Policy and International Relations - What Are They? (Primer for EAs, Part 2)

By Davidmanheim @ 2020-01-02T12:01 (+26)

(This post follows the introduction to this sequence. I think it's worth reading first, but it can be skipped, as it mostly discusses why this matters, and who should care.)

Before trying to understanding how to use political science, public policy, geopolitics, and international relations in informing our decisions about how to have a maximally effective impact on the world, we need to know what they are. This post is intended to explain some of the basics, and address some key questions what these disciplines are, and are not. (If "what they are" seems obvious, you need to read the post.) Before the overview, I'll explain why I think the providing this overview is useful, and after providing the overview, I'll try to justify why all of the different approaches are important to consider.

What does Political Science have to do with Effective Altruism?

Effective Altruism is essentially a practical discipline, not an academic one. For that reason, it should have the same relationship with these disciplines that politics does with political science, or coaching does with playing a game. Just like the best coaches may not be good at playing, experts are not always good at predicting the future, much less improving it. Still, the right coaches can make players play better, and the right academic insights can help effective altruists do a better job${}^1$. That's why I think it is important to know a bit about the different academic disciplines, their approaches, tools, and objects of study.

But academia doesn't usually align with practical needs, and the various disciplines have related and partly overlapping goals, methods, and subjects. At the same time, they are in many ways different from each other, and different from history, economics, law, and politics, but there is significant overlap and room for confusion. This very brief introduction is useful to understand where to use which term, what tools they use, and what they can offer. In future posts, I'll do a deeper dive into some of the specifics.

A Brief Overview of Academic Approaches

Political science is a broad and all-encompassing term that in the broadest sense, deals with how groups of humans make decisions and interact with other groups. With this very high-level view, international relations, public policy, and geopolitics are considered specific areas of focus${}^2$. This can be misleading, however, because there is a very wide range even within what is usually called political science, and there are also theories and methods used in political science that are different from, or parallel to those used in each of the sub-disciplines. Some of these are related to tools and approaches used in applied areas, but different terminology is used, and I will try to draw these parallels and point out where people are looking at the same thing with different lenses, or calling the same lenses different things.

Some of the primary approaches in political science are rational choice, behavioral, institutional, and Marxist theories. These will be discussed in the next post - but as with the current post, if you think that it's obvious what each approach is, I'd suspect that you will benefit from reading more. And despite fundamental flaws in each approach, they have useful tools, and are all valuable${}^3$.

International relations and international affairs lies at the intersection of global politics, economics, and law. The focus tends to be on political units and how they act. This includes not just states, but international organizations like the UN, as well as NGOs, and to a lesser extent, non-state civil groups such as religious organizations or aid groups, and international corporations. Modern international relations is different from political science applied internationally because of a greater focus on history and law. Another key difference is that international relations tends to look for general theories of international relations. These often include economic, legal, historical, and other features. Some key approaches in international relations are rational actor models, process models, and social constructivist models.

Geopolitics deals with the different parts of the world, and how they related to one another. It is a primarily qualitative and descriptive approach. As the name suggests, geopolitics is related to geography, certainly much more so than international relations. It often attempts to provide an external view of the world, focused on global trends and how physical and human geography relates to politics and international relations.

Geopolitics also focuses on how people view the world, and what conceptual approaches exist, and how they matter. For example, consider the terms "the axis of evil," "the arab world," and "muslim countries," and how they imply different ways that a large part of the world is understood by those using the terms. Similar points can be made about the phrases "the free world," "western democracies," "liberal democracies," "first world countries," and "developed economies." In both cases, the different terms used imply a viewpoint about the world which may be insightful, or misleading. This matters both because our understanding informs how we think about the world, and because we want to communicate with others whose conceptions differ from our own.

Public policy and policy analysis is${}^4$ in some ways the odd-one-out in this list. The academic discipline of policy analysis focuses on quantitative (econometric) analysis of the impact of policies. The tools used are sometimes identical to the approaches used in evaluating effective altruist interventions, though the results are often presented, evaluated, and used quite differently.

The applied discipline of public policy focuses on supporting policy decisions with quantitative decision support models such as those used in effective altruism, as well as qualitative tools like wargaming and comparative historical analyses. The approach tends to be interdisciplinary, and draw on all of the above areas of political science, as well as many others. In general, it focuses on analysis of the impacts of decisions and supporting decision making, rather than positive descriptions of the process. I confess that its inclusion here is a reflection of my background in the area, but I will argue that despite the fact that there is overlap, and many of the ideas are well known, the knowledge and experience of policy analysis has much to offer effective altruism in achieving the goals of improving the world. Not only that, but it offers a paradigm for how to reasonably pull from multiple disciplines in helping make decisions - exactly what this series of posts is trying to help with.

Why Not Just Pick One Approach?

Now that we have a broad overview of each topic, I'm going to make a very general argument for why we want to consider all of them, or at least more than one${}^5$. In short, because decision making in these areas is hard (as I argued in the last post,) we don't have confidence in any single approach. Despite this, each of the approaches has some valuable insights, and combining multiple incompatible models works really well in practice. By considering different approaches, we can also mitigate the problems of expertise-induced blindness.

There are trade-offs to this approach. First, it's a lot more work. Each approach costs time and money to investigate, and both are limited. Second, it makes things much less conceptually clear, and harder to explain. Instead of having an explanation for how things will work, you have a group of explanations, and each makes different claims. Lastly, it's much harder to come to a clear conclusion. When four models say something will improve the world, and three say it will make it worse, it's not as though we can have the approaches vote and be confident in getting the correct result.

On the other hand, we can see counter-arguments on each point. First, making decisions without considering all the approaches can lead to mistakes and bad decisions. There are reasons to limit the investigation, but investing a week of thinking using one approach seems obviously inferior to spending a day considering each of five approaches. In fact, one wrong but conceptually clear and easy to explain answer is exactly the hedgehog approach that Tetlock's research showed is ineffective${}^6$. Finally, reality itself is complex, and ignoring uncertainty doesn't make it go away. If we aren't confident that something is good after considering many viewpoints, that's a sign that we shouldn't be confident.

(1) In a later post, I hope to talk more about how to apply expertise to substantive questions of interest. This relates to the discussion at the end of the post, but actually applying insights requires further discussion.

(2) Interpreted even more broadly, the field includes not just political economics, but macroeconomics generally, as well as law, history, and almost any social science - and while some of political science does overlap with each of those areas, calling everything political science is unhelpfully broad.

Note also that Political science, like economics and many other fields, can be somewhat ambivalent about whether it has a normative approach (i.e. what norms should exist, or what should be normal,) or a positive approach (i.e. describing what actually is.) I will focus on the positive aspects of political science.

(3) Even if the tools and ideas weren't valuable at all, knowing the differences and different perspectives will be useful when interacting with experts.

(4) Note also that the related area of public administration is more closely related to business school and management methods than political science, but shares many tools with applied public policy.

(5) Obviously not all of these are relevant to a given question, but when looking at, say, how to regulate bioweapons, I'm arguing that it's important to consider many lenses - geopolitics, neoliberal theory, rational actor approaches, and international realism.

(6) See the previous post.