Republicans say Type I errors are more costly than Type II errors in elections, but democracies are defined by participation inclusion more than sovereignty
Political battles over voting rights are being waged on many fronts in the United States. Democrats in the House have passed H.R. 1, “For the People Act of 2021,” which would provide sweeping voting rights protections. At the same time, Republicans are seeking reforms in many states that would restrict access to voting. Republicans and Democrats claim their efforts are focused on improving Democracy and guaranteeing the integrity of elections; however, each party is operating under a completely different theory about the role elections play in democratic governance—but only one party is motivated by values consistent with democracy.
Each party wants us to believe that they are fighting for the very soul of America and they seek reforms to reinforce US democracy. Each party claims that the other party’s reforms are a threat to democracy and pose too great a risk to the country. The proposed reforms generally contradict one another (e.g., expand or restrict early voting). So which party seeks to strengthen democracy and which party’s reforms pose greater risks?
Ultimately, Republicans maintain their primary concern is about ensuring election integrity. That is, they want to ensure that only those who have a legal right to vote are allowed to participate in elections. On the other hand, Democrats maintain their priority is to ensure that everyone who has a legal right to vote is readily able to exercise it.
The severity of an error depends on its context
To use a science analogy, when it comes to voting and elections, Republicans are worried about Type I error and Democrats are worried about Type II error. These error types refer to challenges associated with causal inference. A Type I error refers to a false positive, such as a doctor telling you, you have a disease that it turns out you do not actually have. The Type II error refers to a false negative, such as a doctor informing you that your tests indicate you are not sick, when in fact you are. Both of these cases describe a faulty test, and both are problematic; however, they are not equally problematic. Further, it is not the case that a false positive is always worse than a false negative, or vice versa. Rather, the level of severity of the error is related to the context of the test.
In some cases, a Type I error, or false positive is a worse mistake than a false negative. Suppose you are a scientist working to develop a vaccine for a horrible disease and you run a test to determine whether or not your vaccine is effective. If the test for the effectiveness of the vaccine produces a false positive, or Type I error, the result is that an ineffective vaccine is given to people, who now believe they are immune to a disease to which they are not.
If the test produces a false negative, then an effective vaccine would be deemed ineffective, and tossed out. Which is worse? Giving an ineffective vaccine to millions who behave as if they are immune? Or throwing out an effective vaccine because you believe it’s ineffective? In this case, the false positive, or Type I error is much worse because it may result in people becoming sick because they thought they were immune but weren’t.
On the other hand, you could also imagine a case where a false negative creates a much worse outcome than a false positive. Imagine a doctor tests you for heart disease. If the test produces a false positive, it means you have a healthy heart but are told you have a diseased one. You may change your diet and exercise more in an effort to combat a disease you do not have. Of course, such lifestyle changes would improve your health regardless if you started with a healthy heart or a sick one.
If the test produces a false negative, or Type II error, the test indicates your heart is healthy when in fact it is diseased, meaning a diseased heart might go untreated. In this case, a Type II error is clearly much worse than a Type I error, because leaving a diseased heart untreated is more likely to produce a negative outcome than treating a healthy heart as if it were diseased.
Whether a false positive or a false negative produces a worse outcome is entirely dependent on the context of the situation. The same is true with voting rights.
Sovereignty versus inclusion
Republicans remain concerned about false positives when it comes to elections and voting. When a non-citizen illegally casts a vote, it is like a Type I error where the vaccine was delivered to market when it should not have been. A false positive is an event that occurs in error. The voter, like the vaccine, was delivered in error.
Democrats, on the other hand, remain focused on Type II errors, or false negatives, when it comes to voting. When someone who has the legal right to vote is prevented from doing so, it is similar to a diseased patient being told they are healthy. A false negative is a needed event that does not occur. Erroneously preventing a legal vote from being cast is a Type II error because it means someone who should get something, doesn’t get it.
When it comes to elections and voting, which is worse, a Type I error or a Type II error? Republicans argue that it is worse to allow non-voters to vote than it is to prevent legal voters from participating in elections. But Democrats argue that it is much worse to prevent legal voters from voting than it is to mistakenly allow non-citizens to vote.
Of course, we could observe that the false positive event of illegal votes being cast occurs much, much more infrequently than the false negative event of legal voters facing barriers to voting; however, we could also look at the costs associated with each error.
In elections where many false positives occur, the integrity of the election is undermined in the sense that elections are intended to be a mechanism to measure the preferences of people in society so that leaders can be selected and enact the voters’ preferences. If non-eligible voters participate, then it interrupts the sovereignty of the state. Americans would not expect to follow laws set by Canada or Mexico? So why should elections for selecting American lawmakers include non-Americans? Election integrity is a matter of defining who is a legal member of a state or nation. Strict rules about who can participate are about enforcing the boundaries of government sovereignty. The costs for violating this boundary include a weaker sense of citizenship and sovereignty. Questions of sovereignty and defining who controls the levers of government have been core themes of Republican politics for several decades. It’s easy to see why Republicans remain concerned about Type I errors in elections.
In elections where many false negatives occur, citizens who are governed by a legitimate authority do not participate in selecting that authority. These citizens are expected to follow the laws of the state, but they have not participated (by choice or force) in selecting the representatives who make the laws. Since the 1960s, the Democratic Party has aimed to promote democratic institutions that maximize inclusivity, although this aim has often been imperfect. The costs for violating the principle that democracies are participatory systems of self-governance are that the non-participants wind up being ruled by systems they did not select. Since elections are the primary institution that qualify a system to be called democratic, any rule that limits participation in elections is inherently anti-democratic. It’s easy to see why Democrats remain concerned about Type II errors in elections.
At the end of the day, you can have a democratic system with weakly defined sovereignty, but you can’t have a strictly sovereign system that systematically excludes participation from some groups and call it democracy.
If the goal is to build a democratic society where the systems have integrity and the barriers to participating are low, it is clear that the costs of Type II errors are much higher than the costs of Type I errors.
But it’s likely all parties involved with these debates already know this, even if they use slightly different language to describe it. If you’re claiming that strict sovereignty is more important than broad participation, or Type I errors are worse for America than Type II errors, you’re not arguing for democracy. Rather, the claim that sovereignty overrules inclusivity is a claim about who is American. Dog whistles or not, arguments that sovereignty is more important than inclusivity is an explicitly racialized claim.
Strictly controlling participation, and undervaluing the possibility of exclusion, is antithetical to democracy, which requires inclusivity. Those who prioritize the boundaries of the state and defining who is allowed to participate are fighting to maintain a system in which their group has been historically powerful. They are understandably reluctant to give up that power, but let’s be clear that what they’re asking for is a system that is decidedly not democratic.