The "Meet Our Researchers" series showcases the incredible scholars at Stanford’s Center on Democracy, Development and the Rule of Law (CDDRL). Through engaging interviews conducted by our undergraduate research assistants, we explore the journeys, passions, and insights of CDDRL’s faculty and researchers.

What role do middle or “auxiliary” powers — like India, Brazil, or Turkey — play in this evolving landscape of great power competition?

This is one of the biggest differences between today and the Cold War. Back then, the system was much more binary. Today, the world is more fragmented. I think of it as a race: the U.S. is ahead, China is closing the gap, and everyone else is running behind. But they’re running. They have agency. They’re not just sitting on the sidelines.

Countries like India, South Africa, Turkey, and Brazil are swing states. They’re not going to line up neatly with Washington or Beijing. BRICS is a perfect example — democracies and autocracies working in the same grouping. The U.S. has to get used to living with that uncertainty. We need to engage, not withdraw.

And at the same time, while the U.S. seems to be retreating from some of its instruments of influence, China appears to be expanding. What worries you about this divergence?

It’s striking. We’re cutting back on USAID, pulling out of multilateral institutions, shutting down things like Voice of America, Radio Free Asia, Radio Free Europe, and cutting back on diplomats. Meanwhile, the Chinese are expanding their presence, their multilateral influence, their media footprint, and their diplomacy.

If the autocrats are organized, the democrats have to be organized too. We can’t just step back and assume things will turn out fine. That’s not how competition works.
 

During the Cold War, despite intense rivalry, the U.S. and USSR cooperated on nuclear nonproliferation and arms control. How do you see cooperation taking shape in today’s U.S.–China rivalry?

That’s a really important point. Cooperation in the Cold War wasn’t just about deterring the Soviets — it was also about working with them when we had overlapping interests. The Nuclear Nonproliferation Treaty of 1968 was a monumental achievement. It was signed at the height of the Vietnam War, while we were literally fighting proxy conflicts, and yet we found common ground on nuclear weapons.

I think something similar can and should happen now. Even if we’re competing with China, and even with Russia, there are areas where cooperation is in everyone’s interest: nuclear arms control, nonproliferation of dangerous technologies like AI and bioweapons, and climate change. These are existential issues. We cooperated with our adversaries in the past; we should be able to do it again.

One of the big debates in international relations is about the role of ideology. How much does ideology matter in this current geopolitical context?

It matters a lot. My book isn’t called Great Powers — it’s called Autocrats vs. Democrats for a reason. I believe ideas and regime type shape international politics.

Putinism and Xi Jinping Thought are exported differently. Putinism — illiberal nationalism — has ideological allies in Europe and here in the U.S. Xi’s model is more economically attractive to parts of the Global South. Power matters, of course, but it’s not the only thing.

You can see this clearly if you compare Obama and Trump. There was no big structural power shift between 2016 and 2017, but their worldviews were radically different. That’s evidence that ideas and individuals matter a great deal in shaping foreign policy.
 

You’ve warned about the dangers of U.S. retrenchment. Are there historical moments that you see as parallels to today?

I worry about a repeat of the 1930s. When Italy invaded Ethiopia, Americans said, “Where’s Ethiopia?” When Japan invaded China, they said, “Why do we care?” Then came 1939. Stalin and Hitler invaded Poland, and we still said, “That’s not our problem.” Eventually, it became our problem.

If we disengage now, we may find ourselves facing similar consequences. That’s part of why I wrote this book — to push back against the idea that retrenchment is safe. It’s not.

To close, what advice would you give to students who want to build careers like yours? And, could you recommend a book or two for young people entering this field?

Be more intentional than I was. Focus on what you want to do, not just what you want to be. Develop your ideas first, then go into government or academia to act on them. Don’t go into public service just for a title. I saw too many people in government who were there just to “be” something, without a clear agenda. The “to do” should come first; the “to be” comes later.

As for books, my own book, Autocrats vs. Democrats: China, Russia, America, and the New Global Disorder, is coming out soon — you can pre-order it. But the two books that shaped me the most when I was young are Crane Brinton’s The Anatomy of Revolution and Guillermo O’Donnell and Philippe Schmitter’s Transitions from Authoritarian Rule.

Motivation and Summary

Modern states depend upon bureaucracies effectively delivering services and enforcing regulations, from public health to environmental protection and postal services. When bureaucracies are plagued by inefficiency and incapacity and are unable to “deliver the goods,” elected representatives suffer. This can open the door for citizens to throw their support behind authoritarian leaders who promise to deliver more effectively.

A central question of bureaucratic design concerns the degree of autonomy that ‘principals’ — political actors like legislators — should delegate to bureaucratic ‘agents.’ Calibrating autonomy is crucial because principals cannot perfectly monitor or control agents, whose preferences often diverge from their own. On the one hand, highly autonomous bureaucrats may become unaccountable to their principals. Think of national security agencies that create secretive dissident watchlists or even authorize assassinations. On the other hand, bureaucracies with little autonomy will find their decision-making hamstrung by “red tape.” If public health agencies required extensive legislative approval for every aspect of vaccine delivery, infection rates would skyrocket. How, then, can bureaucracies be designed to achieve their goals, creatively respond to new problems, and minimize corruption, all while being perceived as legitimate?

In “Calibrating autonomy,” Katherine Bersch and Francis Fukuyama disaggregate the concept of autonomy into two facets (independence and discretion), present hypotheses concerning how each facet relates to capacity and quality, and then test these hypotheses, primarily using survey data from Brazilian bureaucrats. Their findings caution against overly blunt efforts to calibrate autonomy across multiple bureaucracies, suggesting instead that policymakers should understand a given agency’s capacity to perform.

Argument and Hypotheses

The article builds on earlier research by co-author Francis Fukuyama, who in 2013 hypothesized a key moderating role for bureaucratic capacity in the relationship between autonomy and quality. Briefly, low-capacity agencies — lacking expertise and resources — will likely struggle to utilize their autonomy and deliver, whereas high-capacity bureaucracies will deliver effectively for opposite reasons. However, autonomy is itself a complex concept that ought to be disentangled before one can begin to spell out how it is linked to delivery.

Independence, the first “face” of autonomy, concerns the degree to which bureaucrats are constrained by actors who are close to politics, such as elected leaders or politically-appointed agency heads. More independent bureaucrats might allocate waste management contracts on the basis of cost-effectiveness or service quality records. Less independent bureaucrats might allocate according to the whims of politicians who wish to reward their allies. The degree of independence is oftentimes a function of the number and extent of political appointees in a given agency. 

The second face of autonomy is discretion, or rather, the extent to which agencies are constrained by laws, rules, or protocols. Waste management agencies with limited discretion may struggle to respond to sporadic garbage pile-ups, whereas those with high discretion may set unpredictable collection schedules. 

These two faces of autonomy lead to different expectations about how they are connected to quality. The authors hypothesize that independence and quality stand in a linear and positive relationship: agencies that are more unconstrained by political actors will deliver more effectively. By contrast, discretion and quality stand in a non-linear or “Goldilocks” relationship: too few and too many constraining rules will reduce quality. Following Fukuyama’s earlier research, capacity moderates these relationships; for example, less independence in specifically low-capacity agencies may strengthen quality, perhaps in cases where legislative actors are able to appoint highly qualified experts.

Methods and Findings

The authors select Brazil as their case, in part because it exhibits considerable internal variation in bureaucratic quality: some agencies are professional while others are incompetent. By focusing on one country, they avoid comparing fundamentally different national bureaucratic contexts. Further, selecting Brazil — with both strong and weak bureaucracies — reduces ‘selection biases,’ i.e., mistaking characteristics of only high- or low-quality agencies as causes of overall performance. From a global standpoint, Brazilian bureaucracies are solidly middle-ranged, neither outstanding nor abysmal. 

The main empirical part of the paper involves a 2018 survey of over 3200 Brazilian federal bureaucrats. About 60% of the respondents are political appointees and 40% are civil servants, which enables looking at bureaucrats who stand in different relationships to political actors. The authors exclude military personnel, teachers, nurses, and local police officers.

As hypothesized, respondents’ perceptions of agency independence increase their perceptions of quality in a linear way. This finding likely reflects Brazil’s coalitional style of government: politicians benefit electorally from strong government performance; without bureaucratic independence, political pressure and influence from the many coalition partners would likely hinder bureaucrats and weaken performance. Meanwhile, perceptions of discretion align with the authors’ non-linear expectations.
 

Image

Figure 3: (a) Impact of independence on quality at varying levels of capacity. (b) Impact of discretion on quality at varying levels of capacity.
 

These findings are mediated by capacity, which the authors measure in terms of resources, career length, salaries, and the proportion of agents in core or expert careers. The findings hold even when controlling for political factors like party membership or appointment type (political vs. civil servant), individual characteristics such as gender or years of service, and agency differences, including budget or size. These controls are based on administrative data from over 326,000 bureaucrats across Brazil’s 95 most important federal agencies.
 

Image

Figure 4: Quality at varying levels of capacity.
 

Contributions and Implications

“Calibrating autonomy” makes important conceptual and empirical contributions to our understanding of bureaucratic performance. By disaggregating the concept of autonomy into independence and discretion, it helps make sense of seemingly contradictory empirical findings, namely that both minimally and highly autonomous bureaucracies perform well. And by evaluating their hypotheses about the two faces of autonomy using Brazilian data, the authors guard against selection biases or problems from comparing countries with quite different bureaucratic landscapes.

The paper serves as a caution against overly simplistic or blunt “solutions” to poor bureaucratic performance. Merely limiting discretion or increasing legislative oversight can make matters worse if the relevant bureaucratic culture is not properly understood, especially when it comes to capacity.

*Research-in-Brief prepared by Adam Fefer.