I came to the Bank because I thought that research here could have far more impact than research anywhere else. What makes research here unique is easy access to the resources of the entire World Bank Group. Since I arrived, people who work in the various practices and in the regions have expressed an eagerness about pursuing all the possibilities. People in units like ECR and Treasury were also excited to hear about projects where they could help. My initial conversations with ITS have already yielded progress toward easing access to the resources that anyone working in DEC might want.

The mood in DEC

The mood is DEC is different. Many people have told me about the frustration and disappointment they have experienced while working in DEC. The unifying complaint is that they do not get the recognition they deserve.

In a unit with technical professionals, disappointment and hurt can easily evolve into disengaged cynicism. The risk is that this feeling shows up in interaction with those outside the unit, and this can generate more frustration and disappointment.

Imagine that the figure above shows the utility possibility frontier, with utility for people who work in DEC on the horizontal axis and utility for everyone else in the Bank on the vertical axis. If those of us in DEC devote our energy to bemoaning the current position A and arguing that we should be at a point like B that is better for us, we could miss chance to move to a point like C that is better for everyone.

The corrosive effect of cynicism on trust

Before I arrived, DEC sent a memo (/assets/fellowsProgram.pdf) to the Steering Group on Staffing Plans. Members included Senior Vice President for Operations Kyle Peters, former Vice President for Budget, Pedro Alba, Vice President and Controller, Bernard Lauwers, and Vice President of Human Relations, Sean McGrath.

To free up resources for new initiatives, the Bank as a whole had been cutting positions to shift resources from lines of work that are now considered less important. DEC was perceived as being unwilling to consider any such realignments and had not cut any positions. In fact, the DEC memo on staffing proposed a program that would add 90 new staff positions to be filled by people doing work previously done by short-term consultants. Note to outsiders: The main difference between the cost of staff and the cost of consultants (STC’s) arises because staff receive benefits. Take-home pay of $100 for an STC costs $100. Take-home pay of $100 for a staff member costs $100 * (1+x). The benefits charge x used to be 50%. It has increased to 70%.

The attitude expressed by people inside DEC was that

… the ‘resource neutrality’ of the new program will be ensured by the fact that the new positions would only be filled if there are adequate funds.

The problem with this approach is that it depends on a nonstandard and potentially misleading interpretation of “resource neutrality.”

To the authors of this memo, responding to the Steering Group meant doing some play acting. They seemed to think that DEC should not face any limits on positions. To get approval for the positions they wanted, the memo made opaque, misleading arguments to people who have better things to do than sort through make believe.

People in DEC are frustrated that they do not have more money and seem puzzled that their repeated requests for more have not answered. To my knowledge, what no on in DEC has asked is whether its refusal to engage honestly with processes that are designed to encourage the efficient use of resources might be part of the reason why its requests for more money are treated with skepticism.

Many hands touched the memo that DEC submitted, including people from other units such as HR. But in the end, it came out as a memo from the Chief Economist, so we have to own it.

In describing the budget impact of the proposal to hire more staff and reduce spending on consultants, the meaning of budget neutrality morphs from the standard usage to the new one that DEC had in mind: Note for outsiders: The key to this passage is the difference between resources that are part of Bank Budget and those that come from trust funds. True to its name, the Bank borrows at a low rate and lends at a higher rate. The gross income on its loan portfolio is allocated under the Bank Budget to cover the Bank’s expenses, including its spending on research and data collection in DEC. The Bank also accepts donations into trust funds that support these same activities. The point here is that the true statement “Hiring more staff will not cause a deficit in our bank budget because we will pay for them with money from trust funds” morphs into the false statement “Hiring more staff is budget neutral in the sense that it will not increase total spending.”

By the end of page 5, an unwary reader could conclude that a switch from consultants to staff with benefits will cause no net increase in spending.

I refused to back this plan. I couldn’t tell where the bodies were buried, but its arguments didn’t pass the most basic plausibility test.

When I told people in DEC that the rest of the organization would not trust us if we tried to push this request forward, the type of response I received was, “What is the evidence that no one will trust us?”

When I tried to get people in DEC to explain the numbers to me in detail, what I got back were tendentious and evasive responses. One conversation closed with a condescending smile and a recommendation that “because you are so busy, let us work this out with someone else in the front office.” Note for outsiders: The original memo proposing this plan went out under the name of my predecessor, but it is quite possible that he accepted an offer like this and did not realize what it said.If you think that these kinds of conversations have no effect on trust, ask yourself why I won’t let anyone send out email from my personal account.

I eventually worked out what the calculation in the memo shows. (See my handwritten notes on the version of the memo you can download from /assets/fellowsProgram.pdf): Note for outsiders: The table in the memo cites a cost of $50,000 for staff. This is roughly equal to take-home pay of $35,000 when the benefits charge is 50%, which is what the authors of the memo mistakenly assumed. Correcting for this mistake makes the proposal for shifting from consultants to staff look even more expensive.

The memo notes that research assistants in Ph.D. programs work for salaries this low. It suggests that to match the graduate school experience, these staff will only work 85% time, but the comparison to the consultants they replace assumes that they work full-time.

So in short, the calculation shows that if DEC can hire research assistants with a BA or MA degree, pay them less than clerical staff with less education, and have them work full-time doing the work that consultants once did, the cost will be about the same. This is not what most readers would expect from the sentence:

The current perception of STC as cheaper than Term staff has to be questioned.

Getting past denial

I accept the reality of the bad feelings people in DEC have experienced and have no view about who ultimately is to blame. All I can say is that as someone who was recently treated like an outsider by people in DEC, I can assure you that the feeling of disengaged cynicism comes shining through.

Speaking now as an insider, we cannot deny the damage that our cynicism and disengagement do. Moreover, this is not a subsidiary issue. This represents an existential threat to DEC. There are only two possibilities:

If people in the Bank cannot believe everything DEC writes, they can’t believe anything we write. We need to take control of our fate.

Empathy and Trust

To work effectively with others, we must act now to

Empathy gives us insight into the professional obligations and organizational objectives facing our counterparts. This insight lets us predict how they will behave. If you don’t have empathy, you will be caught off guard. People in DEC were surprised that I would not endorse the nonsense in the memo about the Fellows program. They were also surprised at my observation that people at the level of Vice Presidents of the Bank might lose faith in the things DEC says if we give them a memo filled with nonsense.

Investing in trust means maintaining a reputation for giving our counterparts all the information we have that would be relevant in any negotiations; providing it whether or not it causes discomfort for them or us; and presenting this information clearly and concisely. To reiterate this point because it is so important, writing well is essential if we are to build trust. We cannot write any more sentences of the form “The current perception of STC as cheaper than Term staff has to be questioned.”

It also means that when we promise to do something, we do it. If we say we will do something, and discover that we can’t, we say so immediately so our counterparts can adjust.

Specific Decisions

  1. I told the Senior Management Team and the Secretariat to the Board that we will not deliver the four corporate flagship reports that were slated for publication in the fall of 2017. (See /assets/Flagships.pdf. ) We are canceling the Global Development Outlook and postponing the next volume of the Poverty and Shared Prosperity Report so that it comes out in 2018. With half as many reports, we can concentrate on delivering a World Development Report and a Doing Business Report that meet new, higher standards for the quality of our writing. I’ll come back to this again below. Clear writing is essential for trust.
  2. For now, I also decided not to challenge the cap of 387 staff employees for Development Economics that the Senior Management Team had previously approved. I also decided that I will not make a plea for any additional Bank Budget to support the corporate mandates we accepted, including work that supports household surveys.
  3. Before I am willing to make any pleas for more resources, we must build empathy and trust, first inside DEC then outside. To do both, we have to show that we will take actions to carry out the promises we have already made to support new corporate initiatives. In the absence of new resources, this means that we have to stop work on things that we have always done and shift resources toward the new initiatives. (Here by the way is a recent comment by BPS: “During the Business Review, a study of DEC’s MOUs, planning submissions and strategy notes highlighted a number of strategic shifts. No evidence was found to support budget resources being reallocated to match these strategic shifts.”) As a result, all units in DEC, including the front office, will start the process of looking for volunteers for mutually agreed separations and identifying positions that are no longer needed.
  4. In response to concerns expressed by the Board that the World Development Report is too reliant on support from trust funds, we will reduce the total cost, and thereby increase the fraction of the cost that we fund through bank budget, by stopping the practice of compensating units that supply people to the team.
  5. I’ve stopped funding any new proposals under the RSB program. I’m holding all the uncommitted funds in reserve so I can be sure to live up to my commitment that DEC absolutely will not overrun its budget again this year. It is far preferable for us to give some of the funds back than overrun again. If you want, you can think of anything we give back as repayment of funds we took from the rest of the organization by overrunning last year.
  6. I have recommended to DFID that we not fund any grant applications submitted in response to the third call for proposals by the SRP program. I will work with officials from DFID to find an alternative process that might better support the goals they set for this program.

Issues that need to be discussed

  1. I am not persuaded that maximizing the number of journal articles DECRG publishes will maximize the impact of Bank research. To be an intellectual leader and to take full advantage of our unique resources, we have to be able to make our own judgments about the expected value of any research project. If people in the Research Group want to pursue a project with a high potential impact that offers little prospect of a journal publication, it does not make sense to have a compensation and evaluation system that deters them from doing so. How might compensation in DECRG be restructured?
  2. We have an opaque system for deciding which questions from the rest of the Bank the Research Group will answer. A more transparent system would make it possible for people in the Research Group to get the credit they want for the cross-support that they do. The obvious way to get credit is to make the output visible. A more transparent system will also make it possible to track the work done against the priorities of the entire organization. One natural outgrowth would be something that is the functional equivalent of wikipedia, a place where anyone in the bank can find answers we have already provided to specific questions.
  3. Many opportunities for research with impact will require the use of information technology and/or the collection of new types of data. There are two very different ways to collect data–unilaterally or via collaboration with national statistical agencies. These data could be collected via surveys; or via new technological options such as web scraping, natural language processing, or low-cost sensors. For any given project, we must ensure that our organizational structure lets us mobilize teams with research, statistical, and tech skills that are well suited to the project and are not dictated by the unit in which someone sits.
  4. There is an opportunity to create some program that would provide useful experience for young people who want to continue their studies in economics and also provides useful inputs into the DEC’s work; perhaps even to the work of many of the global practices. As I’ve mentioned several times, RAND has a Ph.D. program. Why shouldn’t we consider our own graduate program, one that we structure to take advantage of our unique circumstances? There is a way to build such a program and be transparent about what we are doing.

Progress already

I started by asking ITS if I could have a server to experiment with. They gave me the one I’ve use to set up this blog. As we explored the possibility of letting anyone in Development Economics so something similar, it soon became apparent to Denis Robitaille, the acting VP of ITS, that it was not a viable option for the VP of DEC to be able to do what I wanted, allocate computing resources directly to people in DEC.

All it took to keep working toward a solution was for me and Denis to agree that it was important for DEC and ITS to build trust. Then I signaled that I understood and respected the mandates and constraints that ITS faces. Denis signaled that he understood my desire that researchers have easy access to cloud servers.

The solution that Denis came up with was to experiment with embedding someone from ITS inside of DEC, just as BPS embeds people in DEC. So within the next week or so, I’m very pleased to report that Dave Newsom, who is part of the ITS senior management team and has recently been running ITS Operations Strategy Unit will be moving from his office in ITS and will sit in the DEC front office. To help out, Parshant Tanavade, who has been working in the Operations Strategy Unit, will also be working from an office in DEC.


Research in the Bank can have stunning impact.

Our best shot at something big is to mobilize some of the many people inside the Bank who would love to work with us. With ITS, I’ve tested an approach to cooperation based on trust and empathy and have already seen the benefits it can yield.

To work with others, all those of us who work in DEC have to do is develop empathy and build trust. Our fate is in our hands.