Debate is frustrating and pointless and vital to our Democracy.

Debating is the natural consequence of people having different political views in our society. The debates I am referring to take place around a single topic — gun control, illegal immigration, the Electoral College, you name it — and are contests of facts, not which “values” you possess. They take place in both formal setting (like competitive or parliamentary debate), and casual setting (like arguments with your friends or family). I have taken part in both types, and have found it to be a very important — yet fundamentally flawed—activity. Let me explain.

At the end of the day, debate comes down to which side the facts point to. Now, if that is all that debate were, it would be a boring activity. Debate’s strength is being able to optimize which facts are the most important. Because both sides of the debate are looking to shift the conversation to the facts that benefit their side, observers are able to get a broad view of the discussion. Any observer could plainly see which areas clearly benefit one side, which areas clearly benefit the other, and where the main points of contention lie.

These points of contention are where the facts really become important. And they are precisely where debate starts to become a pointless activity. Imagine the debate over gun control, where the debate comes down to whether one specific fact is true: whether making background checks and 24-hour waiting periods universal, and increasing funding for background checks, would reduce a significant number of gun deaths. Now, in gun control and other debates, there is often more than one key question (and perhaps this question isn’t the best). But I also think that people often exaggerate just how many key questions there are. Most debates, I have found, could be easily settled if two or three specific facts could be conclusively determined.

Let’s get back to this gun control debate that comes down to one question. The argument (again, usually in the real world there are more than one, but stick with me) in favor of this gun control is that it would be harder for criminals or those about to commit mass shootings to purchase guns. The argument against is that these people could just as easily purchase guns on the black market. When there is a question like this, the only conclusive way of answering it is to turn to research. But there are two big problems.

1. Difficulty of Research

The first issue in answering the question through research is that it is based on a hypothetical and thus difficult to measure. We can maybe look at other countries that have implemented such bans, or look at regions of the US that have, or look at past instances of the US implementing the bans (in this case it actually did implement these in the 1990s), but none of these are a perfect comparison. Other countries may have different gun culture or different neighboring countries (that guns can come from) than the US does. Regions of the US often cannot restrict guns coming from their surroundings. And US culture may have changed significantly since a ban was last attempted, or maybe the ban simply wasn’t in place for long enough to become effective (not to mention that for most other issues, the US has not adopted debated proposals in the past).

Yet even if you could find a perfect comparison, it is fundamentally difficult to measure the effect of this proposal. You could look at how gun deaths as a whole were effected, but these deaths could also be affected by other trends of the day. This is especially a problem for policies that take a long time to take effect, like gun control (as it would only affect new gun sales, not guns already in place). You could try to see how many people got guns from the black market, to more intricately answer the question, but that is a very difficult thing to measure.

And yet even if there were perfect ways of studying something, there is still a large issue to overcome: which study do you look at?

2. Narrowness of Scope

For most key questions, there is (at least a little bit of) research favoring both sides of the debate. And even if 99% of the research supports one side of the argument, media sources with political biases are quick to highlight the 1% of studies that support their side.

Now, debates last a finite amount of time, so you cannot highlight literally all research that has ever been done. Debates are thus extremely vulnerable to cherry picking, picking only the research that benefits your own side. Cherry picking by itself is an issue essentially impossible to overcome. But debate’s flaw is more than just people pointing to specific studies. A bigger problem is that it is near impossible to refute a fact that the other person pulls out, even if this fact is fundamentally flawed.

In some debates, such as some competitive or televised ones, this problem goes as far as being impossible to tell whether someone is just flat-out lying. Sure, you can check after the debate which lies independent fact-checkers point out. But this does not affect the flow of a debate in real-time, which is the only part that really matters. Independent fact-checkers themselves are also vulnerable to bias in which sides they check more, and in how they choose to classify facts (there is often a spectrum from true to false).

But often, facts are not so much straight-up lies as much as they are based on flawed studies. Studies can be flawed for a variety of reasons — from factors not accounted for to unrepresentative samples to simple mistakes in code (i.e. the computer output was just wrong due to a programming mistake!). And many of these issues require scientific experts to diagnose, nothing that an ordinary debater could do. Even if the debater quotes an expert in debunking one study, the other side can just pull out a different study. Then what is the debater going to do?

The only practical solution to all of these issues is to take a broader view of studies. This means looking at many different studies, each of which may use a different comparison metric and measurement technique, and trying to find a consensus.

In practice, there are two different ways of showing this off in a debate. The first, often used in internet debates, is to find a page that presents an overview of research. Many new sources, from Vox on the left to National Review on the right to academic papers , have articles that do this for a variety of issues. With these, we find the same problem we do with just citing studies themselves: cherry-picking. You have to trust that the page that is doing an honest assessment of the available data, which has become much harder now that traditionally trusted news sources like the New York Times and Washington Post have come under fire for bias.

The second way of illustrating the consensus in research is to point toward the opinions of experts. In practice, this is the same as pointing to an academic overview of papers (i.e. a research article that simply reviews all available research). Here, we run into the same issue of trust. Scientists themselves are supposed to be non-biased, and can pretty much be trusted to not flat-out make up data. However, it is significantly more difficult to trust a scientist in their judgement in reviewing all available research. This is where politics and traditional research differ: in politics, people are much more invested and emotional over issues, so it is harder to trust that someone is doing a fair assessment of data. In issues of social science, additionally, experts quite often disagree, so even citing them is vulnerable to cherry-picking!

At the end of the day, there can simply be no definitive winner — often not even close to a definitive winner — in a policy debate. Which is a little frightening, given that most policy debates shape the lives of millions of people. But debate is still a worthwhile activity.

The value of debate

Debating, both taking part in a debate and simply observing, can still have a lot of value.

Firstly, as I mentioned previously, debates can help give an overview of political issues to observers. While some issues in a debate come down to facts and are thus impossible to truly settle, many issues in a debate do not. Good debaters recognize that it is necessary to concede on some questions in order to shift the focus to more two-sided issues. For example, a debater arguing against gun control may not make the claim that gun deaths are not very prevalent in the United States. Instead, they may choose to question the claim that more gun control would actually reduce gun deaths. I think that many observers are often less informed than many give them credit for (I realize that sounds extremely snarky but it is my honest belief). Debates can thus serve as good ways of introducing different policy issues to observers.

Secondly, debates are often worth having simply because most other forms of political communication are much worse. A presidential debate, for example, is often perceived as not extremely useful, given that the audience generally perceives their favorite candidate as having won the debate. However, on most networks, the alternative to broadcasting a presidential debate is often just one-sided coverage. At least with a debate there is someone on stage presenting both sides.

So, although debate has some serious flaws at the center of it, it is probably still worth doing. I think it is also a ton a fun. But maybe that’s just me.