WEEK 22

There is an old song, by Doris Day:

Que sera, sera

Whatever will be, will be

The future's not ours to see

Que sera, sera

What will be, will be

I am still obsessing about our ability to predict the future. I want to know what will happen to people who will outlive me. But also for purely intellectual reasons: I want to understand why we are so bad at it. In retrospect, we pretend to understand why history took a particular course: I spent a good part of my academic life explaining historical patterns and I thought I understood at least some of them. Yet, as I reported earlier, some weeks ago I asked a friend, a renown political scientist, what he expects. His answer was "It depends on the weather." This is about where I am still: every time I think I know what to expect, I talk to someone and my expectations change. So perhaps, "The future's not ours to see." Yet we incessantly trying to see it.

Samuel Huntington, who ex-post became the guru of the "third wave" transitions to democracy, published in 1984 an article declaring that the fall of communism in Eastern Europe was impossible. Jean Kirkpatrick, Reagan's national security adviser, went public with the same claim as late as in 1987. Juan Linz wrote the same in 1989 and had the misfortune that when his book was published a year later, it was all over.

Here I am willing to make a claim. I became persuaded that something big was underway in Eastern Europe in 1987. In 1988, at a meeting in Brazil and at another symposium in Montreal, I spoke about transitions from communism to capitalism in Eastern Europe. I was shouted out of the room, accused of being an idiot, a class enemy, and everything else. My intuition, which is all it was, was not derived from any lessons from history, indeed not from any systematic knowledge, but just from one conversation I had in 1987 while walking the streets of Warsaw with a friend who was a prominent communist. At one point during the walk, he said "You know, we are thinking of having a competitive election at the local level," to which I instinctively responded "But you will lose." His reply was "It does not matter whether we will win or lose but what we will lose." He meant that as long as the stake in the election were limited, as long as lives, freedom, and livelihood were not at stake and as long as Communists could freely compete in future elections, they were willing to compete and lose.

This one conversation oriented my academic work for several years to come. But I also tried to analyze the logic of reforms initiated by communists and I concluded that they were a slippery slope. Reading Soviet economists, I thought that one cannot introduce markets to a part of an economy while arbitrarily setting prices in other sectors. When you begin doing what Gorbachev and his team of reformers were planning to do, there is no way to justify the rest of the communist economic model. Once you take that first step, you have to go forward. It's the "bicycle theory": if you don't keep going, you fall.

I plunged into this retrospection because I am trying to understand what it takes to predict the future. In the case of the fall of communism, drawing lessons from history -- which is what Huntington, Kirkpatrick, and Linz relied on -- turned out to be misleading. In turn, just a vague intuition combined with some logic turned out to be accurate. Yet, having read several memoirs of the 1930-1938 period in Germany, I was struck that no one, from eminent politicians to ordinary housewives, predicted what would eventually happen. Even in Chile, where by late spring of 1973 everyone knew that a coup was imminent, no one expected that it would be so bloody and that the dictatorship would last 16 years: the modal prediction was that the military would depose Allende, send him to Cuba, announce a new election that would be easily won by Eduardo Frei, and that would be it.

In all these situations -- Communism, Germany, Chile -- we just did not have a theory on which we could rely. We had no science that would generate valid predictions, not even attach probabilities to the possible outcomes, courses of history. As King and Zheng have argued, inductive learning from the past works when the current conditions resemble the conditions that occurred several times in the past. But drawing lessons from history works only if we correctly identify the relevant conditions and -- this is what I learned from some statistical analyses I reported earlier -- their combinations. We need theory: logically interconnected propositions that say that "if this and that, then this," the last "this" being observable. The brutal fact that we find it so difficult to predict what will happen under our current circumstances is evidence that we do not have theories we can rely on. Economics relies on "forecasts," the reliability of which decreases steeply as time horizon for which they are made increase. Political science has an elaborate theory that predicts outcomes of elections but at least in the US it becomes reliable only after the Labor Day. And without theory, we can rely only on conjectures, intuitions, or guesses.

This is where we are. But I cannot stop myself from speculating. I have done it before, so some of what follows is repetitive, just updated by recent information. To anticipate my conclusion, I now see the mid-term elections as the only possibility of throwing some sand into the MAGA wheels. If Republicans retain the control of both Houses, Trump's power will be despotic in the sense given to this term by Montesquieu: his will, including his whims, will be the law. If Democrats win at least one house, conflicts will not be over but their outcome will become even more uncertain.

The economy is the object of forecasts. I looked at six most recent ones, from Namura, Barclay, BNP, Deloitte, Conference Board, and the OECD. The range of forecasts of GDP growth for 2025 is 1.4 to 1.7 percent; for 2026 it is 1.3 to 1.6 percent. The range for core inflation is 3.0 to 3.9 percent in 2025 while everyone expects it to be 3.2 percent in 2026. Forecasts about the job market predict a slight increase of unemployment and a slight decline of job creation. While all these expectations are worse than under the two last years of Biden, none is catastrophic. The most grim forecasts are for the housing sector: interest rate on 30 year loans are expected to remain at almost 7 percent, housing starts are falling and expected to keep falling, and prices are expected to remain high. Still, a survey of Chief Financial Officers of large corporations, conducted by Teneo, show that 53 percent of them expect the American economy to improve in the second half of 2025; the consumer confidence index is recuperating from a historic low, and the stock market is not collapsing. Hence, those who expected the economy to improve under Trump should be disappointed but those who expected the economy to just tank will also be. The economy is not performing well, positive expectations are frustrated, but there have been many periods like this, so the forecasts predict nothing dramatic.

Trump's overall approval ratings as a few days ago were 42% approving and 54% disapproving according to The Economist, 44% to 52% according to the NYT, and 44.7% to 51.0% according to Nate Silver. All these three sources show that his current approval is at the lowest level since he took office. On particular issues, he is judged most negatively on health care and negatively on all economic aspects, particularly inflation, which is most salient for the respondents. Strikingly, he now also gets net negative evaluations on immigration. According to Silver, his net approval on immigration declined from a peak of +11.1 points on March 6 to -6.5 as of the most recent date. Hence, as of now Trump does not get net positive evaluations on any issue.

This is what we observe. Now theory. While the prevailing belief is that economic conditions, in particular inflation, is a good predictor of election results, there are all kinds of nuances. The economic forecasts summarized above do not provide a strong prediction. Excepting the post-Covid years, the rates of inflation predicted for the remaining period of 2025 and for 2026 are somewhat higher than the historical average but not much. Income growth data are somewhat lower than their historical average but again not much. In turn, the fact that Trump's overall approval continues to be clearly higher than the approval of all his particular policies indicates that his core base of support is solid. This is all, I think, we can tease out from the current information.

These are the tangibles but now we have to enter into the realm of speculations. What is the size of Trump's core base, by which I mean people who will support him no matter what? Trump is still supported by 42-45% of survey respondents. These numbers have changed very little from mid-April: his net scores declined mainly because of an increasing disapproval. I have a bet whether his approval will decline to below 38 percent by the end of 2025, with me betting it will not. G. Elliot Morris in his Substack estimates "hard core Republicans" to be as low as 25 percent. This seems to be too low for me, but I do not know how to estimate it. Moreover, one should not assume that dissatisfaction with Trump automatically benefits the Democratic Party, which remains catatonic. But then people who vote in mid-term elections tend to be more educated than those in presidential years, which should favor Democratic candidates. Finally, the irregular elections that took place here and there in the past six months have been good for Democrats. So I am willing to believe that Republicans would lose the House and perhaps even the Senate in the mid-term election, if elections are held and if they are clean. (P.S. I do not like elections being described as "fair" because they always take place according to some rules and all rules have an effect on their outcomes.)

Now, I still cannot imagine Trump not holding a scheduled election. I can, in turn, imagine that the election will be preceded by a wave of repression and intimidation, that it will be accompanied in some places by decentralized violence and by fraud, and that every marginal Democratic victory will be contested in the courts. This may or may not be enough. The experience of "backsliders" in other countries indicates that somehow they succeed to win national level elections but also that they suffered defeats in some local ones: Budapest, Warsaw, Ankara, Istanbul. So perhaps Trump cannot prevent Mamdani but can still win the legislative election. The question is whether he can win given the size of his core base and all the levers of power in the hands of the federal government. There is a famous anecdote about the one-time dictator of Nicaragua, Anastasio Somosa, who told his defeated opponent "You won the voting, but I won the counting." So the conduct of elections is yet another source of uncertainty.

I argued some weeks ago that the Democrats would be able to do little if they do win the House or even the Senate. But I still think that the mid-terms are the watershed moment. Perhaps this is because I can see few political forces that would restrain Trump. Dissensions within the White House generate chaos but not paralysis; the non-MAGA Republicans in the Congress are sufficiently intimidated to keep in line; the Supreme Court does nothing to limit Trump's discretion. The only visible resistance comes from governors of the large Blue states, so the issue of federalism may continue to flare up. Hence, if Republicans win both houses, no forces of resistance will be left. Trump's rule will be despotic. In turn, Democratic control of the House would allow them to investigate ICE, attract public attention to corruption by the administration officials, and scare Republicans still holding to marginal seats, perhaps loosening Trump's control over the party. It may be not more than sand in the wheels of power, but it would galvanize the prospect that Republicans would lose the presidential election on 2028.

One can speculate further, asking what we should expect to be the MAGA troops reaction to an electoral defeat. Their stakes will be very high, ideological but also personal. Given what I believe about the conditions for elections to be peaceful, I do not think that they will accept the defeat quietly. In fact, we have already seen what happened in January of 2020. Hence, as I see it, Republican victory in the mid-terms would be the end but their defeat would only intensify the conflict.

Going beyond the immediate future, the question that looms is "How can it all end?" One possibility is clear: Democrats win the 2028 presidential and congressional elections, dismantle the apparatus of repression, try to selectively restore essential social programs and services, and we will be back to "normal." So is another one: Republicans win the 2026 mid-terms, the 2028 elections, and usher in an oligarchical, repressive regime for an indefinite future. Trump may die in the meantime, which would affect the probabilities of these outcomes but, whoever would be his successor, these outcomes would be still institutionally determined. All other outcomes would be more dramatic and unprecedented in the history of the US. One is that Republicans do not accept a defeat, either in the mid-terms or in 2028, or generate some event like the Reichstag fire which they would use as a pretext to declare a state of emergency and attempt to impose their rule by force. The partisan postures of armed agencies, including the armed forces, would then become decisive. Finally, it is perhaps also possible that Trump's popularity tumbles to very low levels, street protests reach tens of millions, and the Republicans, freed from his control, seek some kind of a compromise.

Now, as I realize that I am chewing over and over about the same uncertainties, it is time for me to quit, at least for some time. I do not see us making any progress in predicting the future until some uncertainty is resolved. Economic forecasts are quite reliable within the window of a quarter while election predictions become solid about two months before elections are scheduled. Hence, unless some major random events intervene, all I can do is to wait. I have been a political scientist for over 60 years, I know some economics and some sociology. I had some good hunches about things to come in two other countries where I lived. And yet I am not any more clear about the future of the US than anyone else. So I am left with "What will be, will be."