Andrew Selee’s recent book, What Should Think Tanks Do: A Strategic Guide to Policy Impact, begins with a story about how our organization, the Center for Global Development (CGD), moved an idea for advance market commitments from an academic curiosity to a $1.5 billion vaccine program that has averted millions of childhood deaths. Other CGD “home runs” include the design of a $30 Nigeria debt-relief package and the creation of a satellite-based, high-resolution monitoring tool that is transforming efforts to end tropical deforestation.

Think tanks like ours can play a powerful—occasionally even transformative—role in the social sectors, offering donors impacts that extend far beyond those available through traditional charity. But many fall short. From a dozen years of failures and successes, we recently distilled twelve lessons that we hope can benefit other groups. We hope that others will read it in the spirit of one of our unofficial CGD mottos: “We take our work seriously, but not ourselves.”

  1. Start fresh to stay fresh. All institutions risk complacency, but this is especially relevant for think tanks whose job is to produce fresh ideas. Younger tanks have an advantage.
  2. Articulate an inspiring and results-driven mission. CGD’s founding chair Edward Scott, Jr. imagined a new type of development think tank that would not just write reports—it would work aggressively for practical, real-world changes. Scott set out to make the world better, not create a “sandbox for economists.” That drive has become part of our DNA.
  3. Begin with flexible money—but not too much. Ample, diversified financial support is critical for a think tank that aspires to remain independent and do more than chase contracts. We’ve found that a healthy donor portfolio for us is about three-quarters programmatic funds targeted for specific work, with the remainder unrestricted. Funder priorities are an important market signal of the value of our work, and engagement with donors on specific issues ensures that ideas flow in both directions. At the same time, flexible money allows a think tank to make venture investments in ideas that may not yet be ripe for donor support.
  4. Give great people plenty of freedom and responsibility. A think tank is its people. Attracting and retaining the best people means offering them the maximum amount of freedom and responsibility. CGD takes no institutional positions on issues, while fellows are encouraged to follow their investigative instincts to reach their own conclusions. Essential to our success is a corps of junior staff who typically work at CGD for two to three years before moving on to graduate school, or to jobs in government or other organizations. The steady turnover of youthful employees ensures a constant influx of new ideas and energy, while building a network for the future.
  5. Share leadership. Our president’s ability to participate in a shared leadership model is both unusual and a big part of our success. A lean management team (comprising the president, COO, CFO, vice president for communications and policy outreach, research manager, director for Europe, and a rotating pair of senior researchers) meets biweekly. The agenda is a mix of major institutional decisions (for example, should we recruit a fellow in Brazil?), and seemingly minor questions that are important for efficiency and institutional culture (for example, what should we do at our staff retreat?).
  6. Share ideas early and often. Writing papers in isolation and then publishing them is not a good way to engage policymakers and other researchers. Instead, we’ve found that sharing ideas in their formative stages through informal meetings, research-in-progress seminars, blog posts, and online working papers allows us to adjust our ideas and how we explain them.
  7. Don’t plan. Experiment! Other think tanks often ask for our strategic plan, and we elicit quizzical looks when we respond that we don’t have one. Our “strategy” is to build a stable of engaged scholars and a stock of well-researched ideas, and to be prepared to react to policy windows when they suddenly appear. We then track our ideas as they move up the agenda and (fingers crossed!) become reality.
  8. Partner with people, not organizations. Unlike a commercial joint venture, nonprofits lack a clear way to allocate the costs and benefits of collaboration. We’ve found that the best partnerships are specific, output-oriented ventures between individuals rather than vague partnerships between institutions. Similarly, CGD working groups bring together a broad range of perspectives from other organizations to tackle very specific problems over a limited time period.
  9. Resist the growth inertia. Pressures to grow are tremendous. At CGD, we believe our ideal size is large enough to have a critical mass of senior in-house staff members to cover a breadth of issues, but also small enough to maintain a family-like culture. The urge to expand the research agenda is also a constant. Staying small inevitably entails confronting tough tradeoffs.
  10. Keep it fun. CGD’s annual calendar of social events includes holiday parties, movie nights, a soccer team, a yoga class, a pumpkin carving contest, and a proud softball team called the Angry Birdsalls. These activities boost morale and reinforce a collegial culture.
  11. Celebrate (and try hard to measure) success. Tracking success in the think tank business is a difficult and elusive task. We maintain a list of CGD impacts on our website to collect the evidence of our successes, while also providing a check on claims that are hard to substantiate. We are also experimenting with more formal measures for assessing our impacts, such as an expected return tracker and an online policy pitch tracker, which measure progress in terms of singles, doubles, triples, and home runs.
  12. Ask tough questions. Our final lesson is to keep asking ourselves tough questions about what we do and how we do it. Among those currently under vigorous internal debate: How much time and effort should we invest in aid effectiveness (traditionally an area of core expertise for us) when aid is a shrinking share of finance to developing countries? What is the best way to engage with emerging powers? How much of our work should we devote to climate change—an urgent yet crowded policy issue? How can we stay ahead of the game in tracking our real-world impact?

We don’t know the answers yet, but we’ll keep asking.