Some defenders of President Trump have tried to justify his trade rhetoric as just that: words. They argue that Trump will never enact the tariffs, or do so at much lower rates, so firms and individuals will not change their behavior. However, this is a mistake. We all make choices in an uncertain world and our actions depend, in part, on our internal calculations of the likelihood of different events (not to mention our own risk tolerances). With these calculations, we make our actions. If something comes along to change our perceptions, then we also adjust internally.*
Let’s, for the sake of argument, say that Firm X is considering expanding operations in the US. Let’s say they want their annual profits in the US to be $50 million/year. They need to spend $100 million to enter into the US and set up. Through various research, the firm concludes that, given current expectations on 1 November 2016, they have a 95% chance of achieving that goal (or a 5% chance of earning nothing and losing their $100 million investment. The binary nature of this situation, while perhaps unrealistic, is done so the point doesn’t get lost in the math).
So, the expected value of their operation on 1 November is ([95/100]*50)+([05/100]*-100)=$42.5 million.
Now, let’s say that Trump wins the presidency and the likelihood of tariffs has increased. Given that the firm imports some of their material and may wish to import more in the future depending on conditions, they now conclude the likelihood of earning $50 million per year has fallen to, say, 80%. So, now, the expected value of the move is: ([80/100]*50)+([20/100]*-100)=$20 million! That’s a decline of 52.9%!
The firm is now faced with a decision, just as all firms are, based on their assessments of probability. The expected value of the firm’s move has fallen, which means the firm is less likely to enact the move at all. And all this is based off a change in probability, not just the tariffs themselves.
Another important thing to note is this still holds with small changes in probability, too. Let’s say, for example, the firm’s initial probability of 95% still holds. However, after the election the firm determines there is still a remote, but still real, possibility of tariffs and thus their probability estimate falls to 94.9%. Their expected value drops to $42.35 million, a decline of just 0.04%. But it is still a decline. What this means is many marginal investments may be cancelled. These now nonexistent marginal investments, which would have produced jobs and goods for Americans, are weighing on US economic growth. All because of words.
With rhetoric flying about, it becomes harder and harder for firms to determine their probabilities. In this sense, it is no different than arbitrary policy changes that causes other forms of regime uncertainty. In short, rhetoric matters, even if it doesn’t translate into policy.
*For the sake of clarity, I am not saying people walk around and manually calculate probabilities before each and every action they take, but they often act as though they do. For example, if a person knows a particular part of town has a higher crime rate than another, he may avoid that part altogether. However, if that part of town becomes more safe, he may venture in there more often. These decisions are made based on his perception of probabilities, not his actual probability.