The Inadequacy of Election Reform

With the modern assumption that elections are the very essence of democracy, each specific electoral problem engenders a reform proposal (or menu of reforms) intended to remedy it. Elections themselves, however, are never questioned. Before considering a radical concept such as sortition, it is prudent to first examine whether the failings of American democracy can be fixed using any of these commonly proposed solutions. In this chapter, I will discuss several of the most widely recognized specific problems with the American election process, present some corresponding reforms that have been proposed, and examine their likely efficacy (or lack thereof). While many reforms certainly have some merit, we will see how they all fall woefully short of fulfilling the democratic vision.

There is a profusion of reforms that have been proposed to improve the way Congress and other legislative chambers are elected. Many of these reforms also deal with changing the incentives created by the way elections are conducted so as to affect the functioning of legislative bodies. The reforms discussed below are not intended to be an exhaustive list, but are illustrative.

Term Limits

Incumbents who seek re-election almost always win. Of course, a few incumbents who have been caught in scandal, or have become unpopular for other reasons, may choose not to run again in order to avoid suffering the embarrassment of defeat. Typically, around 98 percent of members of Congress who seek re-election succeed. On occasion, there are elections in which news analysts talk of a “red wave” or a “blue wave”, with many incumbents being turned out and the House or Senate changing partisan majorities. But such “sea change” elections, such as those in 2012 and 2018, still typically see approximately 90 percent of incumbents who run being re-elected. The longevity in office of elected politicians is sometimes blamed for corruption, a distancing from average citizens, and a counter-productive focus on re-election. Enacting term limits has been offered in response.

Term limits were historically seen as a useful defense against tyranny at the time of the nation’s founding. The 1776 Pennsylvania Constitution mandated term limits, as did the Articles of Confederation. The Federal Constitution contained no term limits (though a presidential term limit was added by the ratification of the Twenty-Second Amendment in 1951). Term limits for state legislators were adopted in over a dozen states and in many local jurisdictions toward the end of the twentieth century. Imposing term limits on members of Congress would require a constitutional amendment.

Term limits at the level of the state offices, it is hoped, could reduce the long-term concentration of power and thus the potential for abuse. Researchers Michael Smart and Daniel Sturm found that this perceived diminution of accountability indeed allows incumbents to promote their sincere preferences, for good or for ill (Smart and Sturm 2013). In a study by Yacov Tsur, entitled “Political tenure, term limits and corruption”, the author found that the more rapid turnover resulting from term limits affected “the learning curve of incumbents in conducting their regular task (public goods provision) and the learning curve of opportunistic incumbents in becoming effective embezzlers. I find that stricter term limits increase the frequency of corruption incidents but reduce the expected cost per incident” (Tsur 2022).

Term limits may promote a sort of mentoring process for both good and ill, as term-limited incumbents promote potential successors and in turn simply seek another, often higher, office for themselves.

One argument for term limits is the notion that it would reduce the tendency for “over-spending” by legislators seeking to “bring home the bacon” to their districts to buy votes for their re-election. However, there is no evidence of this effect. A study by H. Abbie Erler entitled “Legislative Term Limits and State Spending” reported that term-limited states actually tend to spend more than other states (Erler 2007). While term limits remain popular with voters in the states that have enacted them, studies have shown that term limits do not increase the number of women, nor generate a more representative body overall (Carroll and Jenkins 2002).

Improving Voter Turnout

As noted above, the United States has some of the poorest voter turnout rates in the world. Typically, between 10 and 20 percent of eligible people vote in local elections (and in many municipal elections, it drops into single-digit percentages), around 40 percent vote in congressional election years, and only in presidential years is it common for over half of the eligible people to cast ballots.¹ Many US election reformers, especially those on the left, focus on ways to make voter registration and casting a ballot easier, although many on the right express concern about unsupported allegations of non-citizen voting and voter fraud.

The reform strategies have sought to make registration easier and include same-day voter registration, or universal automatic voter registration, which is the norm in most electoral democracies. However, even if registered, many pro-spective voters fail to vote in most or all elections. So, in addition to making registration either more convenient or automatic, many reformers focus on making voting itself more convenient.

To make the voting process easier, reformers have proposed extending the voting period, making election day a holiday, and, most recently, voting by mail. The COVID-19 pandemic made vote by mail, including ballot drop-off boxes, the standard rather than the exception in many jurisdictions. Some states automatically mail voters a ballot, without a request. Researchers found that universal vote by mail resulted in only about a 2 percent increase in voter turnout, however (Thompson et al. 2020).

It is now apparent that institutional barriers to registering and to voting are almost peripheral to the real causes of low voter participation in the US² Curtis Gans, director of the Center for the Study of the American Electorate at American University, was widely considered to be the “go-to guy” for issues related to voter turnout in the US. In a press release accompanying his analysis of election turnout in the 2008 presidential election, he concluded that the participation problem is, at heart, not procedural but motivational. In that November 6, 2008, press release about the limits of “convenience voting” measures, Gans wrote: “It has always been abundantly clear that, after four decades of making it easier to vote and having turnout decline (among most groups) except for elections driven by fear and anger, the central issue governing turnout is not procedure but motivation” (Gans 2008).

A crucial question is whether higher participation among marginalized citizens, for example, would change the narrow assortment of candidates they get to choose from. In the vast majority of cases, candidates of both major parties are from the same income, class, etc., and there is little evidence that increased voter participation results in more representative legislatures. Voter participation rates, by themselves, tell us very little about the strength of a democracy. Authoritarian regimes often orchestrate elections with nearly 100 percent turnout.

Campaign Finance Reform

Campaign finance reform is another widely proposed remedy for our unrepresentative legislatures. In a nutshell, it takes resources to run winning campaigns, and only certain sorts of people have ready access to the necessary money.³ This restricted pool of potential “serious” candidates is thought to be a fundamental reason legislatures are so unlike the population they represent.

However, that crucial dynamic is generally overshadowed in the public’s mind by the corrupting potential of the fact that politicians have a powerful incentive to cater to monied interests that can fund their campaigns. This corruption is well known in general terms, though the countless specific examples are essentially ignored or unknown by voters. A typical explanation of this by political scientists Bernard Manin, Adam Przeworski, and Susan C. Stokes, goes: “Because voters do not care about policies that have only a small impact on their welfare, politicians can sell policies to interest groups that inflict only a small cost on each individual voter but which concentrate benefits on the particular interest groups and spend on electioneering the resources contributed by interest groups in exchange for these policies” (Przeworski, Stokes, and Manin 1999). Voters may generally despise the influence of money in politics, but relatively few make an active effort to address it.

If the corruption of the campaign system is ever fixed, I expect people will look back at this time with amazement at how any population put up with this legalized bribery. There are limits on the size of contributions directly to candidates and political parties. However, since political advocacy is protected speech under the First Amendment, there is no limit to nominally independent spending (not coordinated with the campaign). While campaign finance reformers focus on the 2010 Supreme Court ruling Citizens United v. Federal Election Commission, that decision only expanded the problem by granting corporations and unions the same right to unlimited election-related spending that natural persons have. Even getting the court to reverse that one ruling would have no impact on election spending by individuals, such as the hundreds of millions spent by Elon Musk in 2024.

One reform that has received significant attention is public campaign financing (in exchange for swearing off large donations), to free politicians from going hat in hand to big money interests.

There are many variants, requiring candidates to raise a large number of small qualifying contributions (as in Maine)⁴ or providing all adult citizens with “democracy vouchers” that they can give to candidates of their choice (as proposed by reformers such as Larry Lessig). These reforms still tend to favor candidates who have access to enough money or celebrity to be well known prior to launching a campaign. It could, however, reduce the importance of wealthy backers and theoretically allow for a somewhat wider range of candidacies. The hope that public financing would reduce politician corruption is plausible, and some international studies lend credence to the notion (Hummel et al. 2021). However, publicly financed candidates in the United States face the reality that their spending can be dwarfed by nominally “independent expenditures” by the wealthy and special interests. Thus, public financing may allow more people to enter the political fray, but it cannot level the playing field.

While the hope of reducing corruption is worthwhile, it does not, by itself, lead to better representation. There is no evidence that public financing results in more representative chambers. An analysis was done of:

public financing programs in Arizona, Connecticut, and Maine to causally identify the effect of legislators’ funding status on how closely they represent constituent preferences. Using multiple identification strategies, we show that candidates who exclusively use public campaign financing are more extreme and less representative of their districts than non-publicly financed candidates. (Kilborn and Vishwanath 2021)

An international study found that public financing of political campaigns

disproportionately aids radical-right parties. This is a counter-intuitive result, as public financing of elections has recently become an important issue for those on the political left as a method to restore power to the majority. Rather, it seems to aid not only those on the right, but those on the political fringe. Through a cross-national analysis of 328 lower-house elections in 20 advanced democracies, I demonstrate that public financing serves to significantly increase vote-share of the radical-right. (Bichay 2020)

Interestingly, however, the study also found that “these same effects are not shared by extreme-left parties. Due to important and unique characteristics of far-right parties, public campaign financing seems to have a disproportionate and exclusive effect on them”. Most of these countries use some form of proportional representation voting method, so the relevance of this finding in the United States is unknown.

After the election, rather than corruptly rewarding big donors or, at the very least, preferentially taking their calls and input, legislators elected with public financing would presumably have more time to engage with ordinary constituents. Former US Senator Ernest Hollings, in a 2006 Washington Post opinion piece, estimated that nearly one-third of a Senator’s time is spent fundraising (Hollings 2006). Since it is illegal to use government offices for fundraising, both major parties maintain offices with phones nearby and encourage members of Congress to spend approximately 30 hours per week fundraising, either “dialing for dollars” or traveling to fundraising events. Publicly financed legislators wouldn’t have to spend a huge percentage of their time raising money for the next campaign.

But even public financing is no panacea when it comes to forming a legislature that is genuinely representative of the population. One might hope that more working-class people would become legislators if access to big money were removed as a prerequisite, but that, unfortunately, does not appear to be the case. Mitchell Kilborn studied this issue and writes,

Using occupational data on state legislative candidates in public financing state Connecticut and two paired control states to execute a difference in difference analysis, I demonstrate that when public financing is available, fewer low SES [socio-economic status] candidates run for state legislative office, and those who do run are not more likely to win and are less likely to utilize public financing. (Kilborn 2018)

In Maine, which has led the nation on public campaign finance, over 80 percent of legislators ran using public financing. But National Conference of State Legislatures (NCSL) data showed that Maine’s legislators are no more likely to have working-class occupations than legislators from similar New England states.

Although census data indicates that less than one-third of Maine’s workforce is in higher-income occupations, NCSL data showed that around two-thirds of Maine’s legislators who are below retirement age come from this class (there is no previous occupation information for legislators who are retired). Fixing the money problem alone might have some effect, but would not remedy the countless other biases that are ubiquitous in American legislative elections, such as gender and age distortions.⁵ Public financing of election campaigns in Maine has shown no impact on gender bias. The state does not stand out in terms of the number of female politicians either, and males still predominate.

Approaching the same subject from another direction, the winners of State House races in my small state of Vermont often spend less than a thousand dollars. Money is simply not a substantial impediment to running a low-level winning campaign in Vermont. Yet, as I noted earlier, when I examined the housing status of my fellow Representatives, I found an almost total absence of renters. This probably reflects a built-in bias in regards to what type of people are willing to run for office in terms of such things as class, self-confidence, and employment situation (affording them the ability to take time off work for campaigning, for example). Making public campaign funds available would probably make no difference—taking time off from work in order to campaign would still be unaffordable for most people, even if they had the desire. No public campaign finance reform proposal includes stipends for living expenses during the campaign. Using campaign money for such “personal” purposes is expressly illegal under all such proposals.

Campaign finance reform may have a beneficial impact on corruption, but the evidence so far indicates it will not create more accurately representative legislatures.

Redistricting

An almost universally acknowledged flaw with the American electoral system is rampant gerrymandering—the redrawing of district boundaries for political advantage, and so-called “safe-seats”. Although this has traditionally been a decennial practice tied to the census, in 2025 President Trump called for Texas to re-gerrymander its Congressional districts, prompting a state-by-state, tit-for-tat partisan gerrymandering spree using out-of-date census data.

As a legislator and city councilor I participated in drawing and redrawing district boundaries numerous times, and am well aware of how cleverly the party in control can stack the deck.📝Inside Story The art of drawing favorable district lines has evolved from an art, when the term was first coined in Massachusetts in 1812, to a computerized science.

Sophisticated computer programs now allow layering all kinds of census, demographic, survey, and internet-mined metadata onto district maps in order to tailor districts to legislators’ re-election interests. The legislators, in effect, get to pick their voters rather than the voters getting to pick their legislators. Gerrymandering is almost a uniquely American problem tied to the fact that the US mostly uses single-seat districts. Most modern elective governments use proportional representation with multiple seats per district, making district boundaries and gerrymandering non-issues.

Arizona reformers passed a constitutional amendment in 2000 to eliminate gerrymandering using an independent redistricting commission. With legislators removed from the process of drawing their own district lines, and with lofty goals and guidelines, the hope was that elections would be more competitive. However, the elections under the new districts were even less competitive than before, with incumbents from both parties winning every single State Senate race in which they ran. In a study published in 2018, researchers evaluated

whether maps drawn by independent commissions are more electorally competitive than those produced by party-controlled legislatures … counter to expectations, we find that independent redistrictors produce virtually the same degree of insulation as plans devised in legislatures or by politician commissions. Overall, our results suggest caution in overhauling state redistricting institutions as a mechanism to increase electoral competition. (Henderson, Hamel, and Goldzimer 2018)

Because like-minded people tend to live near each other (what author Bill Bishop and others refer to as “the big sort”), if Arizona were to elevate having competitive races as the top priority, it would have to give up other priorities, such as compact districts that respect communities of interest, as well as risk a far less representative legislature. Other states have also pursued independent redistricting commissions since then, with Colorado amending its constitution in 2018 to establish a 12-member commission (four Democrats, four Republicans, and four unaffiliated) to approve redistricting plans.

As final evidence that gerrymandering is not the root of unrepresentative legislatures, one need look no further than the US Senate, whose district boundaries, state lines never change, and thus cannot be gerrymandered. Gerrymandering cannot explain the Senate’s mal-representation. Competitive elections are seen as essential to voters’ ability to hold legislators accountable. It turns out that, while gerrymandering to create non-competitive districts is a slimy process, it also is not the core reason that we have non-representative legislatures.

Reforming Primaries

Polarization between Democratic and Republican legislators has increased over recent decades. So-called “moderate Republicans” and “boll weevil Democrats” of the 1960s (when some Republican lawmakers were to the left of some Democrats, and some Democrats were to the right of some Republicans) are gone. One reason often cited for this trend is the party primary system of nomination.

Most Congressional districts are partisan safe-seats, with no meaningful contest in the general election. The party primary is therefore the key election, in which name recognition and campaign cash can be decisive. Yet, voter turnout in primaries is a fraction of the already low turnout of general elections. Since so few citizens vote in primary elections, incumbents often seek to placate the more extreme adherents within their party who are more likely to vote in primaries, in order to avoid a primary challenge. The influence of Tea-Party-Republicans, and more recently, Trump-Republicans, and Black-Lives-Matter-Democrats are often linked to this factor. Extremist wealthy campaign spenders, such as Elon Musk, can also simply threaten incumbents in safe districts with being “primaried” (used as a verb).

Some reformers have laid the blame for polarization on the primary nomination system, and have promoted “top-two” primaries, also called a “non-partisan open primary”, as a remedy. This reform essentially eliminates party primaries and puts candidates from all parties into a first-round election, with the top-two plurality leaders advancing to the general election. Thus, there could be two candidates from the same party in that election, with no other party included. Alaska, California, Louisiana, Nebraska, and Washington State currently use some variant of this system. Political scientists who carefully examined the impact on polarization of California’s non-partisan redistricting commission combined with its top-two primary found that “California’s electoral experiments did not bring their hoped-for effects. If anything, legislators strayed further from their district’s average voter in 2012” (Kousser, Phillips, and Shor 2016)

Although the notion that primary voters are more extreme is widely assumed, studies have shown that this is not actually the case (Sides 2013). Another study concluded that for the top-two reform to work to mitigate polarization, voters would need both access to fine-grained knowledge about various candidates within each party and sufficient motivation to devote the needed time to deeply study candidate distinctions. There is no evidence that voters employed either of these needed elements (Ahler, Citrin, and Lenz. 2016).

Some states today, including my home state of Vermont, have a less extreme form of “open primary,” in which registered voters, regardless of party loyalty, are allowed to vote in any single party primary A national study conducted by a team of political scientists, led by Eric McGhee, published in 2013 found that “the openness of a primary election has little, if any, effect on the extremity of the politicians it produces” (McGhee et al. 2014).

Proportional Representation

“Single-seat winner-take-all plurality voting” is a bit of a mouthful.⁶ It is the voting method used in the United States, Canada, the UK, and a handful of other countries. It is very much the exception among developed countries around the world. This method of voting is prone to the “spoiler” problem if there are more than just two candidates in a race. Voting for a minor candidate, whom the voter prefers, may split the majority and inadvertently help elect a candidate with less than 50 percent of the vote, whom a majority of voters may agree is the worst choice. This dynamic discourages voters from considering minor candidates and discourages third-party candidacies. So, this generally results in domination by just two major parties within a jurisdiction. Polls show that most Americans are dissatisfied with the two major parties and would prefer more choice on the ballot, but the plurality voting rules, which generate the spoiler problem, make this problematic.

Voting reforms such as ranked-choice voting (instant runoffs), approval voting, Borda count, Condorcet, score voting, and other methods have been proposed to alleviate the spoiler problem, with ranked-choice voting reform actually making significant headway in a number of states. Such voting reforms may allow for more candidates on the ballot with less concern about the spoiler problem, but they fail to elect a more accurately representative chamber overall. Being all single-seat voting methods, under these reforms, minorities would still be under-represented in legislative bodies. The obvious response is to instead institute some proportional representation voting method, which is what all but a handful of modern electoral democracies do.

First developed in the nineteenth century, proportional representation (PR) seeks to assure that the majority can elect a majority of a legislative chamber, but also that substantial minorities can elect a fair share of seats proportionate to their support. Under winner-take-all systems, everyone may have the right to a vote, but not the right to representation. The basic principle of PR is that the majority gets to decide, but everybody has a right to be represented at the table. This typically means having multiple parties in a legislature.

While PR allows for much more accurate partisan representation than the usual winner-take-all election rules used in the US, it isn’t enough. All voting methods, including PR, provide voters with a relatively narrow partisan menu of choices and assume that voters choose candidates or parties based on either policy positions or fundamental interests, rather than, for example, technocratic competence. However, incumbent governments are often voted out of office for simple competency reasons, rather than policy direction. An economic downturn, high unemployment, crime, a spate of shark attacks, etc. (which may be outside the control of any ruling party), can result in a change of government, even if the majority of voters oppose the policies of the replacement coalition of parties. But even if voters did rely solely on their policy preferences in deciding how to vote, and voters had complete knowledge of party platforms and the candidates’ records (which can never happen), this is still unworkable. Here’s an example: A retired, pro-life, Hispanic, Gay, economic libertarian voter who also favors taxing the rich and gun rights has no ability to consistently “vote his interests”, even if there are several parties to choose among. While Marxists may assert that class interests are the only fundamental interests, most voters have “false consciousness”—or simply disagree with that view, and make voting decisions on other bases.

A potential danger of pure PR, noted by supporters of winner-take-all election methods such as the American two-party system, is the risk of polarized pluralism, where otherwise centrist parties reach out to the political fringes for allies, leading to irreconcilable intolerance that can put electoral democracy itself at risk. Defenders of a two-party system used to point out that it encourages divergent interests to work out compromises prior to the nomination of candidates, in the hopes of cobbling together a voter coalition capable of winning the election. This is the basis of the “big tent” adage. This analysis, however, has become less and less valid in twenty-first-century America, with a roughly evenly divided polarized electorate.

Proportional representation typically leaves this coalition formation until after the election. This also shifts the deal-making into the hands of party insiders, rather than voters. Maurice Duverger describes this tendency: “By dispersing voters among numerous independent parties, PR prevents the citizens from expressing a clear choice for a governmental team. It transfers the choice to the party leaders” (Duverger 2003, 78). PR may help elect a legislature that somewhat more reasonably reflects the diversity of policy preferences of the public, but which policies will actually be pursued and which jettisoned by the ruling coalition isn’t known until the party leaders hash it out after the election.

While there may be more points of view represented in these negotiations than in winner-take-all election situations, it is a small set of elite party leaders and their sponsors who call the shots, rather than the mass of voters. As Benjamin Barber, the author of Strong Democracy, wrote: “PR may multiply the number of special-interest and one-issue political groups represented in the adversary system, but it can only fragment the citizenry and further impede efforts at public seeing” (Barber 2009, 206).

Proportional Representation and Women

In addition to electing politicians from more than just two political parties, PR can also elect far more women legislators. Nearly every advanced country using PR has a far higher percentage of female legislators than does the US. Some countries also have laws regulating the number and placement of female candidates on each party’s list of nominees as a form of affirmative action. One such procedure is referred to as a “zipper list” that requires parties to alternate male and female candidates on their nomination lists.

Zahra’ Langhi, the co-founder of Libyan Women’s Platform for Peacespeaking, at a TEDx-Women Talk in 2012, spoke about the aftermath of the elections that followed the Libyan revolution, which had used this zipper rule to elect an unprecedented number of women legislators:

Our society, shaped by a revolutionary mindset, became more polarized and has driven away from the ideals and the principles—freedom, dignity, social justice—that we first held. Intolerance, exclusion and revenge became the icons of the [aftermath] of the revolution. I am here today not at all to inspire you with our success story of the zipper list and the elections. I’m rather here today to confess that we as a nation took the wrong choice, made the wrong decision. We did not prioritize right. For elections did not bring peace and stability, and security in Libya. Did the “zipper list” and the alternation between female and male candidates bring peace and national reconciliation? No, it didn’t. What is it, then? Why does our society continue to be polarized and dominated with selfish politics of dominance and exclusion, by both men and women? Maybe what was missing was not the women only, but the feminine values of compassion, mercy and inclusion. Our society needs national dialogue and consensus-building more than it needed the elections, which only reinforced

polarization and division. Our society needs the qualitative representation of the feminine more than it needs the numerical, quantitative representation of the feminine. (Langhi 2012, 5–20, emphasis added)

More Choice is Not Enough

Even in a two-party system, most voters know very little about most of the candidates. In a PR system with more parties and candidates, voters may know proportionately less about the candidates. A group of researchers using more than sixty thousand elections in Brazil over 20 years as their database has also found that an increased number of candidates on a ballot, facilitated by PR, may actually suppress voter participation (Cunow et al. 2023). They suggest the burden of assessing a large number of candidates is simply too daunting for many voters. “Although conventional wisdom suggests that providing voters greater choice may enhance democracy, our analysis shows that there is an overlooked cost to providing voters more choice,” Andrew Janusz said (quoted in Science X 2023).

It is not all that surprising that in a parliamentary system of government using a PR election system, it is sometimes difficult to negotiate a governing coalition. An extreme example of this occurred in Belgium in 2010–2011, when the political parties failed to agree on a governing coalition for 589 days. Though basic government functions were maintained by the bureaucracy, there was nobody steering the ship of state.

This electoral failure contributed to a growing interest in sortition in Belgium. A 2013 book by a prize-winning Belgian author, David Van Reybrouck, entitled Against Elections, makes the case for sortition (Van Reybrouck 2018). Van Reybrouck was invited to speak all over Europe and the world, including on De Zevende Dag (The Seventh Day), the flagship talk-show on Flemish public television, for a wide-ranging discussion about his book with former Belgian Prime Minister Jean-Luc Dehaene. The former prime minister endorsed the need for democratic innovation and urged the government to start exploring sortition. Van Reybrouck’s efforts were an important factor in the subsequent mushrooming of hundreds of lottery-drawn citizens’ assemblies across Europe.

While the gender and class make-up of legislatures in countries using PR may be more diverse than in the United States, legislators still constitute a separate “political class”, and do not accurately mirror the diversity of the population. Perhaps as a result, even in countries using PR, politicians are often held in low esteem. As just one example: a study in Spain found that politicians were the least trusted professional group in society, with 70 percent of those interviewed saying they wouldn’t want their children getting involved in politics (Spanish Foundation for Science and Technology 2012).

In an interview about his book, Democracy Erodes from the Top: Leaders, Citizens, and the Challenge of Populism in Europe, political scientist Larry Bartels (2024) argued that elite control is inevitable whether in a two-party system or a PR system.

One thing I’ve worked on in my career is to estimate the impact of ordinary people’s preferences on policy outcomes, both in the US and other places. The implication of that work is that ordinary people’s views have very little impact on what their governments do. That was shown first in the United States, and people imagine that it has to do with particular features of the US, like our campaign finance system or the weakness of labor unions or our separation of powers. But then people began to do parallel studies in European countries that were supposed to be much more advanced democratic systems, like Sweden and Germany. They found much the same thing in those places, in spite of the fact that they have very different institutions and very different political cultures. That suggests to me that there are much more fundamental factors that limit the impact of ordinary citizens on the behavior of leaders in any political system. (Bartels 2023)

He concludes that the goal should be to “socialize” the elites to behave better, since they are in charge regardless of the electoral system. He does not consider the option of replacing the elected politicians with sortition, where everyday people directly make the decisions without a role for politicians.

I don’t want to leave the impression that PR is worthless. It almost certainly would be an improvement over the current US winner-take-all election system. My point is that PR is not a worthy endpoint goal for democratic reform. Democracy reformers in Australia, which has used a form of proportional representation with ranked choice ballots for generations, are in fact among the leading advocates of sortition, in response to what they view as the inadequacy of proportional representation.⁷ Though the institutionalizing of sortition bodies is already underway in many countries, as we will discuss below, full development and refinement of sortition as a tool for democracy will take more time. Therefore, interim reforms that may decrease the power of plutocrats, such as campaign finance reform and PR, may be worth pursuing as well.