Philippa Kaur

In August 2022, a statewide referendum in Kansas saw citizens overwhelmingly reject a plan to insert anti-abortion language into the state’s constitution. It comes as a slew of similar votes on abortion rights are planned in the coming months – putting the issue directly to the people after the Supreme Court struck down the landmark Roe v. Wade ruling.

But are referendums and citizens initiatives good for democracy? It may seem like an odd question to pose on International Day for Democracy, especially at a time when many feel democracy is imperiled both in the U.S. and around the world.

As someone who researches democracy, I know the answer isn’t simple. It depends on the kind of initiative and the reason that it comes to be held.

First, some simple distinctions. Referendums and citizens initiatives are mechanisms of direct democracy – instances in which members of the public vote on issues that are commonly decided, in representative systems, by legislatures or governments. While with referendums it is typically the government that places questions on the ballot, with citizens initiatives – more common at the state level in the U.S. – the vote originates outside of government, usually through petition drives.

The Chicago Center on Democracy, which I lead at the University of Chicago, recently launched a website that tracks many of these direct democracy efforts over the past half-century.

Appealing to the masses or settling scores

That a majority of democracies retain some form of direct democracy is a testament to the legitimacy with which citizens’ voices are heard, even when, in fact, most decisions are made by our elected leaders. Often, national governments call referendums to bring important questions directly to its citizens.

But why would governments ever decide to turn a decision over to the people?

In some cases, they have no choice. Many countries, among them Australia, require that constitutional amendments be approved in popular referendums.

In other instances, such votes are optional. United Kingdom Prime Minister David Cameron, for example, was under no obligation to undertake a 2016 referendum on continued EU membership. Colombian President Juan Manuel Santos had plenty of legislative support that same year to ratify peace accords with a rebel group through an act of congress. But he turned the decision over to the people, instead.

Pushing a Brexit referendum backfired on then-Prime Minister David Cameron.
Brian Lawless/PA Images via Getty Images

One reason leaders voluntarily put important issues before voters is to solve disputes within their own political parties. The Brexit vote is a case in point. The U.K. Conservative Party was deeply divided over British membership in the EU, and – as Cameron later acknowledges in his memoirs – his position as head of the party, and thus as prime minister, was increasingly threatened.

In these instances, the government is in effect using the people as a referee to decide an internal dispute. It is a high-risk move, though. For Cameron, going to the country meant the end of his premiership. And six years on, the U.K. is still dealing with the fallout of that vote.

Sometimes leaders seek public support on issues about which they expect powerful opposition upon implementation. Colombia’s Santos expected resistance to the peace deal from opponents, including wealthy landed interests. He used the people as a kind of force field to protect the policy. But again, the strategy backfired. The Colombian accords were defeated, and have since faced powerful resistance when subsequent attempts were made to implement them through legislative approval.

But do these two high-profile instances illustrate fatal flaws in referendums, and direct democracy in general? Perhaps not.

Though plenty of disinformation circulated before both votes, the results probably fairly accurately reflected the people’s preferences. Moreover, they illustrate the perils to political leaders of placing issues of crucial importance before voters – they can’t be sure they will like the results.

And when their referendums fail, they may set back causes that these politicians care about. For example, Brazil held a referendum on gun control in 2005. It failed, and later pro-guns rights president Jair Bolsonaro used its failure to try to loosen restrictions on firearms, claiming that the failure of the referendum allowed him to do so.

Tool of demagogues

Sometimes the prime minister or president does prevail. A kind of referendum was used in Australia in 2017 to pressure the legislature into legalizing same-sex marriage. Conservative politicians were willing to hold a vote, with the same kind of “referee logic” as in Brexit – they were opposed to same-sex marriage, but preferred to go along with the public’s will, rather than continue to fight over this internally divisive issue.

In the end, the pro-marriage equality prime minister opted for a “postal survey” rather than a formal referendum. And the gamble worked for Australia’s leader – a very large majority expressed support of same-sex marriage and the prime minister got his way.

For every Colombia-style debacle, in which a leader holds an optional referendum but fails, one can point to governments putting matters to a popular vote to produce a force field, and winning. The approval of the public can make policy immune to –- or at least undermine – later opposition. Such was the case of same-sex marriage in Ireland, passed by referendum in 2015. The following year, Ireland settled the issue of abortion access, overturning a ban by a two-thirds majority.

Referendums are not only used by democratic leaders but also by autocrats and demagogues. Russian president Vladimir Putin put a series of constitutional reforms before voters in 2020, including one that overturned Putin’s prior term limit in office.

Accusations of fraud and intimidation followed the vote. The process could hardly have been more at odds with direct democracy and the autonomous expression of the people’s will.

Getting policy to line up with people’s will

There are no national referendums in the U.S. But American voters have a great deal of experience with initiatives at the state level – and with state-wide referendums, as well. These votes have the potential to force governments to abide by the people’s will in cases where legislators may be resisting popular policies.

Yet problems can arise with these exercises in direct democracy. Even though they are presumably citizens’ initiatives, the influence of political parties, special interests, lobbyists and big money can turn them into something quite different, as was the experience of California in the 1990s – which in turn undermined the public’s satisfaction in the initiative process.

But recently we have seen a spate of state initiatives that seem more promising – where majorities of citizens are demanding that their state legislatures bring policy more in line with public opinion. Florida voters approved ex-felon voting; Arizona voters approved bigger budgets for public schools; Missouri voters forced a reluctant legislature to expand Medicare in their state. All of these initiatives were backed with popular public support.

Most recently, Kansans said “no,” in referendum, to inserting pro-life language into their state´s constitution.

‘Let the people decide!’

The potential for mechanisms of direct democracy to improve citizen representation depends on the context in which they are held, including the manner in which they are placed on the ballot and the motives of those who placed them there.

At one extreme are autocrats like Vladimir Putin who held votes that augment his power and the length of his term. At the other are citizens frustrated by legislators whose actions stray far from public opinion. In between are measures sponsored by governments that may want to insulate policies they care about with the help of the people’s backing, and parties that throw their hands up, in the context of internal divisions, and say, “let the people decide.” More

A county court in Santa Fe, New Mexico, on Sept. 6, 2022, became the first in more than 150 years to disqualify a person from public office because they participated in an insurrection.

District Court Judge Francis Mathew found that Couy Griffin, a former county commissioner and founder of the group Cowboys for Trump, had participated in the violent U.S. Capitol attack on Jan. 6, 2021. Mathew invoked a nearly forgotten part of the 14th Amendment, called Section 3, which can disqualify certain people from state or federal office if they have “engaged in insurrection or rebellion” or given “aid or comfort” to the United States’ enemies.

The clause was first adopted after the Civil War to keep former Confederates from participating in politics. The amendment says that disqualified people are barred for life from either running for or being appointed to office. But Congress can vote by a two-thirds majority to waive this ban.

The clause fell into general disuse after 1872, when Congress gave amnesty to most former Confederates in a move toward reconciliation.

Some observers have argued that Section 3 disqualification should be dusted off to address the Jan. 6 mob and to stop other people who have threatened and committed violence – or tried to disrupt federal elections – from serving in government.

Mathew’s decision has also renewed talk among Democrats and good-governance groups about finding a way to use Section 3 against former President Donald Trump in order to disqualify him from ever holding office again.

We are scholars of comparative constitutional law who have worked on democratic backsliding around the world. In a forthcoming article, we point out that disqualification is potentially a useful tool to protect democracy, but it can also be dangerous – it rubs up against the basic idea of democracy as a system in which anyone can run, and voters can decide.

The disqualification of Couy Griffin is one of the latest efforts – but the only successful one – to remove people from office or prevent them from running because of their support for the Jan. 6 Capitol attacks.
Brent Stirton/Getty Images

Recent attempts at disqualification

The disqualification of Griffin is one of several efforts voters and advocacy groups have lobbied for after Jan. 6. Most of these efforts have failed to remove someone from office or prevent them from running. But the examples are still useful in understanding how disqualification might be an alternative to more punitive criminal law options.

A suit filed by a group of voters to disqualify Rep. Marjorie Taylor Greene, for example, foundered in July when a Georgia court affirmed a lower court ruling that she had not “engaged in insurrection.”

Arizona and Wisconsin state judges have also rejected efforts to use Section 3 as a sword against those who supported the Jan. 6 insurrection. But none of these targets actually participated in the mob at the Capitol. Mere support of the rioters, or questioning the election outcome, is protected political speech under the First Amendment.

Griffin, though, engaged in a physical invasion of the Capitol.

Mathew’s careful opinion contains extensive factual findings and legal analysis. So it tees up nicely the question of whether and how disqualification from democratic office is legitimate, justified or effective in defense of democracy.

This is especially important in the U.S., where there is a lack of recent historical experience with disqualification of people working in politics.

When disqualification makes sense

Other countries make much more extensive use of political disqualification than the U.S. does, as we show in our forthcoming study.

Israel’s courts, for example, have repeatedly disqualified candidates for lack of “good character.” In Pakistan, the supreme court disqualified sitting Prime Minister Nawaz Sharif in 2017 after he was named in the Panama Papers because of corruption.

There are various costs and benefits to disqualifying someone from office, and there are also open questions of how to correctly interpret Section 3. We focus on the first question of costs and benefits here.

Democracies require robust protections for free speech and association. But these freedoms can be abused by those seeking to undermine democracy itself.

For example, Mathew documents Griffin’s persistent efforts to cast doubt on the legitimate outcome of the 2020 election and to instigate violence to derail President Joe Biden’s inauguration. Most of Griffin’s actions, however, fell far short of the threshold necessary to justify criminal penalties for incitement – the First Amendment requires that the violence be imminent.

Griffin, nonetheless, participated in a concerted threat to American democracy. Disqualification is a way to address such threats without the heavy hand of the criminal law.

Section 3, more generally, is another way to address high-level misconduct in politics. As we have explored in another study published in 2021, Congress has rarely impeached a U.S. president – and an impeached president has never actually been removed from office. Given partisan dynamics, it is unclear if impeachment could actually remove and disqualify a sitting president.

This might leave Section 3 as the best alternative.

The risks of disqualification

Mathew’s opinion suggests that Section 3’s “aid and comfort” language can go uncomfortably far. It could potentially chill legitimate political speech – including criticism of the government, or support for a foreign power – that doesn’t threaten democracy.

For example, plaintiffs justified the need to disqualify Griffin by saying that he committed “actions that normalized and incited violence” by “dehumanizing the opposition as ‘wicked’ and ‘vile.’”

Dehumanizing speech about political opponents is indeed often unhealthy for democratic practice, but it has become routine in politics.

Another challenge is that the text for Section 3 is not entirely clear about how disqualification actually works. Does it apply automatically to anyone who engages in insurrection? Or does it require some sort of either judicial or legislative process?

There is no settled answer.

In an 1869 decision, Supreme Court Chief Justice Salmon Chase suggested that Section 3 did not apply automatically – rather, disqualification occurred when Congress, or perhaps a state legislature, authorized it.

Mathew rejected the position that only Congress could make the determination and instead held that Griffin could be disqualified by order of a state court.

The potential breadth and ambiguity of Section 3 creates a risk that the measure could be repurposed, against its original aims, in a way that hurts democracy.

While a few presidents have been impeached, including Donald Trump twice, they were not removed from office.
Drew Angerer/Getty Images

The bigger picture

Disqualification, then, is a superficial remedy to a profound problem. It might be effective against a low-level official like Griffin, but the bigger the target, the less power everyday voters have.

Imagine that some court was persuaded to disqualify Trump from a state’s 2024 presidential ballot. Such a move could be considered to disenfranchise his supporters. This could play into Trump’s beliefs that the “game is rigged.”

Whatever the correct legal answer, there is a strong case for eliminating the uncertainty around how Section 3 works. We’ve argued for a carefully crafted federal statute that clearly explains when it applies and how it works.

If disqualification is to become an effective sword to defend democratic politics, it must not become a two-edged one that later weakens the democratic process in the U.S. More