If you had to summarise the government’s Whitehall reform plans in a single word, ‘data’ would not be a bad choice. The dominant theme of Michael Gove’s Ditchley lecture on the public service was the need for a more numerate, statistically aware government. This autumn, the government will have an important opportunity to show it is serious about this commitment, as the Department for International Development is merged into the new Foreign, Commonwealth and Development Office.
To the layperson, DfID’s defining characteristic was its deep-pocketed largesse – as the guardian of the UK’s commitment to give 0.7% of gross natonal income as overseas aid, no other department spent so much to such altruistic ends, ends which also benefitted the UK in terms of global security, disease prevention etc. But from a Whitehall point of view, the most interesting part of DfID’s departmental anatomy was not its big heart, but its sharp, quantitative brain.
Behind the scenes, DfID was exceptional for its commitment to statistical rigour, to evidence-based policy, and to robust impact evaluation. It was well connected to those within the international development field who stood up for the use of data and evidence, often in the face of significant opposition. It is no coincidence that DfID’s last chief economist had previously run Harvard University’s renowned Jameel Poverty Action Lab, a sort of Mecca for evidence geeks, which pioneered the use of randomised controlled trials in development (and controversially exposed a number of common aid practices as being ineffective in the process). DfID’s commitment to data and evidence also led it to make invaluable investments in open economic and development statistics around the world, such as its substantial contribution to the International Monetary Fund’s programme on data dissemination.
Unusually for a government agency, DfID’s work attracted praise from the most demanding enthusiasts for evidence in the world of philanthropy and development, including institutions like the Gates Foundation, but also members of the so-called Effective Altruism movement, a loose grouping of what Number 10 might admiringly call ‘freaks and weirdos’, who have developed their own quantitatively driven philosophy for how to do good in the world. DfID’s commitment to experiment led to high-risk, high-return successes, such as the DfiD-Vodafone M-Pesa banking project in Kenya. Just as importantly, it meant that when projects do not work they can be terminated quickly and lessons learnt.
The principles behind DfID’s statistical work – using evidence rigorously, collecting data methodically, publishing it openly, and using it to drive decision making – chime perfectly with Michael Gove’s call for a more data-savvy civil service. The challenge now is how to preserve these characteristics in what most people assume is a merger of unequals, with the ancient and prestigious Foreign Office.
The Foreign Office has no shortage of clever and effective officials. But even its staunchest supporters would agree that its skills lie more in diplomacy, guile and grand strategy than in the geeky worlds of statistics, randomisation or data science.
The essential challenge for the new Foreign, Commonwealth and Development Office will be ensuring DfID’s strong culture of evidence, data and experimentation are preserved, and that the new department learns and builds on the enviable statistical legacy that DfID has built. This will require careful management, and a certain amount of experiment – to understand the value of quantitative methods and what they can offer to the Foreign Office’s traditional mission, as well as to the world of overseas aid.
This will also be a wider test-case for the Whitehall reform programme. Building a more data-savvy civil service will no doubt require investing in some entirely new capacity – such as the new Downing Street ‘skunkworks’, 10ds. But the task will be a futile one if it cannot harness and build on the significant data skills that already exist across the civil service, in departments, in agencies, and in professions, especially the Government Statistical Service. Sometimes these capabilities need to be better connected to policy staff, more plugged into to ministerial decision-making, and more integrated with other departmental functions – but neglecting or ignoring them will be a big mistake.
DfID’s data capacity is an important, if under-recognised, asset – and one that has won impressive international respect. If government can keep these skills alive, invest in them, and make them central to the new Foreign, Commonwealth and Development Office, it will offer an important lesson for the future of the Whitehall reform programme. It is vital that we rise to the challenge.
Prof Ian Plewis is chair of the Royal Statistical Society International Development Section