Voting to Send a Message
Breaking down if protest voting is a good idea.
How do people decide on what to vote for? Likely it’s the outcome that they have deemed “better”. Better for who? Maybe it’s better for them, better for their community, better for their country, or better for the world. All very subjective, but what matters is that they are voting for someone or something because they believe it will lead to better outcomes than the alternative.
There is another reason to vote for a specific outcome, and that is signaling. You may not think the Proposition or candidate will directly lead to better outcomes, but instead you want to send a message, hoping for improved outcomes in the long-term. Let's look at a few examples of how and why someone would signal with their vote:
Example 1: Say you are passionate about curtailing crime. There is a recall on the ballot for an alleged “soft on crime” DA. You investigate the details and see that the DA has only been in office a few months, and while perhaps they could have been tougher, "soft on crime" is probably an exaggeration. Despite this, someone voting to send a message would vote to recall this person anyway. Why? Because what’s more valuable than either outcome is sending a message that people are fed up with crime. This message could then influence other politicians to get tougher on crime to avoid the fate of being recalled themselves.
Example 2: Say you are passionate about improving the lives of the less fortunate and reducing income inequality. On the ballot there is a Proposition for raising the minimum wage. You remember from economics class that raising the minimum wage can have serious negative outcomes, such as higher unemployment for the people that the minimum wage is trying to help. If you’re voting to send a message, you’ll vote to increase the minimum wage anyway, even if you believe raising the minimum wage could harm more than it could help. Why? Because what’s more valuable is sending a message that people are fed up with inequality and the inability of people to make a living. This message could then influence politicians to think more about enacting policies that will do that.
Example 3: Let's say you are disillusioned with the current state of politics, and you hate both parties. You know that you don't hate each party exactly equally, and the most rational route is to vote for "the lesser of two evils". However, you decide to send a message by voting for neither party. You instead vote for the Green or Libertarian parties, or you don't vote at all. Your goal is to send a message to those in power that they need to start fielding better candidates. (This is an example of a protest vote, which is a certain type of signaling vote).
I believe the intention of someone signal voting is the same as voting for something because it’s “better”. With a signaling vote you are doing the same thing, just a couple steps removed. You may know that the decision is not ideal (like not voting instead of choosing the lesser of two evils), but if the message that you are trying to convey is received (candidates get better), it will be “better” in the long term. At least that’s the idea.
“Sending a message” implies that there is someone receiving it. Who are we sending our message to? The first order receiver is probably just everyone, specifically politicians, the media, and the general voter. Although the ultimate receiver is likely to be politicians, broadly defined.
Is this a good method? In my opinion, sometimes, but not usually. Determining if you should vote to signal, I would consider two factors: the importance of the decision and the reception and interpretation of the message.
The importance of the vote
As we made clear in the examples, usually when you are voting to signal, you may not be voting for a side that you think is better. You are essentially conceding that you are willing to tolerate things to get temporarily worse, with the assumption that they will eventually get better (assuming the message is received).
Looking at Example 1 (recalled DA), you know that recalling DAs usually leads to a transitional messy period, and this DA wasn’t even that soft on crime. You concede that things will probably get slightly worse if they are recalled. However, if your message is received and DAs across the state get tougher on crime because of it, then it will be worth it. If you want to get tougher on crime, I think it makes sense to recall the DA because the “importance of the vote” is relatively low. I think it's likely that the value of the end message, relative to the downsides of a recall, is worth it for this person.
What would be an example of an important vote? Let’s come up with an extreme example to make the point. Say there was a Proposition to set the minimum wage to $1,000 an hour. Or a presidential race between Abraham Lincoln and Pol Pot. “Importance” is subjective, and while these examples are intentionally absurd, they are indisputable in regards to their importance. In either of these examples, it would be a really bad idea to decide to signal. You likely should not abstain from voting for Lincoln just to “send a message”. If you are going to do a signaling vote, make sure the vote in question is relatively low stakes.
Reception and interpretation of the message
The next factor in deciding to cast a signaling vote is reception of the message. In other words - will your message be received and interpreted correctly? Usually, any result will lead to some sort of message, but will it be yours? The resulting message of an election can easily be hijacked.
One example of an attempted hijacking was the successful recall of former San Francisco DA Chesa Boudin. A lot like our example, he was soft on crime, and was recalled, rightly or wrongly, for that reason. Those in Chesa’s camp tried to change the message from “San Franciscans are fed up with crime” to something along the lines of “San Franciscans are all closet conservatives conned by right wing billionaires into thwarting national criminal justice reform.” Thanks to local journalists like Heather Knight (no relation), this message never took hold.
The problem, of course, is that it’s hard to know when your message is going to get hijacked. I think the best approach is to assume the more controversial and polarizing the decision, the more likely the message will be hijacked.
But even if there isn’t an active campaign to hijack the message, there is still a chance your message is misinterpreted by those you are wishing to signal. Let’s look at approving a high minimum wage, for example. We want it to be interpreted as “the people are fed up with inequality”, but it could also be read as: 1) people are voting for their self-interest without concern for second-order effects, 2) voters “just don’t understand economics”, or even 3) no message at all. What are the odds our preferred message is the one that people take away?
Or take our example of not voting to signal to political parties to improve their candidates. Our intended message is for the political parties to put out better candidates, but low voter turnout could be interpreted as: 1) we need to expand voting access, 2) the younger generation is awful and self-absorbed, or 3) we didn’t have enough money to spend on advertising. Again, how sure are you that the right message will get across?
Out of all the examples we talked about, only the DA recall has a clear interpretation. As long as the recall narrative was about “being soft on crime”, it’s difficult to imagine other politicians and the public interpreting it in some other way.
Conclusion
To put it all together, I think voting on something to send a message, and not because you believe it’s better, is sometimes an ok strategy, but usually not. If you are going to do this, I would consider the following:
Is the direct outcome of the vote relatively unimportant to you and non-consequential?
Are you confident that your message will be received, interpreted, and actioned on in the way that you intended?
If the answer to both is not an astounding YES, I would avoid signaling and just vote for whichever outcome you think is better.
Some good self inflections questions asked and answered. I think the only other piece that could come into signal voting is preventing unintended consequences that there is no way to verify if they would come true until you let the scenario play out. Who knows if making marriage a constitutional right for any person of any age could lead to child marriage to an adult. So maybe you’re signaling clean up the proposition and then get my stamp. For instance, just saying gay marriage is a constitutional right.