All scholars of international politics and perhaps others should read this interesting book. It makes valuable contributions, both specific and general. It develops an original, convincing explanation of the phenomenally rapid spread, especially during the 1990s, of bilateral investment treaties [BITs] involving developing countries, and of the treaties’ puzzling content. It also demonstrates an insightful way to check and enrich rational choice analyses of many other political phenomena.

Most of the world’s nearly 3,000 BITs are quite similar; they grant foreign investors vague, far-reaching rights, and typically do not give the host government any rights to make claims on foreign investors. Most give the investor the right to settle any dispute outside the host country’s court system, for instance at the World Bank’s International Centre for the Settlement of Investment Disputes (ICSID). There, often the arbitrators are private commercial lawyers, who have authority to order a settlement of unlimited size that is binding on the host government, with few real options for appeal. In an extreme 2012 case, a tribunal awarded a U.S. company compensation from Ecuador equal to almost 7 percent of the government’s budget, despite acknowledging that the investor had broken Ecuador’s laws and its contract with the government. An equivalent claim on the United States, proportional to its GDP, would have come to $458 billion.

The book’s puzzle is “why did practically all developing countries suddenly rush to sign largely identical treaties, which significantly constrained their sovereignty? Why did they expose themselves to expensive investment claims and give such a remarkable degree of flexibility to private lawyers to determine the scope of their regulatory autonomy?” (p. 5).

Poulsen challenges explanations given by many political scientists and lawyers. A prominent argument claims that governments signed these treaties because they were rationally competing with other host countries to attract new foreign investment. A host government that is not bound by a BIT would have incentives to renege on promises later, so it needed the treaty to make credible its commitment to treat investments fairly. If competing countries signed BITs and it did not, investors might go to the countries with BITs. Governments weighed benefits against costs, strategically decided to sign, and maximized the agreement’s value. This rational-competition hypothesis has been supported by significant results from cross-national statistical testing.

The book raises serious doubts about this explanation. Poulsen says a 1991 World Bank survey of foreign investors reported that in fact BITs had played a negligible role, if any, in investors’ decisions. Political-risk premia were not high in most host countries, and investors concerned about risk had ample ways to protect themselves without BITs (pp. 6–8). This survey was available before governments went on their signing spree and later surveys largely confirmed it. Moreover, many former officials say that when they were deciding to sign, they rarely paid attention to what other potential host countries were doing.

Poulsen also doubts that Western capitals coerced weaker developing countries into signing against their will. Typically, host countries initiated the negotiations after liberalizing their investment regimes at home. He dismisses the view that they emulated what was expected by global norms rather than expecting material gain. Many officials genuinely believed their BITs were important for attracting investment and that was the main driver in their decisions.

The central argument is that so many developing country governments signed hundreds of largely identical bilateral treaties entailing large potential risks in exchange for little added benefit because, while they intended to be rational, their rationality was severely bounded. Their decisions in many cases were skewed by information processing biases that are well established in cognitive psychology.

The book offers a simple theory of BIT decisions that modifies standard rationalist thinking by adding these bounded-rationality insights. The theory predicts three ways in which biases will mar developing country decision-making (Chapter 2). This theory does not attempt to predict which countries will sign a BIT or when these biases will prevail or be overridden, but it does make predictions about treaty content.

One familiar bias is attending disproportionately to the most salient information. The ICSID, the UN Centre on Transnational Corporations, and Western government advisers made BITs salient by actively pushing them, telling developing countries they would attract investment. South Africa later used its first model treaty (from the UK), rather than other models in use, for additional countries because the British had been the first model to come to their attention (p. 175). Even after some countries began to experience claims in 1987, officials in other countries typically have not slowed rapid treaty adoption until after their own countries were hit by claims (p. 143).

A second recurring bias was motivated reasoning. Poulsen says many busy officials based their expectations of risk-free economic benefit on wishful thinking. They did not conduct careful, unbiased searches for relevant information to verify their wishes. In 1994 South Africa “completely disregarded the risks” that serious claims could arise under its BIT with the UK (p. 170), even though they had an unusually powerful incentive to search (this treaty could override the most sensitive compromise—over land redistribution—in their new constitution), and even though information was available. International organizations had published the report indicating that BITs had little effect on investors, as well as guidelines alerting officials to pay close attention to treaty terms. After Italian investors filed a $350 million claim in 2005, South Africa stopped signing agreements involving investor-state arbitration.

A third bound on rationality was status quo bias, taken to mean an excessive preference for whatever solution is the “default.” This bias is used to explain the remarkable similarity of treaties over time and across countries. Many host governments accepted short Western-model treaties with tinkering at most, rather than carefully considering alternative treaty designs, which some now regret. In fact, former officials of many countries said they often suggested a BIT for a photo opportunity when traveling ministers wanted something to sign, including with other developing countries that had little capital to export, fueling the spree. Even after experiencing an investor claim, many governments have responded with only incremental change to the models they already know (Chapter 6).

Poulsen gathered a remarkably impressive array of evidence, some of it jaw-dropping, relying most on qualitative techniques and analysis since the hypotheses center on decision making. The book reports no fewer than 18 short country case studies from all regions of the developing world, based on interviews with participating former officials and government documents. He also over-sampled middle-income countries, which are less likely to confirm his predictions than low-income countries where one might expect less expertise and experience. He reports three deviant cases—China, Costa Rica, and Turkey—but most confirm. Statistical analysis is used for two subsidiary questions where defensible cross-national data are available.

Interview evidence is sometimes questioned as possibly self-serving, but much of this evidence makes the speakers and their countries look bad. For instance, when a Swiss firm filed a claim against Pakistan, a Pakistani official asked his Attorney General what a BIT is. He admitted, “to be perfectly honest, I did not have a clue.” The Attorney General of Pakistan typed into Google, “what is a BIT?” (p.xiii). He even had to ask Switzerland for a copy of the treaty his state had signed.

More generally, this book also has implications for rational choice analysis of many other political phenomena. Its inferences are more convincing to me than those drawn from theory and methods that are more prominent in U.S political science today: full rationality, statistical methods, and country characteristics as causal variables. If host governments had been guided by fully-rational decision making, we would observe either no BITs after 1991, in favor of more effective tools for attracting investment, or at most some treaties that fit host countries’ long-term interests much better. The full reasons for this divergence are worth debating elsewhere, and they might include the particular statistical models chosen. But clearly country characteristics are farther back up the causal chain from the behavior to be explained than the decision processes documented here. Future projects using mainstream methods might consider adding bounded-rational, qualitative research that can study missing variables closer to the behavior, or publishing calls for such work to check their conclusions. And for all political scientists, this book illustrates a model that could be fruitful for many projects on a range of subjects.