be replaced by a computer that compiles the opinions of the people and then votes accordingly. But this is really nothing more than a system of direct (referendum) democracy. So if representatives are mandated, they are irrelevant, and if they are not mandated then they will often not be truly representative of their constituents.
Advocates of representative democracy do make some legitimate arguments, however. They claim that it would take too much time for everyone to decide everything. This point is often exaggerated -- people's tolerance for meetings, for example, cannot be judged by their reaction to meaningless meetings today where they have no real power; nevertheless, it is true that not everyone has, or ever will have, unlimited time or enthusiasm for politics.
A second argument on behalf of representative democracy is that representative legislatures are deliberative bodies that debate and negotiate complex resolutions that fairly capture the essence of an issue, whereas the citizenry as a whole would be incapable of such fine tuning. They have to vote a ballot question up or down; they can't reword or amend, even though we know that the precise wording of a ballot question can often skew the results. This is a valid point, one which any alternative to representative democracy needs to take account of.
Direct democracy is an alternative to representative democracy. Under direct democracy people make decisions themselves rather than choosing others to do it for them. There are several variants of direct democracy. One of these is referendum democracy, where every issue is put to the population as a whole. In the past such an approach was simply impossible: there was no mechanism for allowing large numbers of people to cast ballots on a nearly daily basis. But modern technology makes this possible on a vast scale. People could use the internet first to access as much background information as they wanted and then to vote on their preferred options.
But even if technically possible, would we really want to spend all this time exhaustively studying the many hundreds of issues that national legislatures currently take up each year. Those legislators are doing this more or less full-time. Do we all want to invest that same amount of time (while doing some other job as well)? Legislators typically have a staff to make the work manageable. Would each citizen have a staff person? Clearly some means is needed to separate the important issues out from all the rather routine issues that legislators currently deal with.
Beyond this time problem, referendum democracy suffers from another defect: when people make decisions that do not emerge from participation in some sort of deliberative process, their off-the-cuff opinions are more likely to be intolerant and uninformed. Whereas deliberation encourages people to seek common ground and find ways to take seriously the opinions of others, voting in a referendum encourages people to express their pre-existing views on polarized positions.
A second type of direct democracy is where all decisions are made directly by the people living in fully autonomous small communities. Here we can combine the benefits of participation and the benefits of deliberation. But there are nevertheless serious shortcomings with this approach.
First, not all problems are susceptible to small-scale solutions. Pandemics call for a global solution. Environmental problems need a large scale response. Small communities cannot afford expensive technologies, like medical equipment. It is true, of course, that some large-scale technologies create great harm -- like nuclear power plants -- and that much technology is horrendously misused in current society to serve the interests of elites; but this is no reason for us to reject technology out of hand. Technology can reduce human drudgery and provide us the opportunity of undertaking more creative work and leading fuller lives.
Advocates of autonomous communities often reply that their preference for small scale does not prevent communities from cooperating, whether to address environmental problems or to share an MRI machine. But sharing and cooperating need some decision-making procedure involving multiple communities. Otherwise there will be no way to prevent an autonomous community from polluting its neighbors or hoarding medical equipment. And if we have procedures that can prevent the pollution or the hoarding, then the polluting or hoarding communities are no longer fully autonomous.
A second problem with small autonomous communities involves the question of size. Either they are too small, and thus can't function effectively or provide adequate diversity. Or they will be too large to permit face-to-face direct democracy. A meeting of thousands or even hundreds of people is not typically a very participatory experience.
A third type of direct democracy rejects both the self-sufficiency and the referendum models and instead has small councils, linked to one another.
Everyone gets to participate in a primary council that is small enough for face-to-face decision making and for real deliberation. Many decisions will be made in these councils because the decision affects only or overwhelmingly the members of that council. But because there are many decisions that affect more than the people in a single council, the councils affected will have to coordinate their decision-making. This means that councils will have to send delegates to a higher level council. And, if the decision affects more than one of these higher level councils, they would in turn send delegates to a third-level council. And so on.
How would these higher level councils operate? We don't want to have delegates mandated by their sending councils, for then the higher level councils will not be deliberative bodies. As noted previously, there would be no point to anyone speaking or trying to persuade others, or passionately explaining one's special concerns, because all the delegates would have zero leeway -- they would have to vote the way their sending council told them to. This means that no one from council A gets to hear the perspective of people from council B, and there is no possibility of coming to a better position than either A or B alone proposed. On the other hand, if the delegates are not mandated and just do what they want, then we have the problem of delegates becoming like the unrepresentative representatives that characterize contemporary representative democracy.
What makes more sense is to send a delegate who, because she has been part of a council and participated in a deliberative process with its members, understands their sentiments and concerns, and is authorized to deliberate on their behalf with other delegates. But what will prevent this unmandated delegate from becoming an unrepresentative representative? First, the connection between delegates and their sending councils is an organic one, not at all like the connection between constituents and representatives in typical representative democracies. The delegates are part of -- and constantly returning to -- their sending council. Second, delegates will be rotated; no one will be permitted to serve too long as a council's delegate. Third, delegates will be subject to immediate recall. If ever a council believes that its delegate no longer adequately reflects its concerns and sentiments (and all higher-level council meetings are videotaped and easily monitored), then it may immediately replace the delegate with someone else.
Most importantly, however, what prevents the unmandated delegates from usurping power is that the higher-level councils will only vote on matters that are relatively non-controversial. Whenever a vote is close (or when enough citizens or lower councils insist), the issue is returned to the lower councils for a decision.
It might be asked, why not send all issues back to the primary-level councils for a vote? But this is where our concern to avoid overdoing participation with excessive time demands comes in. By sending back contentious issues or those so requested by the citizens or the lower-level councils, we have a check on abuse of power by the delegates to the higher-level councils. But to send everything back would simply be a waste of time.
The notion of nested councils has elicited some confusion, with some seeing the layered councils, each sending delegates to the next higher level council, as simply a system of multiple indirect elections. Indirect elections, however, have serious deficiencies from a democratic point of view and therefore it is essential to understand the problems with indirect elections, as well as the problems of direct elections, and how parpolity seeks to avoid both sets of problems.
In a standard capitalist (representative) democracy, policymakers -- officeholders above the civil service grade -- are chosen in one of two ways. Some are elected and some are appointed. But the appointing is done by individuals or bodies that were themselves elected. Of course, this process can be repeated any number of times, so an official might be appointed by those who were themselves appointed by individuals or bodies who were elected. So rather than referring to elected and appointed officials, we might call them directly elected and indirectly elected.
How subject are these indirectly elected officials to popular control? Some of the factors that make them relatively immune from democratic control are not inherent in the fact that they were elected indirectly. So, for example, U.S. Supreme Court Justices are nominated by the President and confirmed by the Senate -- which is to say, they are chosen by elected officials (indirectly elected) -- but because they have life terms democratic accountability is remote. The public might find the views of a Justice selected decades earlier abhorrent, but they have no means of removing her during her lifetime.
But life terms are problematic from a democratic point of view even with elected officials. Various rulers have arranged to have themselves elected "president for life" -- typically these "elections" were bogus affairs, but even if they weren't, no electorate, no matter how democratic, should be able to bind future generations or even itself forever.
But what happens when officials who are indirectly elected serve only as long as the elected officials who appoint them, or roughly as long? Are these appointees subject to democratic control?
There are three reasons why democratic control would be relatively attenuated with indirect election compared to direct election.
First, there's the analogy to the children's game of telephone. You whisper something in someone's ear, who in turn whispers it into the ear of the next person and so on. By the time the message gets to the end of the line it typically bears little semblance to the original. And the more whisperers involved, the more the distortion. In a direct election, there's already a problem with the will of the voters being carried out: Representatives don't perfectly match the views of their constituents. But in an indirect election, there are more whisperers: the voters don't match the intermediate official(s) and the intermediate official(s) don't match the final appointee. Thus, the source of the divergence between the voters and the ultimate officeholder has multiple sources -- the divergence between voters and the first intermediate level of officials, and then the divergence between these intermediate officials and the final officeholder -- and will therefore be greater than if there had been no intermediate elections.
The second reason why indirect election does not allow adequate democratic control is related to non-unanimity. Say we have a political system where voters in ten districts each vote for a representative and these ten elected representatives then vote for a single person, A or B, to hold some office. If the voters within each district were unanimous in their desires, and the ten representatives all faithfully reflected the views of their constituents, then the officeholder they would choose would indeed be the choice of the people as a whole. But the situation is different when views within districts are not unanimous. Suppose that in each of seven of the districts the vote was 60% in favor of the representative wanting to choose candidate A and 40% in favor of the representative wanting to choose candidate B. In the other two districts, 20% of the voters favored the representative wanting to choose candidate A and 80% the representative wanting candidate B. When the ten representatives meet, they will vote 7-3 in favor of candidate A. But in terms of the wishes of their constituents, a majority, 52% (.7 x 40% + .3 x 80%) preferred candidate B. With the addition of further layers of indirect voting, the disparity becomes even greater.
The third problem with indirect election is that each layer of representatives doesn't just randomly diverge from those who elected them; the divergence is systematic. The representatives tend to have better skills than those who elected them. It is not surprising that representatives tend to be more outgoing, better speakers, and more politically aware than their constituents. This bias may not skew things all that much, but in an indirect election this bias operates at least twice, once at each election.
Indeed, this was the reason that indirect elections were established in the United States. The founders worried about "[a] rage for paper money, for an abolition of debts, for an equal division of property, or for any other improper or wicked project" (James Madison, Federalist Paper No. 10), and so some means was needed to minimize these dangers of popular rule. In at least two significant areas, the Constitution used indirect election to this effect.
First, an electoral college was established to choose the president: voters were to elect the electors and the electors were to choose the president. (Keep in mind that, as initially conceived, the electors were not, as they are today, generally pledged to voting for the candidate favored by the voters.) The purpose of this arrangement was to ensure (in the words of Alexander Hamilton in Federalist Paper No. 68) that "the immediate election" of the president "should be made by men most capable of analyzing the qualities adapted to the station..." Hamilton went on to say that "A small number of persons, selected by their fellow-citizens from the general mass, will be most likely to possess the information and discernment requisite to such complicated investigations." This system, said Hamilton, would "afford as little opportunity as possible to tumult and disorder" -- which is to say, to the possibility that the views of the unrefined public might prevail.
The second major indirect election provided for in the Constitution involved the U.S Senate. Until changed by the 17th amendment in 1913, U.S. Senators were not directly elected by the people of the states, but by the (elected) state legislatures. This undermined the will of the people in two ways: one, the Senators had a longer term of office than the state legislators who chose them, and, two, they are indirectly elected. In Hamilton's words (Federalist 27) because Senators were to be chosen "through the medium of the State legislatures which are select bodies of men ... there is reason to expect that this branch will generally be composed with peculiar care and judgment."
That indirect election reduces democratic control has been confirmed empirically. For example, a study found that in the United States elected regulators of electric utilities are more likely to pursue pro-consumer policies than appointed regulators.
But if indirect elections have problems from the point of view of democracy, so too do direct elections. Democratic theory requires that the electorate keep tabs on their representatives: are the representatives acting and voting as their constituents want them to? If citizens can't answer this question, then they can't punish the representatives at the polls, they can't pressure their representatives, and the representatives have little incentive to follow their constituents' wishes.
But keeping tabs on representatives is not easy to do. It is particularly difficult to do in large electoral districts -- where there is a greater number of voters per representative. And direct elections tend to have larger districts than indirect ones for the same end position. (For example, today a directly-elected U.S. senator is elected by all the voters of the state; before the 17th amendment, the indirectly-elected senator was elected by a hundred or so state legislators, each one responsible to a few percent of the state's voters.)
Few citizens read the Congressional Record regularly to check up on their representatives. Few citizens wade through the roll call vote records to determine whether their representatives have voted "the right way." To be sure, various organizations can compile such data to make it easier for the public to assess their representatives' records, but it is still difficult for the average citizen to keep up with anything but the most significant votes.
This is not -- as is often claimed -- a result of voter stupidity. The connection between voters and representatives is quite remote: in the United States a member of the House of Representatives has some 650,000 constituents and, in most states, senators have an even larger constituency (36 million plus in California). Many U.S. voters don't even know the names of their congressional representatives, but this is likely a result of voter powerlessness rather than a cause of it. The average person can master lots of information (think about celebrity gossip or sports statistics), but she has to be connected to that information in some way -- it has to make a difference -- something that's lacking when it comes to political representation on this scale. Why take the effort to learn about your representative when you're not going to be able to use that information to effect meaningful change? At best, you can vote against a bad representative at the next election (if she stands for re-election), but the odds of defeating an incumbent are small. (And the larger the constituency size, the greater the advantages of incumbency.) In short, citizen ignorance is "rational" in a U.S. Congressional election: in a system where real decisions are made by remote politicians, there is inadequate incentive for citizens to make themselves informed.
The larger the size of a constituency, the more difficult is face-to-face contact between representative and citizen, and thus the need for money for media campaigns and so on, which ultimately reduce citizen efficacy.
It is because of the powerlessness of citizens in huge constituencies that many Indian political reformers, following Gandhi, have suggested a decentralized, multi-tiered, bottom-up political system to replace the current system where the elected representative is far removed from the people -- "an average of one million voters for each Lok-Sabha constituency spread over a large geographical area." Currently, to influence the choice of such a large number of voters, candidates resort to money and muscle power, both of which undermine democracy. The result is that the voters "are in no position to hold the candidate accountable nor does the candidate consider himself accountable to these people."
A study of U.S. senate elections before and after the passage of the 17th amendment found that
"The amendment clearly made senators responsive directly to state electorates, so their selection and accountability once in office were based on a democratically stronger standard. A the same time, the amendment made senators answerable to relative political novices, so they could not be held to that standard (whether through electoral sanctioning or selection) as tightly as they were held to their preamendment standard."
In short, both before and after the amendment -- under both indirect and direct elections -- democracy was not well served. The political system of a good society must avoid the disadvantages of both.
Unlike typical direct elections, a good political system must give people an organic connection to those they elect so they can adequately monitor their performance and remove them when necessary. There cannot be large or remote constituencies that render monitoring impossible or even burdensome.
Unlike typical indirect elections, a good political system must ensure that the people's will does not get attenuated through each intermediate level of voting.
The way in which parpolity achieves these dual goals is through the system of nested councils, with delegates to each council level known personally to the council members who elected them. This maintains the organic connection missing from so many direct elections. But at the same time, there are various mechanisms to assure that the will of the people actually emerges. The most significant mechanism is that a petition signed by a given number of people or primary councils can always assure that an issue is returned to the primary level councils (of which every citizen is a member) for a vote. (The petitions, of course, would be electronic, and the number of signatures needed would be relatively small.) Additionally, a higher level council can always decide to send an issue to the primary level for decision. This will make sense whenever the issue is contentious and at all close.
As noted previously, to send every issue down to the primary councils for decision would be exceedingly inefficient. Most people are happy to let minor, non-controversial issues be decided by their representatives -- so long as they have the ability to have their say when they want to.
Other mechanisms to prevent the nested councils from undermining democratic control include: delegates return frequently to their sending councils, any delegate is subject to immediate recall by the sending council, delegates are rotated, and all deliberations and decisions of higher level councils are recorded and accessible to members of lower level councils.
Majorities and Minorities
One of the biggest dilemmas in designing political institutions involves the relationship between majority rule and the rights of minorities. Obviously we reject a political system where the rich or the well-born or the meritocratic rule over the majority. So in that respect, of course we favor majority rule. On the other hand, we would also obviously reject any political system that allowed the majority to exerminate or enslave or oppress the minority. Moreover, we would consider it improper for the majority to tell the minority what they could read or say or believe. Indeed, there are many things we think majorities should not be allowed to do. In many countries there are constitutions that spell out people's rights, which are really no more than restrictions on what the majoritiy can do. (My right to free speech is the same as the First Amendment guarantee that Congress -- that is, the embodiment of the majority -- "shall make no law ... abridging the freedom of speech....")
A good society ought to have restrictions on the majority, entrenched in some sort of constitution. But, of course, no brief document is going to be able to provide a full elaboration of what these rights entail. Does free speech include pornography? Racial insults? Libel? These sorts of interpretations are typically left up to courts. So in the United States it's the Supreme Court that ultimately decides the actual boundaries of free speech and of other constitutional provisions.
But this returns us to the question of how the Justices of the Supreme Court -- or on comparable judicial bodies in other countries -- are to be chosen. If they are chosen by indirect election and have long or even life terms, then they are removed from majority control. But why should nine individuals, only indirectly elected, who may have taken office when the public had very different views, get to impose their opinions on the rest of us?
On the other hand, if the Justices were chosen by direct election, would they adequately protect minority rights? One can imagine a candidate for Supreme Court Justice running for office on a platform denying rights to criminal defendants or atheists or gays. If the purpose of the Supreme Court is to prevent the majority from oppressing the minority, wouldn't an election by the majority undercut this purpose? Wouldn't an elected court be subject to the same passions and intolerances that permeate the majority? When the U.S. Congress, elected roughly by the majority, voted to ban flag-burning, wasn't it a good thing that there was a Supreme Court not up for election that could declare that law by Congress to be unconstitutional?
There have been various proposed solutions to this dilemma of reconciling majority rule and minority rights.
One approach has been to reject majority rule entirely, insisting that all decisions be made by consensus, which would mean that a single dissenter (or perhaps a very small number of dissenters) could block passage of any law.
Because parpolity emphasizes democratic deliberation, consensus should always be sought. However, to insist on consensus in every case is ill-advised. It is sometimes said that even a large group should be forced to respect and acknowledge the sentiments of a single dissenter who feels strongly on an issue. Respect and acknowledgment are fine; but the question is whether the strong feelings of the one dissenter should invariably be able to block the equally strong feelings of everyone else. To allow the lone dissenter to block action is to deny the overwhelming majority ultimate authority to decide their own fates. There is nothing magical about 50 percent plus one, but it does deserve more moral weight than 50 percent minus one.
Another approach to the majority-minority dilemma has been to insist on absolute parliamentary supremacy: that is, that anything the majority wants to do should be permitted. The argument here is that the protective benefit of Supreme Courts is overstated. Most of the time, the Court reflects rather than checks the passions and prejudices of the majority. (So, for example, in 1986 the Supreme Court upheld state laws criminalizing homosexual relations between consenting adults; when the Court reversed itself in 2003, six out of ten Americans agreed that homosexuality should be legal.) And historically the minority to which the Court seems to have shown the most deference has been the rich. Nevertheless, the long history of majoritarian discrimination against racial, religious, sexual, and other minorities should leave us uneasy about unchecked majority rule.
Parpolity's approach to this dilemma is analogous to the jury model. Choose a small group at random from the population to constitute "council courts." These courts will review decisions made by councils to see if they interfere with basic rights and constitutional protections. Each level council above the primary-level will be assigned a court, with the court assigned to the highest level council being the High Council Court. Like current-day juries, these courts will be deliberative bodies, though unlike juries they would have a term longer than a single case -- perhaps staggered two year terms. As a cross-section of the population, these will be democratic bodies: democratic bodies serving to check the democratic councils. The logic here again makes use of the finding, noted above, that when people make decisions through a deliberative process, the result is likely to be less intolerant than a simple poll of public opinion.
Redesigning political institutions alone cannot assure a decent political system. A good political system needs a good economic system. Put another way, without an economic system that is equitable, democratic, and partipatory, no political system will be able to offer the values we seek. But if we have a partiicpatory economy, with equity in terms of gender, sexuality, and national, ethnic, an religious differences, then parpolity seems to me the appropriate political structure.
Nested councils offer the possibility of a political system that promotes democracy, participation, and equality. But nested councils can only achieve these goals if issues can, whenever desired, be returned to the primary council level for decision.
1. See "Political Vision: Making Decisions in a Good Society," Z Magazine, Oct. 2004, pp. 42-48; "ParPolity: Political Vision for a Good Society," ZNet, Nov. 22, 2005; "Parpolity: A Political System for a Good Society," in Real Utopia: Participatory Society for the 21st Century, ed. Chris Spannos, Oakland, CA: AK Press, 2008, pp. 25-31. The first half of this essay draws from these other works.
2. See James S. Fishkin, The Voice of the People, New Haven: Yale University Press, 1995.
3. For a discussion of the some of the complications of these terms, see Benjamin Akzin, "Election and Appointment," American Political Science Review, vol. 54, no. 3 (Sept. 1960), pp. 705-713.
4. They can be impeached, of course, but this is supposed to be a remedy only for departures from "good behavior," not for diverging from the will of the people.
5. In the United States, some states have elected judges. Some others have appointed judges where they must be reconfirmed after a fixed period. In New Jersey, for example, "Justices of the Supreme Court and the Judges of the Superior Court shall hold their offices for initial terms of 7 years and upon reappointment shall hold their offices during good behavior..." (NJ Const., Art. VI, sec. VI, paragraph 3).
6. A third indirect election in the Constitution is one that remains in place today: Supreme Court justices are not elected by the voters, but appointed by the president with the advice and consent of the Senate.
7. Today the problem with the Electoral College is only trivially that any particular elector might not vote for the candidate desired by the voters. The so-called "faithless elector" problem is relatively rare and has never made a difference in the outcome -- though it could. Much more serious is that (in every case except two) all of a state's electoral votes are assigned as a block to the candidate who received a plurality of the state's popular vote. So a one-point difference in the vote total in California gives a candidate all 55 of California's electoral votes. This leads to the result that in principle one can be elected U.S. president by winning the 11 most populous states by a single vote in each and not taking a single vote anywhere else in the country. On several occasions, most recently in 2000, the candidate who (was said to have) won failed to receive a plurality of the vote.
8. Timothy Besley and Stephen Coate, "Elected Versus Appointed Regulators: Theory and Evidence," Journal of the European Economic Association, vol. 1, no. 5 (Sept. 2003), pp. 1176 -1206.
9. Ranging from 950,000 people per representative (Montana) to 523,000 per representative (Wyoming). See Wikipedia, "List of U.S. states by population," accessed 7/13/09.
11. See, for example, Michael Krashinsky and William J. Milne, "The Effects of Incumbency in U.S. Congressional Elections, 1950-1988," Legislative Studies Quarterly, vol. 18, no. 3 (Aug. 1993), pp. 21-44.
12. Edward L. Lascher, Jr., "Constituency Size and Incumbent Safety: A Reexamination," Political Research Quarterly, vol. 58, no. 2 (Jun. 2005), pp. 269-278.
14. Sean Gailmard and Jeffery A. Jenkins, "Agency Problems, the 17th Amendment, and Representation in the Senate," American Journal of Political Science, vol. 53, no. 2 (2009), pp. 339-40.
15. Texas v. Johnson, 491 U.S., 397 (1989), a 5-4 decision.
16. For a more detailed discussion of this matter, see my "ParPolity: Political Vision for a Good Society," ZNet, Nov. 22, 2005, section 5.7.
18. Frank Newport, "Six in 10 Americans Agree That Gay Sex Should Be Legal," Gallup News Service, June 27, 2003. Following the 2003 Supreme Court decision there was a slight backlash.
19. Fishkin, Voice of the People. For more details on the council courts, see my "ParPolity: Political Vision for a Good Society," ZNet, Nov. 22, 2005, section 6.
Stephen R. Shalom teaches political science at William Paterson University in New Jersey. Among his books are Which Side Are You On? An Introduction to Politics and Socialist Visions. He works with ZNet and is on the editorial board of New Politics.