In democracies, why are debates among political elites far more consensual on national security issues than on other subjects? What is the significance of the fact that, precisely in democracies, state bureaucracies concerned with security issues have continued to grow in size and in scope? And what are the potential risks of these phenomena for democracy itself? These are some of the questions addressed in the research project Elite Ideological Consensus and National Security: A Cross-national, Over-time Study Using a New Methodology which will be supported by the Swiss National Science Foundation for three years at a cost of some 680,000 CHF. David Sylvan, project leader and professor in the Institute’s International Relations/Political Science Department, explains the project.

Why did you decide to work on the transformation of democracies and security issues?

The possible transformation of democracies into totalitarian regimes is a subject that was widely written about in the 1940s in both scholarly works and in novels. In political science, Harold Lasswell – one of the most influential social scientists of the twentieth century – himself focused on the subject in a widely cited article that was both theoretical and empirical. Lasswell, who among other research topics studied elites, war-time propaganda, and the ideological reasoning of Soviet leaders, published a piece at the start of 1941 (thus, almost a year before the United States entered World War II) titled “The Garrison State”. The article was built on work Lasswell had done earlier about Japan’s transformation into a regime where the military played an increasingly important role; in the 1941 paper, Lasswell addressed the question of whether Japan’s transformation was a harbinger of similar changes in other democratic states, such as Great Britain or the United States. Although he didn’t answer that question, he did describe a set of troubling structural trends.

Did the Snowden affair influence your coming up with the research proposal?

Not directly, because I had been thinking about the topic for a number of years. In fact, a few years ago, I finished a book on US foreign policy and was increasingly struck by the enormous size of the American national security bureaucracy. The real takeoff point for that growth was the passage of the National Security Act in 1947, which created a unified Department of Defense, a National Security Council to advise the president and, last but not least, for the first time in peacetime, a civilian intelligence service, the Central Intelligence Agency (CIA), which reported directly to the president. A few years later, in the wake of the outbreak of the Korean War, the United States massively increased its national security spending, which in turn rapidly became institutionalised, so much so that in Eisenhower’s farewell address as president, he could refer to the “military-industrial complex”. Arguably, these changes and others (such as setting up the National Security Agency [NSA] in 1952) marked a real break with previous US policies, at least with respect to geographical scope and budgetary size. Up to the 1930s, the United States had been a regional power with a relatively limited military. Two decades later, it had a peacetime army of millions of men, as well as non-wartime alliances with countries in Europe and Asia. This shift, which some scholars later referred to as the establishment of a “national security state”, already had important potential implications for the persistence of democracy. For example, because the CIA and the NSA had operations which were not supposed to be made public, their budgets were not labelled as such but instead hidden in various accounts of the Defense Department, something which was at the least a violation of the spirit, if not of the actual text, of the US Constitution. Now, the interesting thing is that these various changes – security organisations, lavishly funded, secret budgets – were never overturned, up to and including the present day. At the time, of course, the justification was the Cold War. For a brief moment after the fall of the Berlin Wall, people began talking about a peace dividend. But after the First Gulf War, and even before 9/11, things quickly returned to the way they had been during the Cold War. In fact, there’s a sort of ratchet effect: every terrorist attack is cited as evidence that the world is dangerous and that security agencies must not only have their budgets increased but the scope of their powers as well, for example, as regards surveillance. And this is true not only for the United States but also in Europe: France just passed a new law similar to the Patriot Act voted after 9/11 in the United States; last summer, Britain did the same, with only three days of debate. Even here in Switzerland, there is fairly broad agreement among political parties that this is not the time to cut back on the resources or the scope of the security agencies.

In other words, we’re reproducing garrison states at the start of the twenty-first century.

Let’s say that Lasswell’s idea is in sync with current events because long-term processes of growth and institutionalisation of national security-related programmes are justified both by pointing to threats and by providing themselves a whole array of specialists whose job is to look for such threats. As security agencies grow in size and scope, the secrecy of their operations and their funding mechanisms, as well as their tendency to monitor all sorts of communications and transactions, are potential threats not only to civil liberties but also to what Lasswell was concerned with, namely democracy. If we add to this the fact that on any number of other issues – say energy or transportation policy – parliamentary debate is vigorous, with sharply opposed viewpoints being put forward, whereas on national security issues, debate is much more circumscribed, to the extent it takes place at all, the implications for democracy are even more striking. It’s for these reasons that I pose two questions. The first is whether in fact what Lasswell describes is a general phenomenon applicable to all democracies, or whether instead it’s particularly American, spawned by the Cold War and not really true of other democracies, particularly those that weren’t active in the Cold War.

And the second question?

It’s how, if there is indeed a general phenomenon, to explain it. Is it simply due to a sense of threat? Or are there other factors at play? There are at least two other such factors that seem plausible. One is the role of peer pressure in multilateral military alliances. We know that each member of such alliances faces strong pressure from other members to spend money on security-related agencies. In NATO, for example, the United States has for decades argued over and over that the countries with low levels of spending needed to buy planes, tanks, and other equipment; we have some evidence that the same pressures are at work on intelligence matters. The second factor isn’t really an alternative to the first but instead complementary. There’s an important scholarly literature on the way in which states with overseas colonial empires created specialised police and intelligence agencies to keep track of actual or potential independence or insurgent movements. This is true of France, Britain, Portugal, and other colonial powers (and also of the United States, in its colony of the Philippines). But what happened after decolonisation? Those agencies weren’t entirely, or even largely, dismantled; instead, they were reconverted into security-related tasks in the metropolis. Thus, if there is a general tendency for democracies to convert themselves into garrison states, the explanation may not be solely a matter of organisational inertia connected to an ideology of security against potential threats, but also due to allied pressures and/or to the persistence of former colonial structures.

I imagine that answering these questions won’t be simple from a methodological point of view.

This is exactly what I’ve been trying to think through for several years. It’s easy to imagine a general phenomenon of the garrison state, but it’s very hard to measure it in a systematic and cross-national way. For two years, I taught a seminar at the Institute on the garrison state, and I was lucky enough to have some very bright students try out different ideas about how to study this or that aspect of the garrison state. But the problem they ran up against was always the same. Because security organisations are secret, it’s not possible to know anything about their budget. In political science, the first thing we do when studying organisations is to look at their budget. Here we can’t do that, nor can we look at the number of employees they have or at their operations (e.g., spying, surveillance). So instead of doing this, the idea behind the project is to look for indirect measures: the traces, so to speak, of agencies that have become institutionalised. Here’s the reasoning: if security agencies keep getting funded year after year, this means that there is parliamentary approval, year after year, of their undisclosed budgets, whether or not there’s a particularly acute threat in a given year. And while of course the actual budgets are hidden among all kinds of accounts, those positive votes should be signalled by consensus about the national security issues dealt with by the agencies.

What do you mean by that?

The hypothesis is that the continued existence of security organisations, not to mention their long-term growth, is only possible because political elites, whether in parliament, in the executive or in the mass media, see the world in roughly the same way. Those elites may strongly disagree with each other on other subjects, such as education policy, taxes or the environment, but they’re united on the idea that the world is dangerous and that it’s out of the question to try to economise on the agencies set up to deal with those dangers. If Lasswell was right, there should be a consensus on national security issues though not on other issues voted on by parliaments; arguably, that consensus should grow over time as formerly dissident members of parliament retire, are voted out of office, or even (rarely!) change their mind. In this way, we can use the split between consensus on national security issues and dissensus on other ones, as well as an increase in time for the first of these, as a kind of indirect measure of the garrison state or, to be precise, of its traces. It’s a sort of thermometer, if you will, that lets us study a phenomenon whose details are secret.

How will you study this consensus?

The idea is to look at parliamentary debates and see if the way of reasoning is the same even for those who end up voting on opposite sides. For that, I’ve come up with a way of pulling out from parliamentary speeches the reasoning used in those speeches – how does a speaker connect point A to point B? – and then, as a way of checking on consensus, of seeing whether that reasoning is the same across speakers, especially those who, as I said, vote differently. Of course, given the hypothesis, this means that we have to make comparisons both across issue areas and over time. So the idea is to measure consensus for three types of debates: those on national security issues, those on non-security issues that involve foreigners (for example, development assistance or immigration), and those on domestic political issues (for example, the welfare state or agriculture). Similarly, we’ll look at three time periods: early in the Cold War, late in the Cold War, and post-9/11. Where things get complicated is when you add in multiple countries: since there are two other factors (belonging to a multilateral alliance and having had an overseas colonial empire), we get four combinations of countries. To be safe, we should have two countries in each category; so, we end up with eight countries times three time periods times three types of debates. That’s 72 cases, and since each debate involves a minimum of 20–30 speeches all told, you can begin to see why it’ll take us three years and why there will be three researchers (including a native Japanese speaker and a native German speaker) in addition to me.

Let me come back to Edward Snowden. Presumably you’re planning on carrying out all this work not only because of a scholarly interest but because of your concerns as a citizen.

What really shocked people in the whole Snowden affair, beyond the obvious issues of the NSA’s technical capabilities, was the fact that almost no one had any idea of the sheer scope of its activities: it not only could listen in on phone calls and read emails all over the world, but in doing so, it collaborated with most other countries’ – including, of course, most other democracies’ – own national security organisations. Of course, we were immediately told that most members of parliament and indeed most ministers were unaware of this collaboration. That may well be true, but it doesn’t take away from the fact that if they had wanted to know, they had all sorts of means at their disposal to do so. In effect, there was a deep and largely tacit consensus just to go along and not to ask questions. It’s that consensus I want to look at, and in this sense, not only does the research project have theoretical and scholarly implications, but it also offers a possibility, because all of the research findings will be publicly available, for any citizen of the countries we’re studying to take our results and use them to further, or perhaps open for the first time, debate on these issues. In the largest sense, then, the project offers the possibility of recovering a neglected portion of our recent history.