Richie and Hill argue that democratic justice requires proportional representation. Single-member district systems of election are, in their view, fundamentally defective for several reasons. They are insufficiently representative: they artificially reduce the options for voter choice, limiting the range of viewpoints present in the legislature, and emasculating political debate and discussion. This reduction in the range of viable candidates diminishes the representation of minorities and previously underrepresented groups and tends to reduce voter turnout as well. In addition, single-member district (SMD) systems permit and encourage officials to gerrymander electoral districts in order to increase their own job security, at the expense of making most legislative contests uncompetitive. Richie and Hill argue, moreover, that such systems cannot be reformed by regulating campaign finance, limiting incumbent control over redistricting, or redrawing districts with a view to fixing problems of “misrepresentation,” both because such reforms are politically infeasible and because they do not touch the root defects of SMD systems. The authors argue that there are proportional representation systems that would cure these defects, and that there is a politically practical path that can bring about a transition to such a system. Richie and Hill thus present three arguments: that SMD systems are incurably defective, that there are superior PR systems, and that the adoption of PR is politically practical.

I think it is possible to agree with many of the criticisms of SMD systems. There is no question that SMD systems limit the range of representatives that can be elected. It seems clear that racial and ethnic minorities and women are probably not elected in the numbers they would be under other electoral systems, and this fact of under-representation probably does contribute to limiting the kind of debate that takes place in American legislatures. It is important to emphasize, however, that other ideological or religious minorities are probably under represented as well and these voices would probably appear in greater abundance in PR elected legislatures. In this respect, one of the conditions for full legislative deliberation–a wide range of viewpoints–would be enhanced. What viewpoints would be expressed would depend, of course, on which interests organized themselves into parties for purposes of presenting candidates for election. It is also not clear that meaningful legislative deliberation would actually increase, however, since many of these new voices would likely be tied very closely to narrow ideological constituencies and may not be as willing or able to engage in political accommodation as are current legislators.

I think the authors’ suggestion that SMD systems depress turnout relative to PR systems is somewhat less convincing. Even if one grants that PR offers voters more of a chance to elect candidates they prefer, the magnitude of this effect appears to be very small. Moreover, it is not at all clear that the individual voter’s influence on the ultimate policy produced by the legislature is larger in one system than another. The cross-national comparisons offered by Richie and Hill seem unconvincing when we consider the many other confounding factors–that some nations penalize nonvoting, for example, or that the United States, with more than half a million officials, has many more elections than virtually any other modern nation–as well as the cultural differences between countries. Further, many SMD systems have very high turnout rates: the United Kingdom and France average about 75 percent, and New Zealand voters turned out at almost a 90 percent rate when that country employed SMD elections. Poland employs PR, has a wide range of parties in the legislature (including one called the Beer Drinkers Party), and has a lower turnout rate than the United States.

I also agree with the authors that many of the defects of SMD systems are intransigient and not very susceptible to reform. They are traceable more to the way the population is distributed geographically, rather than to the real but marginal effects of the incumbency advantage, campaign spending distortions, or gerrymandering. It is just hard to see many Republicans getting elected from central cities or Democrats from the non-urban south, almost no matter how the districts are drawn. Thus, even if campaign finance were to be reformed, gerrymandering eliminated, and term limits imposed, it would still be the case that SMD systems would tend to produce at most two viable candidates in most districts, and a high proportion of relatively uncompetititive seats. Given the geographic distribution of minorities, minority underrepresentation in legislatures would no doubt persist (although this is probably more attributable to historical biases in intraparty nomination processes than to the use of SMD elections), though it is less clear that underrepresentation of women would. Thus, if one is concerned, as the authors are with enhancing “descriptive representation” (making the legislature look like the electorate) or fuller legislative deliberation (making the legislature sound like the electorate), reforming the electoral system the direction of PR seems an attractive way to go.

I also agree with the authors that the political impediments to such a change are probably less formidable than they might appear. Single-member districting is not constitutionally entrenched, and within many states could be implemented through use of the popular initiative. Implementation at the congressional level would require only statutory change. And, while the authors might be right that a broad progressive coalition might be fabricated for these purposes, I think it as likely that many on the religious right and economic libertarians might find the idea just as attractive. I have rather less hope that a congressional majority, elected under the current system, would be swayed by such a movement–it will hardly appear to be the most pressing political issue even for the members of a reform coalition, since their attraction to PR will be largely instrumental–and I have similar doubts about the prospects for such reform in states without the popular initiative. But, in many parts of the country, PR-oriented reform might well be politically viable and perhaps worth experimenting with.

I am not yet convinced, however, that Richie and Hill have made a case for adopting PR all things considered, because they have not addressed the new problems that PR would likely produce. It is possible, of course, that it is worth adopting PR despite its weaknesses, but that argument needs to be made explicitly. The authors are remarkably casual in dismissing what I take to be the main problems with PR systems, saying that concerns about accountability, transparency, and governability are “insulting” to voters and “as objectionable as arguments against full suffrage.” They also cursorily dismiss the experiences of Italy and Israel with PR systems, suggesting without argument that these experiences are atypical. This failure to give arguments about the attractiveness of PR all things considered is unhelpful for two reasons. First, it is not clear that Israel and Italy are that atypical. One could list as well the French Fourth Republic and Weimar Germany as other instances in which PR systems may have produced circumstances of chronic constitutional conflict. Second, and more importantly, the refusal to analyze actual PR experiences limits our capacity to understand why some PR systems failed and how reformers in this country might avoid those fates. The framers of the German Basic Law and of the Constitution of the French Fifth Republic certainly had theories about what was wrong with the Weimar and Fourth Republic electoral systems, and they attempted to remedy these defects by adopting electoral systems that were less than fully proportional (the French adopted SMDs, while the Germans chose a mixed system with relatively high representational thresholds).

To me, the principal defect of PR is the weakness of electoral responsiveness–the relationship between electoral expressions of public opinion and public policy. PR systems tend to produce fragmented electoral outcomes in which no party wins a majority of seats and, for that reason, interparty coalitions are required to enact legislation (or to form governments in parliamentary systems). Electoral shifts in popular support for or against a party, in such circumstances, typically do not much alter the logic of coalition formation so that similar or identical coalitions may reform even after substantial electoral shifts. The continuous presence of the Christian Democrats in postwar Italian governments is only the most notorious example of this rather ubiquitous phenomenon. As a result, electoral results are often not reflected in policy outcomes (or government composition). In this sense, PR governments are “too stable.”

In another sense, however, the logic of coalition formation in multiparty systems is inherently unstable in the sense that governing coalitions can shift independently of electorally expressed preferences. The frequent breakdowns of governing coalitions in Fourth Republic France and postwar Italy (or Weimar, or Israel, or Poland) are again not atypical. Insofar as PR produces multiparty outcomes (and, as the modern German case indicates, this tendency can be limited by adopting representational thresholds or other institutional checks on PR), the problem of limited electoral responsiveness is real and can produce problems of legitimacy. I must say that I don’t see how it is insulting to voters to mention this. Such tendencies are simply operating characteristics of PR systems. Is it insulting to voters to point out the incentives of politicians to gerrymander in SMD systems? The voter’s could, if they chose, punish incumbents for such behavior.

Finally I want to return to the issue of the amount and kind of deliberation fostered by the two electoral arrangements. It is well to remember that, in the end, the legislature will only adopt one policy and that all those who dislike this policy will be disappointed at the stage of policy choice. PR systems, by increasing descriptive representation, bring more diverse voices into the legislature and require that political arguments about policies happen in the public legislative forum. However, by encouraging wide ranging debate to occur within a broadly representative legislature, PR encourages a politics of mobilization rather than persuasion within the electorate. This, indeed, is why Richie and Hill claim that PR will increase turnout. SMD systems, by contrast, force compromise and accomodation to occur within the electoral process, putting weight on persuading what the authors call “swing voters” (others use the term “median voters”) within the district, and produce a less diverse but possibly more tractable legislature. In effect, the legislative diversity in SMD legislatures is due to interdistrict differences whereas that in PR legislatures arises both from intradistrict and interdistrict variation. Policy choice is presumably easier (and more transparent) in such systems because many of the underlying social conflicts have already been implicitly accomodated earlier in the process.

There is little question that SMD systems have some real defects, especially with regard to the representation of minorities. But PR systems have weaknesses too and it is no accident that many countries have moved away from the purer forms of PR in reaction to particular unhappy experiences. I think that because PR systems promise to remedy some of the defects of SMD systems, it is particularly important that we experiment with the most attractive versions, ones that do not cause other, possibly worse, problems. Doing this requires paying careful attention to the experiences of other jurisdictions with such systems–yes, even Italy and Israel–in order to take advantage of the most attractive features of such systems while avoiding their weaknesses.