We Need More Costly Signaling

The general contours of the saga of the Republican Party and conservative politics in the 1960s are well known to most political observers. In brief, the modern conservative movement was born out of Barry Goldwater’s loss to Lyndon Johnson in the 1964 presidential election, which led to the building of the think tanks, activist groups and fundraising operations that followed over the next 16 years, culminating in the election of Ronald Reagan to the presidency. 

After Goldwater’s defeat, the institutional apparatus of the GOP was taken back by the so-called Eastern Establishment, comprised of more centrist and even liberal Republicans ranging from Richard Nixon to Nelson Rockefeller. Meanwhile, the Goldwaterites were unceremoniously cast out.

A recent conversation with one of the leading figures of the post-Goldwater conservative wing of the party—someone who is not well known outside political circles but whose name appears in every history of the modern conservative movement for his role in building the army of activists who helped bring about the Reagan Revolution—gave me insight on the machinations of the GOP in 1964 and 1965, and why those decisions were crucial in setting the stage for what was to come.

In brief, in the wake of the GOP’s 1964 defeat, actively supporting Goldwater was a costly signal to send if you were a partisan Republican and an up-and-coming party careerist. And it was the costliness of the signal—the fact that Goldwater’s loss had real and immediate negative career repercussions—that allowed them to regroup outside of the party and form their own parallel institutions that eventually brought them to prominence again. Their expulsion created the crucible in which true believers could identify one another after the election and bond to create the modern conservative movement.

That is, the Goldwaterites leveraged a short-term cost into a long-term payoff. 

Costly signals are necessary to differentiate ingroups from outgroups, a designation that’s needed if we’re going to have, well, groups. But such signals are in perilously short supply at the moment. 

For example, it used to be that if one wanted to be a candidate of a political party, it was necessary to signal affiliation with that political party by being a member. This was not a particularly costly signal: All one had to do was change one’s voter registration or perhaps pay a nominal fee to join a local party chapter. 

By doing so, one foreclosed signaling either nonpartisanship or support for another party. The primary cost of being, say, a Democrat was the opportunity cost of being a Republican, a Libertarian, a Green, a non-inscrit. It was a clear way of differentiating who was, say, a Republican from those who were Not-Republicans, which is necessary if being a Republican is to mean anything. You could not simultaneously be a Democrat and a Libertarian, or an independent and a Republican. 

This is no longer the case. The fact that in 2016 the Republican Party nomination was captured by someone who had little allegiance to the party—and the Democratic Party nearly suffered the same fate. Parties are not and were never intended to be democratic institutions; they are firms overseen by leaders and managers with the aim of winning elections. But today the parties are in thrall to political hobbyists and small-dollar donors with little if any party loyalty—and who in some cases don’t appear to actually want to win elections.  

Cheap signals abound, and costly ones carry little weight; party membership itself no longer signals much of anything. This is a fundamental reason our political institutions appear so dysfunctional.

Take candidate rhetoric. Until about 10 years ago, taking a stand on same-sex marriage was a politically costly signal to send; Democrats in particular had to walk very fine lines through the oughts. Now, ostensibly serious candidates can tweet word salads like “Black trans and cis women, gender-nonconforming, and nonbinary people are the backbone of our democracy” or advocate for taxpayer-funded gender surgeries for illegal immigrants. Sending such a signal—whatever it means, other than obeisance to The Current Thing—is effectively costless. 

Or to take an example from outside of politics, think about tattoos. It’s hardly an original observation that visible tattoos used to be a costly signal. Regardless of the specific signal a tattoo sent to the semiotic cognoscenti, the mere existence of a visible tattoo signaled that someone was at most on the periphery of mainstream institutions. Ink was costly because it permanently precluded its bearer from pretty much any position of social responsibility or (legal) employment above the semiskilled level. Visible tattoos were a clear, convincing and none-too-subtle signal. 

Clubs, teams and fraternal organizations also rely on signals of membership, even if the costs are more like political party membership than tattoos. Many of these signals may be inexpensive, but they’re still not cheap. Oaths, uniforms, secret symbols and the like serve to designate who is a member of the group and who is not. Becoming a Freemason or a Boy Scout is a relatively trivial undertaking, but members of these clubs differentiate those who are in from those who are out—and those who are out have no standing within the organization.

The American Enterprise Institute’s Danielle Pletka recently wrote about the virtues of political excommunication. Her point is fundamentally correct—parties and movements have to be able to police their own sides and cast out those who bring the larger group into disrepute—though it further raises the question of what kind of signal one might send that one is either an extremist or a they’re-coming-to-take-me-away-ha-haa nutter where there is actually a cost attached. At the moment, signaling either extremism or lunacy is fairly low-cost, and depending on what one is optimizing for, potentially very high return. (Indeed, to riff on Kevin Williamson’s recent column in The Dispatch, it’s possible that many of these signalers internalize the benefits and disperse the costs.) 

In debates about cancel culture, some suggest that the term is a misnomer, and what we really have is an “accountability culture”; this is essentially an argument in favor of raising the cost of various signals.

But this argument largely falls apart under scrutiny. First, much of what causes people to be on the receiving side of a cancellation is often private or very old speech or writing where a cost is only applied retroactively, sometimes decades after the fact. Second, the entire purpose of a costly signal is that people know the signal that they are sending; cancellation for signals that were not meant to be costly or even sent at all is at best akin to a stochastically applied tax, at worst like a street robbery. 

In general, bets can be a good “tax on bullshit,” in Alex Tabarrok’s memorable phrasing (and the recent Kalshi decision allowing bets on elections should improve the capitalization of betting markets). But even betting markets indicating that something as plainly ludicrous as the ABC whistleblower “story” is, in fact, complete bunkum, doesn’t break through enough to impose any real reputational or financial costs on the people spreading it. 

It’s not clear whether or how we can get back to the ability to send and receive costly signals. It’s institutions (and institutional leadership) that ultimately encode, decode and price signals, and we are living in a profoundly anti-institutional moment. (Somewhat paradoxically, this raises the relative value of signals from traditional elite-credentialing institutions: Ivy League schools, top business firms and the like. As other signals cheapen, the relative value of these kinds of signals increases, even as absolute values decline.) Michael Strong makes some interesting suggestions for improving the quality of signals sent by academics, though implementation of his proposals is both fraught and unlikely. 

One of the many tasks before us as we sort out our epistemological brave new world is to figure out how to once again send and receive costly signals. Speech should be, well, free—as in speech, not as in beer. For the moment, however, we’ve lost a valuable means of sorting wheat from chaff, true beliefs from bluster, real commitments from passing ones.

Until then, cheap talk and cheap signals abound.