How to Do Technical Assistance (Well)

Declining aid budgets are changing how major donors operate. The UK’s Minister for International Development Jenny Chapman has said that UK aid will shift “from grants to expertise” and “from service delivery to system support.”

What that really means is that there is no more money (at least bilaterally), but we can give you some advice. Perhaps not the easiest pitch! In his reviews on aid effectiveness in the 2000s, Bill Easterly maligned “technical assistance” as one of the “worst practices” in foreign aid, on the grounds it was generally seen as ineffective and tied to consultants provided by the donor country.

Through my work with African governments, I have seen clear examples of both good and bad models of technical assistance. Recent evidence suggests that we should think of training and consultancy as making marginal but valuable improvements. Improving the productivity of a government ministry or department by 5 percent would be incredibly valuable, but also essentially impossible to observe. How would you even know? Whilst no one notices 5 percent annual productivity growth, keep that up and you'll double productivity in 14 years.

Can technical assistance actually work?

Good technical assistance seeks to strengthen governments’ ability to make better decisions and implement policy well. This is inherently difficult to evaluate quantitatively, as the unit of observation is in one sense the country. This leads us to the cross-country regression, and the challenges of inherently limited sample sizes and trying to identify causality with observational data. Despite these challenges, there is actually some recent evidence that countries that receive advice from the IMF on taxes do go on to raise more money. A 10 percent increase in the volume of technical assistance is associated with a 1 percentage point increase in Value-Added Tax (VAT) as a share of GDP (though it is worth bearing in mind that IMF advice often comes coupled with loans, so advice without any money might be less effective).

We can also learn from programs where national governments provide technical assistance to local governments. There are only around 200 national governments in the world, but thousands of local governments, so we have much bigger samples to work with. An evaluation of technical assistance attached to funding for local public works in Chile and Guatemala found that the marginal effect of TA (in addition to the funds) improved outcomes by 4–20 percent, which was highly cost-effective given the relatively low-cost of the TA. In education, technical assistance from the state government of Ceara in Brazil to local Mayors improved test scores.

Perhaps the strongest evidence comes from individually randomised training for civil servants. Azulai et al find that a one-day training for civil servants in Ghana increased productivity as measured by independent auditors.

We can also learn by analogy from training in other areas. It’s easy to poke fun at consultants who borrow your watch to tell you the time, but the evidence on private consultants for business is actually a little more impressive than that. For instance, a new paper with data on every business transaction in Belgium over 20 years finds that hiring consultants increases labour productivity by 4 percent. Experimental evidence also finds gains. Advice from international consultants for Indian textile firms increased productivity by 17 percent, and by local consultants for Mexican SMEs by 20 percent of a standard deviation. More generally, a recent meta-analysis on the effect of small business training programs shows effects on profits and sales of 5–10 percent. From the public sector there is a rich literature on the effects of coaching teachers and training school leaders, which again shows small but meaningful effects.

The UK typically spends aid money on technical assistance by a) hiring consultants, b) hiring NGOs, or c) contributing to multilaterals. In business consulting, production and outcomes are more clearly observable than in much policy work. When outcomes are not easily observable you can’t contract on quality, so private providers face the incentive to cut costs. So perhaps technical assistance is better delivered by workers with mission motivation.

Finally, coming back to magnitudes, how cost-effective is technical assistance? Of course, the impact of any individual project might well be zero, but if we could expect a 5 percent improvement in productivity on average, think of a government Ministry spending $100 million per year. Five percent of that is $5 million, significantly more than the cost of a full-time advisor and some targeted short-term consultants. That level of value is supported by our survey of Ministry of Education officials across 35 aid recipient countries, who were indifferent on average between a) a $33 million aid project and b) a $30 million project that included a full-time advisor.

How to do technical assistance well

Technical assistance can provide two distinct functions: ideas and execution. The execution part is probably what most people think of first; building the capacity of organisations to function effectively. In some cases this might constitute “gap-filling” in critical functions within dysfunctional states, but ideally, it is focused on improving organizational management. Not just providing training, but also building better systems.

The other part, though, ideas, may be just as important. It’s no use getting things done if you’re not doing the right things. Here, international experts can provide invaluable outside perspectives, pushing governments to learn from other countries and think longer-term beyond the next election.

The best advisors I’ve seen up close have been focused on execution. They work with mixed teams of national government staff and international advisors. They base themselves in the same location as local officials—often hot and dusty government offices and not the nice modern air-conditioned UN or NGO office. They’re humble, listen and follow direction from senior government colleagues, whilst also providing challenge where appropriate, and above all driving action; keeping momentum through the basics of day-to-day management, ensuring roles and responsibilities are assigned and checking in regularly on progress. Being involved in day-to-day conversations means understanding when and where it might be valuable to bring in more specific specialist skills, or experience from other countries. If you need to reform sales tax you don’t want to just copy-and-paste regulations from elsewhere, but there are a handful of global experts on the process of reforming sales tax that you probably do want to talk to.

Training needs to be targeted at the right individuals. Governments in low-income countries are regularly emptied out with staff away from their desk at donor-funded training workshops that are irrelevant to their job. To do the micro-targeting of providing the right training for the right people at the right time, you need someone to be physically present day-to-day to really understand what the organisational and individual needs are.

Good examples of this kind of technical assistance are CGD’s Scott Family Liberia Fellows program, the ODI’s fellowship and Budget Strengthening Initiative, the Clinton Health Access Initiative (CHAI), what used to be the Africa Governance Initiative from the Tony Blair Institute, and the Oxford Policy Fellowship. The recent experience of the new Lead Exposure Elimination Project is also instructive. There, technical assistance is targeted on a specific technical problem (enhancing and enforcing lead paint regulations), an area where there is little demand for action because the problem is invisible, but also almost no political opposition to change.

One thing all of these groups have in common is their commitment to a specific mission. This is made easier by being nonprofits. I’ve worked with excellent embedded advisors employed by for-profit consultancy firms, but ultimately they were excellent individuals working in spite of the natural incentives facing for-profit consultancies. There are echoes here of debates around the role of the private sector in public service delivery—quality is not always observable or easily contractible for.

In the realm of ideas, Lant Pritchett’s description of experience with the Indian think tank ICRIER is instructive. This wasn’t direct technical assistance to the government or about international advisers, but it was an internationally funded effort to influence the direction of government policy, with phenomenal returns. More broadly, think of last year’s economics Nobel prize winner Joel Mokyr. His key insight is that for sustained economic growth you need a society that is open to change. Growth creates winners and losers and you need to find ways to stop the losers from blocking progress. International funding might be more effective not in directly providing piecemeal assistance, but in supporting a robust domestic ecosystem including think tanks and research centres.

So what should donors actually do?

How do you set up the best project or program for technical assistance? One strong model is Rwanda’s Strategic Capacity Building Initiative. This is government-led, focuses on strategic priorities, with embedded advisers, coordinating the support of multiple donors. Rwanda is now the highest rated country by the World Bank on its “quality of public administration” score, and the best performer globally on the Blavatnik Index of Public Administration (after adjusting for per capita GDP). But not all countries have the same focused leadership. But there are typically at least pockets of effectiveness in even the most dysfunctional states (there are some great specific individuals featured in Dan Honig’s book Mission-Driven Bureaucrats). Finding and supporting these mid-level leaders should be a priority.

I’ll close with five practical suggestions.

1. First, don’t tie aid! This is a more straightforward issue for goods, but less obvious for services. The UK has lots of great expertise that we can share, but we’re not the only country with expertise. If our government is buying services for our international partners, we should buy them the best value advice no matter where the advisor happens to come from. That might mean British experts sometimes, but it also might mean Indian or Kenyan or American experts.
2. Second, where the quality of outcomes are hard to observe, we should seek to fund technical assistance from mission-driven organisations, be they non-profit NGOs, government civil service, or multilaterals.
3. Third, we should embed consultants and advisors with government counterparts, not in separate donor offices. The most effective technical assistance involves mixed teams working side-by-side in government buildings, engaged in the daily rhythm of decision-making rather than periodic consultations. The recipient government should be involved in recruitment, helping to select which experts they want.
4. Fourth, combine ideas with implementation support. Don't just provide policy recommendations—help governments think through the political economy of reform while simultaneously building the management systems needed to execute. The returns to helping governments prioritize what to do may be just as high as helping them improve how they do it.
5. Finally, not every project needs an impact evaluation. Researchers have often under-appreciated management training interventions because individual studies had too small sample sizes to be able to measure a small improvement. But whilst a small improvement in government productivity might be invisible year-to-year, if sustained it could be transformative over a decade. Donors should be open-eyed about the need for longer time horizons and fundamental uncertainty about whether we’ll ever know if a project had impact.

The cuts in foreign aid from several major donors are lamentable and will cause real harm for very little fiscal saving. Given the reality of the cuts, there is some wisdom in focusing on relatively low-cost technical assistance that offers the potential to improve and leverage existing spending by recipient governments. If donors can resist the urge to tie assistance to their own consultants, stay humble about what outsiders with meagre financial resources can realistically contribute, and commit to the patient work of long-term institutional strengthening.