How to Forecast Policy Decisions: Beyond material constraints

My previous note “Geopolitics”: What Is It Good For? mentioned that “geopolitical” analysis based on material constraints has potential shortcomings. This article unpacks that point.

A constraint-based framework of “geopolitical” analysis

Marko Papic's 2020 book Geopolitical Alpha provides a good example of a constraint-based framework of “geopolitical analysis”. Its core thesis is that, when forecasting policy outcomes, politicians’ preferences are “optional”, while constraints are cogent or mandatory. So, the prescription for investors is to focus on the latter and largely disregard the former. In particular, Papic recommends a focus on material constraints:

“…investors should focus on the material world of constraints, rather than the ephemeral of policymaker preferences and beliefs. The world of constraints is measurable and quantifiable whereas the world of preferences is not."

"Preferences are optional and subject to constraints, whereas constraints are neither optional nor subject to preference” (page 18).

Material constraints, such as the state of the economy and public finances, military power, geography, demography, etc. form an important part of the context surrounding policy decisions. This perspective is key to evaluating politicians’ election campaign promises, for example: Once elected, politicians often face material constraints that prevent them from implementing some of their promises. Papic’s book builds a systematic framework out of this core thesis and provides many examples of practical application.

However, rather than objectively identifiable “constraints”, investors need to consider policymakers’ subjective perceptions of constraints (and opportunities), when forecasting policy decisions. Moreover, any policy decision is based on the decision-maker’s expectations that different courses of action will entail different costs and benefits, as s/he tries to pursue a hierarchy of goals. So, what counts in policy forecasting are three things: a policymaker’s (1) subjective expectations of (2) potential outcomes, relative to (3) goals. Let’s unpack this.

The revenge of preferences

In a sense, contrary to Papic’s formulation...

Preferences are determinative: what counts as a “constraint” depends on preferences.

If I intend to go to the grocery store, the fact that I need to cross a busy road to get there is a constraint. But if I need to go to the chemist and I don’t need to cross that road to get there, that busy road is not a constraint. Something is or isn’t a “constraint” only relative to a given preference or goal. Therefore, policy analysts need to start by grasping what the policymaker is trying to achieve – i.e. goals and preferences. Any elaborate analytical edifice that focuses on “constraints” will always rest on that foundation.

This is important because analysts often get that part wrong. Explicitly or implicitly, constraint-based analysis often assumes that political leaders want to avoid damaging their countries’ economies, for example, or favor other “greater good” goals.

What policymakers want

Political leaders tend to have a hierarchy of goals (a series of ranked preferences) with respect to any decision, and the acquisition and preservation of their personal power is typically their top goal. Even highly ideological leaders, who would die for a cause, usually end up believing that furthering that cause requires them to acquire and preserve personal power. This applies to almost all political leaders in all political settings, but it has different implications for the forecasting of policymaking in competitive vs. authoritarian regimes.

In a system where political survival ultimately rests on winning competitive elections, the leader must care about being more popular than his/her competitors. That normally means promoting the interests of a majority of citizens. In that context, the goal of acquiring and preserving power is aligned with “greater good” goals. Hence, in most cases, analysts can mistakenly assume that the policymaker’s top priority is to promote the greater good and still get their policy forecast right. In fact, the policymaker wants to acquire and maintain power first and foremost.

The analysis of policymaking in authoritarian systems, by contrast, is not as forgiving… In such regimes, the leadership does not need to compete with political alternatives for public support and can influence public opinion through its control of the media. In such settings, the policymaker can prioritize self-interest at the expense of harming his/her country if needed, when the goals of self-promoting and the “greater good” are mutually incompatible or misaligned. In such cases, analysts that assume greater good motivations end up with incorrect forecasts.

Common analytical pitfalls

Geopolitical frameworks that rest on material constraints often rule out certain policy decisions as highly unlikely because they would harm the economy. However, when it comes to authoritarian systems, this misses the fact that policymakers: 1) do not place the economy above their acquisition and preservation of power in their hierarchy of preferences (this also applies to policymakers in competitive systems); and 2) the authoritarian set-up enables them to disregard the good of the national economy in order to pursue their personal interests, when those goals are mutually incompatible or misaligned.

Most analysts thought that Putin was bluffing and did not intend to start a full-scale invasion of Ukraine in February 2022, as that would attract massive sanctions, disrupt economic relations with the West, overstretch the country’s military capabilities, etc. In other words, material constraints overwhelmingly (supposedly) indicated that Putin would not invade. Even today, many analysts believe that it was irrational for Putin to take that decision, because it was clear that the country would suffer heavily as a result.

Instead, it is safer to assume that Putin, like the overwhelming majority of political leaders anywhere in the world, cares about the preservation and strengthening of his personal power more than anything else. Putin would choose policies that benefit his country, but only as long as that does not interfere with preserving personal power. In my article Why Did Russia Invade Ukraine?, I have proposed that the full-scale invasion of Ukraine in 2022 was largely caused by Putin’s regime insecurity. Putin’s decision to invade was rational relative to his hierarchy of goals, even if he understood the costs the country would incur.

Regardless of why one thinks Putin chose to invade, the point here is that an analyst seeking to forecast the high-stake policy decisions of authoritarian leaders must first form a correct model of that leader’s hierarchy of preferences. That model determines which factors are going to act as constraints on the decision. What analysts usually take as standard, material policy constraints – i.e. the state of the economy, public finances, military power, demographics, etc. - are often irrelevant for forecasting policymaking in authoritarian systems, where the personal interests of the ruler can override any other considerations.

Perceptions of constraints

There are additional reasons why a framework based on measurable, objective material constraints may fall short when forecasting policy decisions. Even assuming that the analyst gets the goals of a policymaker right, s/he still faces the difficult task of estimating the policymaker’s subjective expectations of costs and benefits attached to different options. Just like analysts, policymakers are also engaged in forecasting when trying to decide on a course of action. More precisely, a decisionmaker often must…

1. …face the limits of forecasting: Even assuming a deliberate, careful and systematic effort to try and predict the costs and benefits of a range of decisions, that prediction would often be impossible for anybody, including the policymaker. This is true when the decision pertains to a complex system characterized by large, irreducible uncertainty, which tends to be the case with high-stake policy decisions. The world is far too complex for anything approximating a confident estimation of the future costs and benefits of a range of policy options.

2. …rely on ideology or mindset: Given the limits of forecasting, the best that a policymaker can do is interpret the constraints they operate under - i.e. the costs and benefits associated with different options. But that interpretation is inevitably filtered through the lenses of a leader’s personal worldview or ideology, as well as mindset and personal background - leaving aside cognitive biases. Again, this means that what counts are not objective material constraints, but how the policymaker perceives and interprets them through his/her particular mentality and worldview.

3. …rely on available information: To the extent that some of the parameters of the decision problem are knowable, the policymaker depends on the quality of the information s/he receives on them. The process whereby bureaucratic machines (e.g. the intelligence services) gather and filter-up information to a political leader is often highly dysfunctional. That is true of all complex bureaucracies everywhere, but autocracies tend to be more susceptible to this, as uncomfortable information is even less likely to reach the leadership.

Putin’s decision to initiate a full-scale invasion of Ukraine in 2022 arguably reflected the three aspects listed above, in addition to reflecting his hierarchy of goals, as mentioned above. He could not forecast confidently how an invasion would pan out - i.e. the exact costs and benefits that would ensue: He had to rely on an estimation of those costs and benefits, which was inevitably colored by his ideology or mindset. For example, the notions of a hostile West bent on destabilizing Russia, that Ukraine is not a nation and its people would not put up a patriotic fight, that the West is too morally decayed and weak to unite in support of Ukraine, … these ideas probably played a role in Putin’s decision, whether they are correct or not (they are not!). He also had to rely on information that was fed to him about the state of preparedness of the Russian armed forces, the state of the Ukrainian armed forces, etc. – such information probably was partial and/or inaccurate.

Theory of policymaking ≠ theory of politics

In Geopolitical Alpha, Papic writes...

“…even the most incompetent policymakers are ultimately pushed to do the right thing by the constraints of the market, economics and politics. As economist Herb Stein said, ‘if something cannot go on forever it will stop’” (page 27).

That is certainly true, but it does not help us forecast policy decisions in situations where the policymaker can disregard “the constraints of the market, economics and politics” in order to pursue his/her personal goals, and where those two goals (personal vs greater good) are not aligned.

John Maynard Keynes famously said that markets can stay irrational for longer than most investors can stay solvent. We can paraphrase that into...

"Authoritarian leaders can pursue disastrous policies (which they see as benefiting themselves) a lot longer than analysts can stay in a job!"

It is true that, eventually, reality will catch up with such leaders. That means that a framework based on measurable, objective opportunities and constraints may be a good theory of politics, but it does not make it a good theory of policymaking, particularly in authoritarian systems. It is policy decisions that we often need to forecast (e.g. invade / not invade), not just the long-term implications of those decisions. The two are very different tasks.

Summing-up

This all means that the task of the analyst is not as simple as calculating objective constraints, but rather to estimate: 1) which goals the decision-maker is trying to pursue, in what order of importance; 2) how the decision-maker is estimating the potential future repercussions of different courses of action; 3) how the policymaker’s worldview and mindset are likely to color his/her interpretation of available information; 4) what information is available to the decision maker on the issues that s/he considers to be key.

Papic writes:

"like the path of a river is shaped by the feature of the terrain, the policymaker will choose the path of least resistance.” (page 27)

In fact, rivers and policymakers are profoundly different, for the purposes of forecasting their “paths”. Although hydrologic forecasting of, say, river floods is far from infallible, it does not have to contend with intentional and strategic actors who: 1) engage in deception about their real goals; 2) estimate the pro and cons of several courses of action relative to those goals; 3) are very bad at estimating those constraints in high-stake policy domains that involve high complexity and irreducible uncertainty; and 4) have to rely on ideas (ideology and mindset, plus biases) to make sense of the world, including opportunities and constraints.

All of this explains why, especially in authoritarian settings, policymakers often do not choose what analysts consider to be, based on objective and measurable constraints, “the path of least resistance”. Certainly, political leaders often end up paying for decisions that “defy gravity”, but we still need to forecast those decisions, as a lot hinge on them. In 1939, analysts could have argued that it would be foolish for Hitler to go to war against so many major powers. Ultimately, those analysts were proven right and Hitler was defeated, but that is small consolation for all that happened in the intervening 6 years.

How do you identify and forecast policy decisions that may affect your business or industry? I'd love to hear about your approach.