Death to the policy report
Think tanks should spend precious time on more useful activities
A recent Washington Post editorial “Industrial Policy for Dummies” savaged a 276-page World Bank report. My first thought was to feel sorry for the authors who were no doubt proud to have their herculean effort finally published, only to be skewered by the hometown paper.
But my next thought: why is the World Bank still publishing 276-page reports?
I have some ideas – and a plea.
No one reads. So why are so many organizations spending so much time and effort publishing long reports?
The World Bank has long known about its weak readership problem. A 2014 internal analysis, “Which World Bank Reports Are Widely Read?” found
About 13 percent of policy reports were downloaded at least 250 times while more than 31 percent of policy reports are never downloaded. Almost 87 percent of policy reports were never cited.
The Washington Post merrily covered that story too while SSIR used it as a hook to ask, “If a Report Is Published and No One Reads It, Did It Really Happen?”
The World Bank is not alone. The United Nations found similar disappointing downloads for the 1,100 reports produced in 2024. The Secretary General explained
Many of these reports are not widely read…. The top 5% of reports are downloaded over 5,500 times, while one in five reports receives fewer than 1,000 downloads. And downloading doesn’t necessarily mean reading.
I’m picking on the World Bank and UN because they’re easy targets. But my fellow think tanks fall (jump?) into the same trap. We use social media, videos, and all kinds of communications tools. But we also write loads of long reports. I recently visited a lovely think tank who proudly decorated their entire lobby with dozens of their report covers. A ChatGTP query estimates 800-1,500 long-form PDF reports were published last year by the dozen most prominent US think tanks.
Most orgs have no clue of the impact of these reports. They can track downloads or media citations, but only rarely can they point to the specific effect a report has on government policy or a real-world outcome. Yet poor readership is a signal that, in my view, likely points to a bigger problem: the report itself.
What are policy reports good for?
That anyone would bother to spend months or years writing a long report they knew few will read suggests they have some implied value. The main reasons to write one:
Share new data, a new detailed technical analysis, or make a new convincing case to a specific audience. Yes, when these are the goals and only a hefty report can do the trick, they can be useful. Sometimes (I would say rarely) this is the clear purpose.
Weight as a signal of credibility. Stacks of data, charts, and words create the appearance of deep research, even if the ideas or conclusion might be better conveyed in another format. For some audiences, only something long can be valid. I once had a senior African official demand “the full report” after I shared a short one.
As a funder deliverable. A report is something concrete. It’s either published or not. You can hold one in your hands. So it’s often what’s shown to donors as proof of activity. “You paid for this.” But outputs are definitely not impact. Some funders are apparently fine with this, but, in my experience, most are not.
I’m sure there are other reasons I’m forgetting. Please email me if you have ideas.
Long reports have a high cost
I have many more reasons why not to write them.
Waste of money. Add up all the staff salaries, reviewers, editors, designers, publication costs, etc and a single think tank report can easily run into six figures. Signature reports at big official institutions can, all in, run deep into seven figures.
Tragic waste of time. The opportunity cost to busy people – who should be world-leading experts on vital social, economic, and political issues – is actually mind-boggling. Imagine the social good such smart capable people could do with all the time not spent on long reports.
Tedious repackaging. Lots of reports I see are mostly cut and paste jobs with little real new information. If you have some new analysis to share, why wrap it in thousands of extra words that have already been written many times elsewhere?
Mistake activity for action. Some official reports are written because they are mandated by members or parliaments. But think tanks are independent organizations that, in theory, should be deciding what they work on. A think tank is not the UN, so they can choose the problems to tackle and what activities to undertake. Writing a report takes a lot of effort, but it’s often the lazy answer to addressing a wicked problem. (What to do about global poverty or climate change? Write a report!)
Too often a substitute for progress. This gets to the core purpose of a think tank. Do they exist to stoke egos and provide a soft landing? Or are they built to solve real problems that governments cannot or will not tackle? I lean heavily to the latter.
What’s worse than a new policy report? My 20-year old policy reports!
I’ve contributed to many reports over my career. Two stand out in contrast that greatly shaped my views – and help explain my current antipathy.
Early in my career, I was on the World Bank team that drafted the Africa regional strategy. We spent over a year gathering ideas and writing. Consultations spanned all relevant parts of the Bank, took the team to Ethiopia to meet with President Meles, and then to a group retreat in South Africa. It was eventually published as a glossy report, I’ll guess about 120 pages long, blandly titled the “Strategic Framework for Investment in Africa” in 2003.
I certainly learned a lot (mostly about how the World Bank works) and met some incredibly smart, dedicated people. But I don’t think the report had any meaningful effect on the bank’s strategy, policy advice, or investment behavior. It’s no longer even on worldbank.org. (I think I have a copy buried deep in a box somewhere in my storage unit.) What I can say for sure is that it consumed me and another full time person for 18 months, plus nontrivial chunks of time from the VP, chief economist, operations director, sector directors, country directors, country economists, editors, designers, and other contractors. I’ll conservatively estimate that the report cost at least $2 million directly and probably 2-3x that once all the time of others involved is added in. And for what? I cannot say.
A few years later, I co-led a working group at the Center for Global Development that was very explicitly trying to reshape the African Development Bank from the outside. The bank had just come through a near-bankruptcy, had utterly lost credibility with its own shareholders, and had just been forcibly relocated to Tunis after a civil war broke out in Cote d’Ivoire. Its survival was not assured.
So we assembled a group of 16 experts to float some ideas to incoming President Donald Kaberuka and to pressure the board to give him space to rebuild. The output of our effort was, “Building Africa’s Development Bank: Six Recommendations for the AfDB and its Shareholders.” The whole thing was 27 pages, with just 11 pages of core text. Its heart were three sharp recs for the president (to focus on growth and infrastructure) and three for the board (to back off and let Kaberuka do his job).
We were deliberately crisp and provocative as a way to constructively stir the pot. Several of us from the working group flew to Tunis to brief the board in a lively exchange. I don’t know for sure if we made an appreciable difference or were pushing on an open door. Yet Kaberuka reoriented around growth and infrastructure and, by all accounts, was the most effective leader the bank has ever had. The shareholders, predictably, did not back off.
The main lessons I draw from comparing these two reports: be crystal clear about your purpose, know your role, and make it only as long as it needs to be.
Or… if a report is not what’s needed, do something else.
To make this point, here’s a quick summary of another career-altering experience. Around the same time I joined CGD, Ngozi Okonjo-Iweala became Nigeria’s finance minister and was grappling with what to do with $33bn in inherited external debt, mainly to rich governments like the US and UK. She asked my boss Nancy Birdsall for any helpful analysis we might provide and, being the new guy, Nancy asked me to take a look and come up with ideas. The US had just helped post-Saddam Iraq get debt relief while Argentina was negotiating with commercial creditors for a deal too and, fortunately, Nigeria had just built up savings from an oil windfall.
One win-win idea that seemed obvious was to have Nigeria buy back its debt at a steep discount, using Iraq and Argentina haircuts as benchmarks. But no one – neither the Nigerians nor the western governments – wanted to be the first to propose this. So after floating the idea with lots of people for feedback, I wrote down the discounted buyback idea in an ultra simple 2-page note. A whole long, complicated story short, the 2-pager was used by negotiators as the starting guide for an eventual deal. The CGD note played a small but catalytic part in saving Nigerian taxpayers $18 billion and opening the door for the country to access new borrowing.
I cannot say for sure that a deal would never have been struck without our note, but I am absolutely confident – and was told so by a senior US Treasury negotiator – that a long report would not have had the same impact as our digestible 2-pager. There was no “full report” needed to share this idea and have it reach the target audience. Instead, it was the opposite.
So, sure, go ahead, write a report if that’s what the world needs, but don’t do it by lazy default
My plea to fellow think tankers is, before you write your next report, ask yourself a few questions upfront:
What’s our objective?
Who do we need to reach?
How to best convey our analysis and ideas to that specific audience?
What kind of language, tone, and format will give us the best shot at impact?
How short can we make it?
If the answer to these questions leads you to write a long policy report, fine. If not, please do something else. You – and the world – will thank me for it.
This is SO good. Thank you for writing it!
Absolutely, knowing your role, your audience, and bringing new information/views is absolutely key. And you hardly ever have an audience craving 300 page reports.