The ever-funny Richard Hanania has published his first book: Public Choice Theory and the Illusion of Grand Strategy: How Generals, Weapons Manufacturers, and Foreign Governments Shape American Foreign Policy. You can obtain it the usual place. The publisher describes it as:
This book argues that while the US president makes foreign policy decisions based largely on political pressures, it is concentrated interests that shape the incentive structures in which he and other top officials operate.
The author identifies three groups most likely to be influential: government contractors, the national security bureaucracy, and foreign governments. This book shows that the public choice perspective is superior to a theory of grand strategy in explaining the most important aspects of American foreign policy, including the war on terror, policy toward China, and the distribution of US forces abroad. Arguing that American leaders are selected to respond to public opinion, not necessarily according to their ability to formulate and execute long-terms plans, the author shows how mass attitudes are easily malleable in the domain of foreign affairs due to ignorance with regard to the topic, the secrecy that surrounds national security issues, the inherent complexity of the issues involved, and most importantly, clear cases of concentrated interests.
Hard to parse? Here’s Hanania’s own summary from the conclusion:
American foreign policy is better explained by assuming that individuals within the system seek out their own personal gain rather than assuming actors work toward goals consistent with some conception of the national interest. This is consistent with a review of the most important aspects of American foreign policy, among them the circumstances under which the United States uses force abroad, how it responds to rising powers, and the major decisions made over the course of the war on terror. The point has not been that American policy is misguided, rather that it cannot be said to be “rational” in the economic sense, meaning working toward specific goals, at least over a substantial amount of time. Rival powers are built through American technology and economic engagement while the United States plans on how to balance against these same states. Countries are invaded based on one justification, and then troops stay for some new goal that has nothing to do with the original purpose of the war as the occupation continues mainly in order to provide cover for domestic political actors. Even on the most important issues, long-term planning is practically nonexistent, because leaders know that future problems will be left to their successors, or that, if they are not, they can make another political decision in the future.
Which also sounds like a typical reactionary view on democracy’s failure with long-term planning. Point being that if politicians are only in power for a short time, they don’t have much time to accomplish long-term goals, rather they will be focused on PR relations so they can win re-election or their party can keep power. In the USA, this is particularly evident due to the 2-year cycle. The idea here is that you instead you had some emperor/king in place, they would be very interested in the long-term goals of the country as they and their descendants would be in power for decades to come. I guess I will be reviewing one of these books soon: A Critique of Democracy: A Guide for Neoreactionaries by Michael Anissimov seems like a reasonable choice.
Back to the book. Hanania particularly likes the analogy to anti-smoking campaigns and politics:
This is understood in other areas of policymaking. Imagine an anti-smoking group concerned that the tobacco industry exerts too much influence over our government. The most inefficient way to go about changing this would be to convince the individual lobbyists that they are wrong about the harms of smoking, and appealing to the consciences of politicians is likely to be only slightly more effective. Rather, a better strategy would be to seek to limit the influence of lobbying. One can do that through laws meant to either curb the power of the tobacco companies or to expose their efforts to the broader public. Similar logic applies for advocates of restraint. It is also worth considering cultural and social changes that different segments of society can take to ensure more transparent and ultimately better decision-making with regard to foreign policy. For example, when referring to former government officials who now work for corporations with an interest in larger military budgets, it is normal for reporters to mention the former government positions of such individuals, but not their current jobs in the private sector. This reflects an unstated and unexamined bias that their past experience in government is more relevant for understanding their behavior than how they currently make money. Because we are not sensitive enough to conflicts of interest in the area of national security, such practices continue to distort our ability to understand much of what is driving American foreign policy. Rather than arguing for the educated public to take an overly cynical view toward the field of national security, this chapter simply calls on us to apply the same standards that we use elsewhere when we worry about conflicts of interest and corruption in politics and government.
So in general, democracies are driven less much by the pure marketplace of ideas, and more by whoever happens to fund efforts to push public views on some topics. It is well known that the public generally is ignorant of politics, and tend to copy the views of their in-group elites, whether these are prominent media people, academics, or politicians. In other words, whatever views will prevail in the in-groups’ media will be the new norm for members of that group. Pick your favorite examples from the changing views of non-heterosexuals, gay marriage, pornography, Islam, protestors etc. Everything becomes left because left people always push these issues effectively.
One point Hanania misses with the smoking campaigns, were that these were in fact heavily influenced by irrelevant factors. Anti-smoking campaigns were delayed for many decades due to association with an enemy. Who? Nazi Germany was among the first countries to pursue serious anti-smoking campaigns, and by effect, this made the Anglos less likely to follow suit. One might even be tempted to say that Nazi association with anti-smoking campaigns killed more people in the USA than the war itself. Wikipedia says some 400k US deaths related to WW2. CDC says some 480k+ people die every year from smoking related mortality. OK, the comparison of deaths is not entirely fair (smokers die old so the years of life lost is lower), but you can see that delaying anti-smoking laws could probably accumulate 400k deaths over the course of like 7 decades since Nazis came into power until now.
Changing policy
The traditional model of international relations suggests that to change American foreign policy one must engage directly in the marketplace of ideas. The public choice perspective suggests focusing more on the incentive structures that lead certain ideas to be promulgated among the educated public and ultimately accepted by policymakers. Ideas are downstream from interests; politicians have long adopted beliefs about topics such as the best way to keep Americans safe and how winnable any particular war is based on electoral and bureaucratic pressures. Within the government bureaucracy that influences the president, the top military officials know that upon retirement they can be making money from those who directly profit from war. Journalists who want access to such figures generally have to reflect the worldview of powerful individuals in order to themselves get ahead. Individuals, not nations, are “self-interested.” The implication of this is that the best way to change policy is to shape the incentive structures that powerful individuals face.
For the last time in the book, I will return to tobacco lobbyists and the politicians they influence. If one is concerned that the tobacco industry has too much influence over policymaking, the most obvious solution to this problem is not to convince the lobbyists that they are wrong, or politicians that they should, out of the goodness of their own hearts, refuse donations from the industry and do what is right. While this kind of moral persuasion could work on some leaders, the bulk of activist time and resources should be spent on seeking to influence policy in other ways. More specifically, one should, through the use of the law and other means, try to reduce the influence of the industry and the lobbyists it hires, and create countervailing pressure on politicians that may incentivize them to take an anti-smoking stance.
This is more some basic advice for conservatives. While socialists constantly take over institutions and change laws in their own favor, conservatives and libertarians generally seem more interested in ‘playing fair’ with the result they always lose. If they want to win, they will have to start playing by the same strategy, or somehow legally prevent the enemy from doing so. I don’t really think conservatives will get wiser, so we will get more socialism.
Who directs American foreign policy? Where is the money from?
Steps to reduce the power of concentrated interests can be broken up into three categories: laws that directly reduce their power, laws that may reduce their power through greater awareness of their activities, and norms that, although they do not have the power of law, can change the incentive structures that powerful officials face. First, in the area of laws that directly reduce the influence of special interests, one simple reform would be to ban retired members of the military and national security bureaucracy from taking jobs with companies that have contracts with the Pentagon. If a lifetime ban is deemed too extreme a step, a prohibition for a number of years can have a positive effect, as the norm now is for many military officials to take lobbying jobs immediately upon retirement (Bender 2010). Other reforms would likely be incorporated into the larger movement to restrict the influence of corporations, a cause that has been hindered by Supreme Court decisions like Citizens United v. FEC.
It is perhaps the second category of reforms, that is, laws meant to increase transparency regarding the influence of special interests, that may have the larger impact. Right now, think tanks are not required to disclose the sources of their funding. We simply have no way to estimate what percentage of the op-eds that are written, the congressional testimony that is given, and the punditry we see on TV is funded by foreign states and government contractors. There is practically no theoretical limit to what the number is. Many organizations simply break down their donors by how much they give. For example, Brookings (2018) lists six donors that gave $2 million or above in 2018, with one of them being the Embassy of the State of Qatar. There is no upper limit to what the actual contribution from that country might be. Think tanks should provide more detailed reporting of their top donors, as they are as 501(c)(3) non-profits required to make public other financial information. Organizations that gather whatever data they can on the influence of concentrated interests, like the Foreign Influence Transparency Initiative at the Center for International Policy, should be supported (Freeman 2020).
Under current law, the Foreign Agents Registration Act (FARA) requires individuals who are lobbying on behalf of foreign states to report their activities to the federal government, with the information provided becoming public. Yet there is an exception for work of a scholarly nature (Laufer 2017). The practical effect of this is that if an American is paid money by a foreign despot to take out a billboard hyping the virtues of his regime, the law requires the public to know about it. But if the same dictator pays another individual to write a book doing the same, there is no reporting requirement. Thus, the law in effect considers paying for crude propaganda something that society must guard against, but takes a “see no evil” approach to sophisticated propaganda. At the very least, if foreign money is allowed to influence American politics, what form it takes should not be relevant to whether the people have a right to know the details.
One interesting thing that Hanania likes to highlight is that media stories about military matters are often based on an unhealthy combined interest from journalists and pro-war military staff. The military staff (deep state) will leak confidential information to the journalists, who then publish the BREAKING NEWS STORY. This gets the journalist what they want, attention, and the military what they want, their view of matters out in the public and changing the policies. A win-win strategy.
Regarding Hanania’s proposal about more openness with funding, I mean, this is sensible, but I think probably not practical. If Qatar wanted to fund some think tank promoting pro-Qatar policies, they could donate the money directly as they do now. Or they could set up some inscrutable set of business network that funnels the money eventually into the desired think tanks. Or they could get with the 21st century and pay money via Monero. Either way, this kind of funding openness law probably would not have much effect and would create more opportunity for powers that be to harass enemies using selective enforcement. Speaking of selective enforcement, that is exactly how the state deals with the leaking of information from deep state actors:
The public choice perspective should lead us to look at practically all existing practices, even those designed to hold government accountable, in a new light. Laws and regulations that seem to protect civil liberties may have unintended consequences once we think carefully about how ideas spread. For example, in New York Times Co. v. United States, the Supreme Court ruled that the government cannot restrict the publication of classified material, although it could arrest individuals for publishing that material after the fact. Today, the top newspapers in the country routinely publish classified information, and although they could theoretically be prosecuted for it, as a norm they almost never are. While this seems to be a victory for civil liberties, it actually advantages reporters who are close to government sources. According to the law, any member of the public has the right to publish information without prior restraint from the government, but that right can in actuality only be practiced by a select few. While top civilian and military officials regularly leak classified information that ends up on the front pages of major newspapers, those who leak classified information that the government truly does not want to get out like Edward Snowden are hunted down as criminals.
Perceived vs. actual expertise
In addition to taking leaders’ explanations for political decisions at face value, there is also a tendency to give too much deference to expertise within the military and national security bureaucracy. Often, organs of the government will through methods such as public interviews, official reports, and congressional testimony engage in forecasting with regard to the likely outcomes of various policy options. Through less formal channels, when the president is considering some policy or another, anonymous officials within the bureaucracy will often provide journalists with predictions about what will happen as a result. Yet a look at the history of the American intelligence community reveals that it would be beyond generous to say that it has been right on major questions as often as it has been wrong (Weiner 2008), and this view is confirmed by Tetlock’s (2017) groundbreaking work showing experts in a field to not be any better than informed lay observers at forecasting. Thus, there appears to be little justification to take the predictions of government officials seriously when they issue grave warnings about the consequences of retreat from a certain region of the world or area of conflict. A world where we did away with any deference to experts in areas in which expertise has not been independently validated is probably too much to ask for, although it is good to have as an ideal. Tetlock’s work does not suggest that there is no way to engage in smart forecasting, only that individuals conventionally described as experts are not necessarily the ones we should be listening to. Policymakers and the wider public should be more aware of this point, and also keep in mind that there is often a clear interest on the part of government officials to hype up threats or exaggerate the costs of a more restrained posture.
Expertise is an interesting topic. Who knows best about some issue? Generally speaking, people favor the views of people with credentials. Credentials in physics certainty provides accurate information about expertise in physics. The same is true for mathematics. But how generally does this pattern hold? Tetlock et al’s results show that it clearly does not hold with regards to many areas of social science, where off-field people do just as well, or better, than on-field people. This has to be one of the most interesting findings in social science, and has broad implications for epistemic heuristics.
Sanctions do/don’t work
One of the interesting chapters of the book concerns the use of economic sanctions. These are very widely used against nations that America dislikes. Wikipedia provides a map, and the government has a website with details.
Apparently, actual academic studies of sanctions show that they generally don’t work to change regimes, but they do work in causing economic disruption and killing people. If true, the result of this is that basically sanctions are a cheap way for the American empire to kill enemy civilians without the cost of having to sacrifice own soldiers or deal with unpopular media images of dead people from actual warfare. Even worse, America does this not because it hates enemy civilians, but just because this is a convenient strategy for politicians to improve their election chances. No grand strategy.
Let’s check out some of this evidence. First, sanctions can be quite extensive even pressuring third parties to avoid the targets:
Sometimes, a president will only sanction certain foreign individuals or businesses rather than enact more sweeping orders. For example, as of 2019 sanctions against Belarus targeted only individuals, such as President Aleksander Lukashenko and members of his staff, and a handful of businesses, rather than entire categories of economic transactions with that country (Belsat 2019).2 The toughest sanctions programs are applied to what are often referred to as “rogue states,” governments that are known for their violations of human rights and hostility to American foreign policy. Cuba is the only country in the world still sanctioned under the original Trading with the Enemy Act of 1917. American law bans practically all exports to Cuba, prevents Americans from doing business with that country, and even prohibits those who do business in Cuba from operating in the United States (Drain and Barry 2010). Likewise, Americans cannot legally do business with hundreds of Iranian individuals and corporations, and third parties that do not avoid dealing with the sanctioned entities can themselves face punishment (US Department of the Treasury 2018; Norton and Egan 2018).
Hanania cites some evidence on economic effects:
In order to estimate the economic effects of these kinds of policies, a 2015 study looked at 67 countries that were subject to American or UN sanctions between 1976 and 2012 and compared them with countries that did not face similar kinds of economic coercion (Neuenkirch and Neumeier 2015). UN sanctions were associated with a GDP reduction of an average of 2.2% a year, for an aggregate effect of more than 25% of GDP over a 10-year period. The effects of American sanctions were smaller but still significant, reportedly reducing GDP by about 1% a year, for a total effect of a 13.4% decline over seven years. The more stringent the sanctions regime, the greater the economic decline. As mentioned above, the GDP of Syria dropped by 75% between 2010 and 2015 (Gobat and Kostial 2016). Per capita income in Iraq likewise went from $3,510 in 1989 to $450 in 1996 (Sen 2003). The economy of Venezuela has been mismanaged for years, yet it did not completely collapse until after the election of Trump, when the United States began to place stringent sanctions on the Maduro government. From 2016 to 2019, GDP per capita income of that country dropped from $9,090 to a projected $2,550 (International Monetary Fund 2019). While all of these countries have suffered due to bad economic choices made by their governments, rarely have modern countries seen declines of this magnitude without the imposition of Western restrictions on trade. Besides perhaps war, it is difficult to think of a foreign policy tool that today causes more economic and humanitarian destruction than economic sanctions.
Neuenkirch and Neumeier 2015: The Impact of UN and US Economic Sanctions on GDP Growth:
Abstract: In this paper, we empirically assess how economic sanctions imposed by the UN and the US affect the target states’ GDP growth. Our sample includes 68 countries and covers the period 1976–2012. We find, first, that sanctions imposed by the UN have a statistically and economically significant influence on economic growth. On average, the imposition of UN sanctions decreases the target state’s real per capita GDP growth rate by 2.3–3.5 percentage points (pp). These adverse effects last for a period of 10 years. Comprehensive UN economic sanctions, that is, embargoes affecting nearly all economic activity, trigger a reduction in GDP growth by more than 5 pp. Second, the effect of US sanctions is much smaller and less distinct. The imposition of US sanctions decreases GDP growth in the target state over a period of 7 years and, on average, by 0.5–0.9 pp.
It passes the initial smell test of not asserting a crazy conclusion. It’s not very surprising that economic sanctions by the US empire would disrupt a country’s economy considering that international trade has a huge influence on the economy and that the US has been the largest economy in the last many decades. Anyway, their results. First, distribution of sanctions:
What happened in 2001? I guess a lot of sanctions were lifted in order to get countries in line with the anti-terrorism campaign following world trade center attacks. Trying to google this, I find CNN basically saying that was the reason US ended sanctions against India in 2001, and a story about the same for Pakistan. The UN sanctions seem to massively increase after 1991, which presumably related to the US led gulf war (Iraq war 1). In general, I would guess that most UN sanctions are driven by the US, so trying to distinguish both in a regression model may not be so sensitive from a causal perspective (both are US caused). Anyway, their main regression results:
The rows with sanctions is that of interest. Why are the sample sizes so small? Authors explain:
Figures 1a and 1b illustrate the frequency of sanctions and their severity over time. The overall number of country/year observations in which UN sanctions are in place (200; 9.3% of the observations) is much lower than that for US sanctions (618; 28.6%). Similarly, UN sanctions have been imposed against only 23 countries, whereas a total of 64 countries have at least one non‐zero observation for US sanctions. In addition, the US sanctions are on average harsher than those of the UN as 21.8% of US sanctions fall into category 3 (compared to 12% for the UN). These findings are not surprising, of course, since UN sanctions have to be enacted by the UNSC, which consists of five veto powers, whereas US sanctions only have to pass the US legislative. Also interesting is the huge increase in the frequency of UN sanctions after the end of the Cold War. The frequency of sanctions is highest during the 1990s due to the First Gulf War, the Yugoslav Wars, and several civil wars in Africa.
The authors are using a fixed-effects design, so one can only include countries that have been at some point sanctioned by the US or the UN (i.e., that vary in the predictor variable). Fixed effects models are good because they by design remove any effects that are (nearly) constant for a country over time, which will then include things like ethnic composition and geographical location. It would have been interesting to also see a cross-sectional regression with every country to boost the sample size to a more reasonable ~200.
Anyway, looking at their table, we see that it is not that convincing. Their 3 models show p values of < .05, < .05, and < .01 for UN and < .10 for US. So really, the last model is the only kind convincing result. I would have preferred to combine the sanctions to a single variable to boost statistical power. Similarly, the level of sanctions should have been used to create a single numerical variable of US+UN sanctions. Authors also look at the severity of sanctions in their own way:
The UN results check out, more severe sanctions have greater effect. The US results are nonsensical. I interpret this as being due to low statistical power. I don’t know why economists are so inclined to do these subgroup analyses with low power and interpret the noise. Why bother?
Anyway, the above model is also not good because it looks at individual years of economic growth. A single year’s economic growth is highly variably and somewhat random, so trying to predict this is difficult. Instead we might want to look at the numerical effects over time, say, 5 years of sanctions should have a bigger effect than 1 year and try to predict longer term economic growth. Their table 4 has this approach:
The results in the simple models (left) are convincing looking with every predictor < .01. OK, all in all, sanctions probably retard economic growth, and this study is not too bad.
On civilian deaths, Hanania quotes a variety of numbers:
Of course, economic sanctions do influence nutritional and health outcomes, and several efforts have been made to quantify these effects in terms of lives lost and other metrics. Economic sanctions tend to kill through harming the most vulnerable: pregnant women, newborns, the sick, and the elderly. The economic sanctions on Iraq in the 1990s were said to have killed more than 500,000 children by increasing the infant mortality rate (Zaidi and Fawzi 1995; ReliefWeb 1999). Recently, researchers have argued that these numbers were manipulated by Saddam’s government, yet even the “corrected” numbers still imply a death toll that may reach into the six figures (Dyson and Cetorelli 2017).3 A 1993 UN report indicated that sanctions on Haiti were killing 1,000 children a month (French 1993). One recent study estimated that 3,968 North Koreans died in 2018 alone because of the impact that sanctions had on delaying or preventing funding for UN programs aimed at addressing health problems and malnutrition (Korea Peace Now! 2019). As the report says, that analysis does not take into account the effects of sanctions on hindering other international aid programs or preventing the North Korean government from addressing its own domestic needs. In Iran, sanctions are estimated to have contributed to the deaths of 1,600 people from the H1N1 virus in fall 2019 alone (Jalapour 2020). Again, this number is based on one kind of death over a very limited period of time during which much of the entire Iranian healthcare sector was and remains in crisis. The effects of sanctions have certainly become worse in the era of COVID-19.
I didn’t check the sources, but it stands to reason that if you prevent already poor or kinda poor countries from importing medical supplies, make their exports less valuable etc., this results in civilians having worse healthcare which will kill a lot of people over time.
OK, but do sanctions help to change regimes then?
Supplementing this work, a more recent wave of scholarship reveals that sanctions can actually be counterproductive in achieving the goals of democratization and greater respect for human rights. Sanctions are associated with greater levels of repression, as leaders become increasingly desperate to hang on to power (Wood 2008). Compared to similar countries, states that are hit by sanctions are less likely to move toward democratization (Peksen and Drury 2010). Empirical research also suggests that sanctions may indirectly contribute to governments deciding to engage in mass killing. States would usually rather not commit atrocities against their domestic population, as doing so can bring international pressure on the regime, cost the government legitimacy in the eyes of the public, and spur disobedience among security forces sent to carry out orders (Krcmaric 2018; Koren 2017). Thus, states tend to kill large numbers of people only when they become desperate, which is why the existence of an ongoing civil war is perhaps the strongest predictor of mass killing discovered by scholars (Valentino, Huth, and Balch-Lindsay 2004; Valentino 2004; Uzonyi and Hanania 2017). Poorer countries are similarly more likely to engage in mass killing, since they often cannot afford to buy off their enemies or rely on more targeted forms of repression (Wayman and Tago 2010).
- Wood, Reed M. 2008. “‘A Hand upon the Throat of the Nation’: Economic Sanctions and State Repression, 1976–2001.” International Studies Quarterly 52(3): 489–513.
- Peksen, Dursun, and A. Cooper Drury. 2010. “Coercive or Corrosive: The Negative Impact of Economic Sanctions on Democracy.” International Interactions 36(3): 240–264.
I skimmed both articles, and everything looked fairly reasonable with p < .01’s everywhere. One should really look more into these studies and conduct a replication with more controls and larger datasets, using something like Bayesian model averaging. And yes, we use this p-value based way to assess papers because it works when we check replication efforts. Any paper claiming stuff that can’t produce p < .01’s across the board for their main results is very suspect.
Conclusion
I am no expert in American foreign policy, but according to the forecasting results, this is maybe a good thing. This book I feel like should have been shorter by maybe 20% as there were some boring parts. Still, at some 200 pages, it is reasonably short. While I wasn’t familiar with the body of academic research beforehand, I wouldn’t say I was convinced of much new by this book. I remember the Iraq and Afghanistan wars and the changing claims over time. These were evident in Denmark too, as we were involved in both wars. The revolving door issues were familiar too. Our minister of state (that’s the prime minister/president), Anders Fogh Rasmussen, pursued this warfare while he was leading the country (November 2001 to April 2009), and then immediately jumped ship to become the NATO general secretary (August 2009 to October 2014), and then after that immediately founded a lobbying company (Rasmussen Global; 2014-present). You know what this lobbying company published? That’s right, a book arguing that US needs to do more warfare/world police: The Will to Lead. This line of events perfectly into the topics of this book. My view is that Rasmussen is a traitor who got Denmark involved in pointless and expensive foreign wars for personal gain, though in a probably legal way. I am sure there are may such cases. The most obvious way to deal with this is as Hanania suggested: political leadership come with a lifetime ban on working for lobbying organizations. It’s almost an ancient Greek idea.
In a broader view, then, this book is essentially an attack on the deep state’s influence over American foreign affairs. America goes to many wars not because of beliefs in democracy etc., and not to gain oil, but because the Pentagon people and their media friends want to make money and can do so by influencing the US government to engage in warfare. The longer the better (new record is 19 years, 10 months, 3 weeks and 2 days). Everybody else loses though. Weirdly, academics and the media seem kinda uninterested in this issue, which one could interpret in far left Chomskian ways.
