On The Optimal Incentivization of Bureaucracy

Incentives matter. That is how I (and every other economics professor lucky enough to teach principles of economics) begin my classes every year.

Consider these sentences:

The central tasks of a modern state can be placed in three broad categories — maintaining the rule of law, providing public goods to citizens and using fiscal tools to redistribute income.

An efficient bureaucracy is essential for a successful state.

The efficacy of a bureaucracy is dependent on the incentives or disincentives that individual civil servants face when they take decisions.


Bureaucrats, policymakers, students of public policy and us professors should read this excellent paper. But if you are a student of economics, just beginning your journey in the Dark Arts, you should definitely read this paper.

Why? Three main reasons:

  1. It is a wonderful way to understand why incentives matter
  2. It helps you understand why government offices (and their cousins) function the way they do
  3. Trying to understand the Indian economy without understanding this aspect of it is an impossible task.

The paper is titled “Bureaucratic Indecision and Risk Aversion in India”, and is freely available here. The authors are Sneha P, Neha Sinha, Avanti Durani, Ayush Patel and Ashwin Varghese. It is an extremely accessible paper, in the sense that there are hardly any pre-requisites for you to read it in its entirety, and if you take notes, the paper is guaranteed to leave you with a ton of reading material. All those are the cherries on top: the biggest advantage is that you get a much better understanding of why the Indian bureaucracy is the way it is, and what could be done about it.

Here are my quick reflections after having read the paper:

  1. If you are a student reading this paper, you might enjoy encountering prospect theory at the outset. Again, it is a great way to connect the dots between what you learn in a classroom and its real life application. (There’s further reading, if you are so inclined)
  2. The second section is (to me) the meat of the paper. The title is “The Causes of Bureaucratic Risk Aversion”, and the authors list out 12 in all. These are grouped under three separate headings. As a mnemonic for myself, I think of them as Structure, Culture and Nurture
    1. Structure is Organizational Design: whether through over-monitoring, not enough discretion being given, an overload of bureaucratic responsibilities, the typical bureaucrat simply doesn’t get the time, the bandwidth or the incentive to not be risk averse. In plain simple English, the authors are saying this: if you or I had been a bureaucrat, we would have done exactly the same things that our bureaucrats have been doing for years. And this is so because like our bureaucrats, we too would have responded rationally to the operating structure we are a part of. It is not the bureaucrats that need to be changed, in other words, it is the organization design of bureaucracy that needs to change. Or that is my understanding of the first half of this section.
      But when it comes to structure, the latter half of this section speaks about the way candidates are selected, about how they are trained and mentored and about how they are (under)paid. Each of these are important to understand, and I especially liked how benchmarking and comparisons were made using examples from Singapore, Malaysia and Indonesia. Speaking of which, please read Gulzar Natarajan on entrepreneurship as a desirable trait in a bureaucrat, and please see this website from Malaysia.
    2. Culture is Institutional Norms and Culture: Please (pretty please!) read the whole section. It is an eye-opener. Plus, it is full of delightful nuggets. For example, I learnt by reading this section that the Official Secrets Act doesn’t define what a secret is, but forbids government employees from revealing them.
      The authors say that accountability to process, rather than outcomes is a problem. Now that gives me a delightful problem to mull over for a long time to come. How to reconcile this point (which is, I think, fair and valid) with this post (which also, I think, is fair and valid? Thoughts and suggestions welcome!
      Shleifer and Vishny (1993) and Becker and Stigler (1974) are two papers cited in the last part of this section, and as a student, they are absolute must-reads.
    3. Finally, Nurture is Political Pressure: This section is about your political boss (or bosses, in some unfortunate cases). Anybody from the corporate world will immediately draw the link between this and the dreaded “dotted line manager”. Similar problem, and similar outcomes.
  3. What can (and should) change is what this concluding section is about.
    1. I loved the idea of linking public sector salaries to private sector wage levels (although as statistician I can’t help but think about how that might actually work in India)
    2. The Committees timeline is wonderful for students, in the sense that gives you a quick way to understand what has happened in this space since independence, and the enthu-cutlets can dramatically expand their “To Read” list.

As additional reading, should you have the appetite for it, here is what I would recommend:

  1. All posts tagged “Bureaucracy” from Gulzar Natarajan’s excellent, excellent blog.
  2. An old, tangential but delightful read: English August
  3. Ch06 of this year’s Economic Survey

India and her cities

In the previous week, Livemint published an excellent article, titled “Why India has the fastest-growing cities“. Today’s post is a rumination on that article, and associated thoughts.

Urbanization, I unequivocally hold, is an good thing. This belief has come about as a consequence of learning development economics over many years. It has also come about because I have had the opportunity to read many great books about the topic, of which I think Ed Glaeser’s ‘The Triumph of the City‘ is by far the best one.

The reason I like that book so much is because it is an unapologetic paean to urbanization. It not just defends urbanization, it actively reveres it. And there is something to be said for that argument. Cities, when designed well, are worth revering! Watch this lovely TED talk by Jeff Speck, for example. The talk is ostensibly about how to make cities more walkable, but it covers a lot more ground than just that.

Here’s the most important reason, I think, that cities ought to be revered. It is because of their most important feature, and their most appropriate definition: they are labor markets, first and foremost.

What are cities?

Cities are simply a lot of people packed into a relatively tight space, most whom happen to be open to new job opportunities. That’s a paraphrased definition, and it is certainly not mine. The best way to truly understand what this means in practice is to read a lovely (but by Indian standards, prohibitively expensive) book by Alain Bertaud, called Order Without Design.

As I said, the book is expensive, but I can recommend three freely available online resources that you might want to read, listen and watch instead.

When I quit my job in the analytics industry in 2009, it was because I wanted to switch over to academia. Switching over to academia meant that I had to come to Pune. Now, Pune is my hometown, and I love it to bits, but the reason I had to come to Pune is because there were many more jobs in academia in this city than any other city in India.

Conversely, if you are looking to set up a college, a student exchange program or a university, Pune is the best place to do so, precisely because of the paragraph that precedes this one.

That’s what Alain Bertaud means when he says this:

“Sometimes when I read the papers of my fellow urban planners, I get the sense that they think cities are Disneyland or Club Med. Cities are labor markets. People go to cities to find a good job. Firms move to cities, which are expensive, because they are more likely to find the staff and specialists that they need. If a city’s attractive, that’s a bonus. But basically, they come to get a job.”

“All the jobs are in the cities” is a phrase that you will hear often enough in India, but reading Alain Bertaud’s book helps you understand that the statement is actually tautological.

But, if you think about it, and to the extent that you agree with what is written above, we’re committing a moral crime by not glorifying urbanization. Strong words? Maybe. But, I would argue, true words as well.

Urbanization in India

But then how come urbanization in India is only 31%? If all the jobs are in the cities, and people in India are crying out for jobs, why aren’t they all moving to India’s cities?

There are three responses to that.

First: we make it difficult, expensive and to begin with, potentially unremunerative for people to migrate to India’s cities. Difficult because of a whole host of laws and regulations that hamper and hinder the development of efficient urban labor markets. Expensive because of poor urban planning which means housing and transport are not cheap for first generation immigrants into India’s cities. And potentially unremunerative because a lot of our welfare schemes require their targeted beneficiaries to be citizens of rural, rather than urban India.

Second, they are too moving away from villages! Hop into an Uber in your city, and take the time out to speak to your driver. More likely than not, your driver is likely to have the following characteristics. He will be a he, he will have a parcel of land back home in his native village, and he’ll have come to the city in search of a job. That he is a he is an indictment of our culture and our labor market. That he has a parcel of land back home is an indictment of our lack of reforms when it comes to land. And the fact that he is working as an Uber driver (services) rather than in a factory (manufacturing) is an indictment of our lack of reforms when it comes to labor and land market laws. But, to loop back to the start of this paragraph, people are certainly leaving India’s villages.

Third, in spite of it being difficult, expensive and potentially unremunerative, they are migrating, but to areas just outside our country’s cities. And therein lies a trifecta of tragedies: of policy design, of incorrect measurement and therefore of a poor urban experience.

The Livemint article…


… has been written by three people: Kadambari Shah, Vaidehi Tandel and Harshita Agarwal. All three of them work at the excellent IDFC Institute, located in Mumbai. One reason I use the word excellent that is relevant to today’s blog post is the fact they produced a very interesting report, a somewhat abridged version of which is this article, that came out in Livemint a while ago.

In that article, they gave us India’s original definition of urbanization, as it was defined in the year 1961.

“India’s three-tiered census definition of ‘urban’—at least 5,000 inhabitants, density of 400 people per sq. km or more, and at least 75% of male working population engaged in non-farm activities—was first framed in 1961 by then census commissioner Asok Mitra.”

By this three-tier definition of urbanization, we’re at 31%. That is, roughly one-third of our population is urbanized, and the remaining is not.

But does that mean that the remaining is rural (and somehow agrarian)?


Because of what we discussed above, in the section “Urbanization in India”, folks migrate out of villages, but not to India’s cities. They migrate to areas just outside of India’s cities: the so-called satellite towns.

So what’s the big deal?

Well, if you stay out of the local municipal corporation’s limit, it is not obligated to provide you the following services: “town planning, slum improvement, public amenities including street lighting, parking lots, bus stops, solid waste management, building regulations and fire services.”

Sure, of course not, you might think. It won’t be, for example, the Pune Municipal Corporation’s job, but that of the satellite town’s corporation. Ah, but because it is a town (a census town, to use the government’s definition), it will not be covered under the definition of an Urban Local Body (ULB). It will, instead, be a Regional Local Body (RLB).

And the RLB doesn’t need to provide (you might say cannot provide, given financial and other capacity constraints) the services mentioned above.

So What?


  • Urbanization is good, even great
  • We don’t have enough of it in India
    • Because we make it difficult for people to move
    • When they do move, we make it difficult for them to find jobs
    • They still move, but we don’t measure the movement well enough
  • We don’t measure it accurately enough because our approach to the measurement is wrong, and woefully out of date
  • As a consequence, when (and if at all) folks attempt to urbanize, they don’t get the kind of urban amenities that they so desperately need.
  • All of this is assuming, of course, that urban amenities are provided, and ably so, by municipal corporations – but a blog post should only be so long, hey?

On the first Monday of March, we’ll come back to the topic of urbanization and India once again.