In November, voters in South Dakota will decide whether to pass Amendment V. If passed, the measure will institute a nonpartisan primary much like the one in Washington state. Under the proposed rules, all candidates would appear on a single ballot without party labels, and the top two recipients of primary votes would advance to the general election.

The proponents of the reform cite numerous potential benefits, including increased voter turnout, the election of more moderate officials, depolarization, and the restoration of functional governance. One supporter called Amendment V “the answer we have all been hoping for to address many of the problems we see with politics these days.”

Yet as well-intentioned at the amendment’s backers may be, there are empirical and principled reasons to oppose the measure.

First, there is no persuasive evidence to show that moving to a nonpartisan top-two primary will affect the behavior of elected officials and encourage moderation. (We can bracket the question as to whether we should engineer our electoral models to promote a desired outcome like greater moderation in the first place; I insist we should not.)

One comprehensive study of legislators’ voting records in all 50 states over a 20-year period, for instance, found that the type of primary used had no significant effect on voting behavior. And a recent analysis of California’s experience with top-two concluded that “the move to the system has had, at best, a minimal impact on increasing the number of moderates getting elected.”

The idea that moving from a closed partisan primary to an open, "blanket" one should result in increased moderation is rooted in a pervasive myth: that those locked out of parties’ closed primaries – independents – are by-and-large moderates. Empirical evidence belies this notion: in fact, there is evidence indicating that the typical open primary voter is more extreme – not more moderate – than the typical voter in a closed, partisan primary.

Support for this observation can be found in the results of the 2016 presidential primary contests. It was not more moderate candidates who benefitted from open primaries, but instead candidates like the more “extreme” Senator Bernie Sanders (markedly to the left of Secretary Clinton) and Donald Trump (who, difficult to categorize as he is, is no moderate).

The 2016 presidential election evinces another reason to oppose South Dakota’s proposed amendment: its reduction of voter choice.

Given that Clinton and Trump are two of the most disliked presidential candidates in 30 years, it is no surprise that many voters are demanding more voices and seeking out other choices: 76 percent of voters want to see Gary Johnson and Jill Stein participate in the debates, for instance, and in four-way mock election polling, 23 percent say that they are undecided, prefer Johnson or Stein, or refuse to choose any candidate.

Yet under top-two there are, by definition and by design, no such alternatives. If the top two recipients of primary votes are widely and wildly unpopular (and Clinton and Trump did receive the most primary votes at more than 15 million a piece), there is no outlet for principled dissent – perhaps outside voter abstention. There is no possibility to broaden the debate by adding a third podium to the stage, no option to allow an insurgent third party or independent candidate to run, and – if the system ends up resembling California’s – no chance to permit a candidate to wage a last-ditch write-in campaign.

While proponents of top-two insist that this model fosters genuine competition and forces candidates to “reach beyond their base,” this competition can take a rather peculiar form. After all, this November, 7 of California’s 53 contests for U.S. House will feature two members of the same party, as well as 15 of 80 state Assembly contests and 5 of 20 state Senate contests. And most frustratingly for many California voters, the U.S. Senate race features two liberal Democrats, Kamala Harris and Loretta Sanchez, inspiring widespread apathy (and, as it appears, failing to foster competition).

This reduction of voter choice is as detrimental to voters as it is to candidates – especially third party and independent candidates. Indeed, in California, these candidates have fared terribly since the implementation of top-two: having quickly acclimated to the new rules, the Democratic and Republican parties have managed to maintain their supremacy over the state's politics.

As some have argued – including election law expert Richard Winger – top-two may not just be bad in practice for third party and independent candidates, it may even be unconstitutional.

One argument against the constitutionality of top-two combines two Supreme Court decisions. In 1971, the Court established a ceiling on the level of popular support that candidates must demonstrate in order to advance to the general election: 5 percent of voters. And in 1986, the Court held that there is no constitutional difference between having to demonstrate support through a petition and through performance in a primary contest.

Together, these decisions indicate that a state cannot restrict access to the general election ballot by requiring a show of support equal to, say, 10 percent of votes cast in a primary any more than it can restrict access to the general election ballot by requiring a candidate to gather signatures among, say, 10 percent of voters. How does this relate to top-two? Because candidates under top-two typically need to receive at least 30 percent of votes in the primary to advance to the general election.

In other words, these rulings suggest that top-two is unconstitutional because the states that adopt it are imposing an overly burdensome requirement on candidates by effectively stipulating that they demonstrate support well in excess of the 5 percent figure.

And top-two is likely unconstitutional for another reason: like the open primary, it infringes on parties’ freedom of association by denying them the ability to do what are perhaps parties’ most important functions – to define their membership and to select their own nominees for office. These rights have been upheld by the Supreme Court on several occasions, including in 1975, 1981, 1986, and 2000. They cannot be abrogated by the general public simply because some independents are frustrated that they cannot participate in the internal election of a party that they refuse to join – even temporarily.

Proponents of open primaries in general and top-two in particular insist that since taxpayers foot the bill for primary contests, then all voters should be permitted to participate. While I, for one, am sympathetic to those who complain about all taxpayers (including independents) having to subsidize private elections, the solution is not to demand or mandate universal access to these elections. A more sensible solution is to recognize that these are indeed private elections and thus to have the parties pay for them themselves.

In short, Amendment V’s proposal to bring nonpartisan top-two primaries to South Dakota is riddled with problems. Based on a misreading of the ideological makeup of independents, top-two cannot be expected to significantly promote moderation (nor, I believe, should a state jigger the rules to encourage such a preferred outcome).

In addition, the electoral model would restrict voter choice in the general election at a time when voters are increasingly frustrated with the two-party system. After all, only 37 percent of voters say the two parties are doing an adequate job of representing Americans, and 57 percent say a strong third party is needed. And among those seeking more choices, some are even looking beyond the Libertarian and Green Party options at independent candidates and lesser known parties.

Moreover, the proposal – like its predecessors – is of dubious constitutional validity. Requiring candidates to demonstrate levels of support well in excess of 5 percent of voters in order to qualify for the general election ballot contravenes long-standing precedent, as does the violation of parties’ right to select their own nominees.

None of this is to say that we should preserve the status quo – far from it. Instead, we should be promoting reforms that respect the rights of parties, enable third party and independent candidates to access the general election ballot, and increase competition and voter choice.

There are many such reforms that can further these goals, including ranked choice voting, the adoption of multi-member districts and fair representation voting, and proportional representation – not to mention the reduction and elimination of restrictive ballot access laws and the abolition of gerrymandering.

The proposal and adoption of any of these reforms in South Dakota (or elsewhere) would represent significant progress in the struggle for more open, inclusive, and competitive elections. Amendment V does not advance these goals. Voters of all partisan identifications and ideological persuasions should reject it.