WaPo's Attempt to Help Liberals Argue With Conservatives Mistakenly Reveals Why We Struggle to Get Agreement

Edited 8/18 for clarity & brevity

On August 10th, Harvard Professor Max Kasy (economics) wrote a piece for the Washington Post (WaPo), titled: "Liberals are terrible at arguing with conservatives. Here's how they can get better."

This is a noble effort and you should read it. It's great to see some big thinkers putting serious and widely-accessible thought into bridging the partisan divide. Instead of, "conservatives are stupid, throw rocks at them," Kasy is putting the burden on liberals (which includes himself) to change their approach. This is great. If either side embraced such an approach seriously we'd be better off. However, there are a two issues with Prof. Kasy’s analysis:

  1. Kasy argues that liberals care about facts and outcomes, which fail to persuade conservatives who instead focus on processes. In reality, both sides care about both outcomes and processes.

  2. Kasy mistakes interpretations of analysis and projections for facts.

Everyone Believes in Both Process and Outcomes

Kasy looks at two divisive issues: healthcare and the estate tax. With each, he claims that liberals want to improve objective economic outcomes, whereas conservatives think it’s unfair to force certain people to pay for others. One side cares about outcomes, the other cares about processes. He claims that if you understand this distinction you can have a real discussion. For example:  

"Much normative (or value-based) reasoning by liberals (and mainstream economists) is about the consequences of political actions for the welfare of individuals. Statements about the desirability of policies are based on trading off the consequences for different individuals. If good outcomes result from a policy without many negative consequences, then the policy is a good one."

He then goes on to say that conservatives don't care about “value-based reasoning” or “the consequences of political actions,” and that they instead only care about how fair the rules of the game are. For example, people have a right to keep what they earn. This is a common conservative position, but when did conservatives stop caring about people having jobs, which is a “consequence of political actions”? When did they stop caring about GDP and income growth? They haven’t, and these are all examples of outcomes.

Kasy says that for conservatives, "as long as property rights and free exchange are guaranteed, the outcome is deemed just by definition, regardless of the consequences."

It’s an inaccurate representation of the conservative position to claim that they do not care whether property rights and free markets lead to better outcomes. If you ask a conservative, "do you think property rights and free markets lead to worse economic outcomes?" it’s unlikely that they’ll respond, "yes, but I don't care because those rules of the game are fair."

Look to Trump's campaign promises: Trump said he would create 25 million jobs in 10 years, emphasizing manufacturing. Policies such as tax reform, and renegotiating NAFTA, imposing tariffs on Mexico/China (or using a border adjustment tax) are all designed to achieve his stated objectives. Again, these are outcomes based on policies, not a detached focus on process.  

In fact, conventional conservative adherence to processes such as free trade and property rights are based--in part--in the belief that they lead to better outcomes. Free-market economist FA Hayek's famous manifesto, The Road to Serfdom, is just one example of conservatives making the case that government overreach leads to bad outcomes --both a decrease in individual freedoms and quality of life. In his other works on free trade, he repeatedly makes the case that free markets will lead to the best economic outcomes.

Now let's look at the other side: do liberals care about outcomes, with no regard to process? One economic analysis by the Heritage Foundation, a conservative think tank, shows that single parenting is positively  correlated with poverty-- more so than race or education. The Heritage Foundation proposes that government policies that encourage two-parent households and lead to greater economic prosperity for the poor.

Most liberals wouldn’t support such a policy. It would appear to many like an old-school, morally-conservative approach to managing people’s lives. Even if they could be convinced that a two-parent policy would lead to a decrease in poverty, many would likely oppose it on principle. Liberal opposition to this policy is a focus on process. Many liberals would not tolerate pressuring parents into a marriage situation they don’t want in order to achieve a desired economic outcome. Liberals, too, care about process.

Kasy even uses examples that reveal a liberal concern for process over outcome: He quotes Senator Kamala Harris (D-Calif) on the ACA:

"I feel strongly that when we’re talking about our sick, when we’re talking about our poor … and we’re talking about something that would deny those in need with the relief and the help that they need, that they want and deserve, it does put in place a question about our moral values.”

Kasy’s believes this a liberal, outcome-based statement: "In other words, if a policy will harm the welfare of individuals in need, it's a bad policy."

That's an incomplete interpretation. The senator emphasizes the poor “deserve” help, which is a matter of “moral values.” In other words, regardless of outcome the poor deserve help, which is a value judgment on the process of providing aid. When discussing taxation Kasy says conservatives think only of process and that people "deserve" to keep what they’ve earned, and that somehow liberals are being empiricists when they say people "deserve" help. The Senator is not talking about a complex analysis of the economic of health care. She's saying that people deserve help so it is just to give it to them.

You can cherry-pick quotes (as Kasy did) to try to demonstrate partisan tendency towards fact-based reasoning or moral rectitude, but ultimately both parties want better policy results that are implemented in ethical ways.

Kasy's "Facts" Aren't Always Facts

A more concerning aspect of Kasy's piece isn't the false process/outcome dichotomy: it's the misrepresentation of what is a fact. The idea that liberals "use facts" to justify their outcome-based arguments is something he doesn't question.

Here's how the Oxford English Dictionary defines "fact:" a thing that is known or proved to be true.

Not the most helpful definition, but we can infer from it (and experience) that facts are things that are absolutely indisputable. Policy analysis--such as economic or political projections--are not facts. They are best-guesses of a future, unknowable reality based on imperfect models and a number of variables interacting in complex ways. "It is a fact that all liberals/conservatives are idiots," is obviously not a fact, it's an opinion. "It is a fact that the poor deserve help," is not a fact, it's a value judgment. "The temperature at such-and-such location is 76 degrees right now," is a fact.

In reality, any policy proposal must depend on projections or models of reality. If someone says, "it is a fact that such-and-such policy will produce such-and-such outcome," that is not actually a fact. It is a projection. You may believe that the interpretation is obvious from the facts at hand, but that is often not the case. For example: if you ask 100 random economists how a tax proposal will impact job growth, public revenues, and the GDP, how many of them are going to land on exactly the same conclusions? Which of these conclusions are "facts?" None of them. They are projections, and entail both uncertainty and the risk of faulty or incomplete assumptions.

The best example of Kasy misinterpreting projections as facts is in his discussion on the estate tax.

Example from the Article: The Estate Tax

Professor Kasy argues that facts lead to the conclusion that estate taxes--taxes on a deceased individual’s wealth--increase public revenue. Conservatives don’t care about this "fact" because they believe that the estate tax is simply unfair.

However, here Kasy assumes he has a fact. What he has is a projection. He projects that increasing the estate tax would increase public revenues. However Stephen Entin, the president of the Institute for Research on the Economics of Taxation, claims that repealing the estate tax would increase public revenues. He makes an economic case for it here. Nobel Prize winning economist Milton Friedman has argued that the estate tax creates a perverse incentive and decreases tax revenues. Friedman and Entin have very different interpretations from Kasy as to the outcomes of an estate tax.

Who is right? I am not entirely sure; I am not an economist. But here is a case in which Kasy claims he has "the facts," when in fact what he has is a projection that is debated by competent economists (and Friedman is a Nobel Prize winner, not just some guy). Both Kasy and the two economists rely on models that anticipate future outcomes based on their interpretive modeling of past events. The past events are facts, but the future outcomes are uncertain.

Kasy claims that conservatives don't care about the fact that the estate tax provides more government revenue because they believe that the estate tax is wrong. Some conservatives may see it this way. However, others provide analyses that yield different conclusions from Kasy's. In other words, Kasy isn’t relying solely on facts, but thinks he is.

It's also worth repeating that conservative economists are talking about outcomes. They do care about public revenues. They do care about economic well-being rather than simply moral righteousness of the free market. Economic outcomes are a common part of conservative thinking on the economy.

And indeed, liberals care about more than outcomes when discussing the estate tax. Kasy says liberals believe it will reduce economic inequality and increase public revenues. However, when discussing how to argue with conservatives, he brings up the tactic to convince conservatives that it is inherently wrong for someone to receive a large sum of money that they did not work for. Might a liberal agree this is true? If so, it is a process value, not an outcome value.

Kasy suggests that “challenging the conservative value framework” is one effective way to persuade a conservative. In the case of the estate tax, “Our conservative likely believes that everyone has the right to keep the fruits of her labor, and free contracts of exchange between any two parties are nobody else’s business.” Challenging this “value framework” requires pointing out that "nobody can be said to consume only the products of their own labor." The assumption that conservatives believe that they only consume the products of their labor, however, doesn’t hold up well to closer scrutiny. From my own anecdotal evidence, I’ve never met a conservative who believes this, let alone conservative economists like Hayek or  Friedman. Again, what we see here is a conclusion that claims to be based on facts, but is instead based on a debatable interpretation.

The Danger of Mis-Labeling Interpretations as Facts

One dangerous thing about Kasy's article is that he (probably inadvertently) uses his position of academic authority to state his ideas as facts when they aren’t--they are economic projections and models, or interpretations of what he thinks other people believe. Facts can't be argued; projections and interpretations can. When one walks around believing that  projections and interpretations are "facts," they're bound to get frustrated when someone presents a competing projection or interpretation.

Ultimately, if you're sitting on a bunch of projections, calling them "facts," and then getting mad that others don't agree with your projections, you're going to make the dialogue problem worse. You're going to reflect your inner monologue ("this person doesn't believe in facts"), shut down the conversation, and alienate your interlocutor.

Lies, Damn Lies, and Statistics

Even when one is using some set of facts, one can fail to represent reality. One can mislead with facts and statistics.

Both sides are guilty of cherry-picking sets of facts and presenting them as accurate pictures of reality. If Donald Trump, for example, references a number of crimes committed by illegal immigrants in a speech, he's using facts. Do you object to that? Why? Probably because he's not accurately representing the rate of criminal behavior among illegal immigrants. The statement is meant to mislead people to believe that illegal immigrants are more criminal than others. But he is using facts; each crime on its own is a fact--it happened. Just because you are using some set of facts--things that are true--it does not mean that you are accurately representing reality.

Even more sophisticated sets of facts--statistics, polling, graphs of data, and the like--can twist reality or even intentionally mislead. Reconsider’s "Data in Policy Debate" series outlines a number of these. One such example is a Mother Jones piece that shows a positive correlation between gun ownership and gun death by state. As far as I can tell, this is accurate. But it implies that more guns means more death in the United States. However there is no correlation between murder rate (all murders, regardless of weapon) and the gun ownership rate. Both of these are facts, but the facts presented in a discussion should  be relevant to that discussion. If one is trying to limit the number of unlawful deaths or murders in the United States, the focusing on the total murder rate is more relevant than only the rate of gun deaths.


Kasy makes a good attempt to understand the conservative mindset. That he approaches the challenge of bipartisan political dialogue from a novel perspective, rather than focusing on some terrible caricature, is admirable. There’s something to be learned from this: everyone thinks they're doing the right thing, for different reasons. Most policy debates represent a clash in how we prioritize process-based values, and sometimes a desire for slightly different outcomes. We have to understand where other people are coming from to have any chance of having a real discussion with them. This is great.

However Kasy's own biases will leave a sour taste in many conservatives’ mouths. Claiming that conservatives don't care about outcomes is going to make them say, "you don't understand me." Saying that liberals always use facts to justify their policy proposals will make conservatives say, "that's certainly not true and hoo boy is that arrogant." Kasy moves closer to reality than the typical demonizing that goes in our current political environment but he's got another few steps to go to bridge the gap between inter-partisan political discussion.


Erik Fogg

We do politics, but we don't do the thinking for you.