On 23 November 2018, Agulhas Director Marcus Cox gave the keynote address at an annual workshop of law and justice development practitioners in Canberra, at the kind invitation of Professor Veronica Taylor from the Australian National University. Drawing on past evaluations of Australian and UK law and justice assistance, the address explores the shortcomings of traditional capacity building assistance, the progress made in shifting towards new forms of support, and why it is unhelpful to think of security and justice as a service sector.


Back in 2012, I led an evaluation team for AusAID’s Office of Development Effectiveness that took a strategic look at Australian law and justice assistance – which at the time constituted a very substantial $370m portfolio or 15% of the bilateral aid programme, off the back of the RAMSI intervention in the Solomon Islands and large programmes in PNG, East Timor and elsewhere.

In 2015, I had a chance to conduct a similar review of UK security and justice assistance, under the ICAI banner. The UK portfolio was significantly smaller – under £100 million or less than 2% of UK bilateral aid, and has since declined further.

The conclusion from both reviews was that the law and justice field was going through something of a crisis of confidence, in two senses.

  • First, many of the practitioners we spoke to were unsure that their efforts were bearing fruit. While people could point to pockets of achievement, the wider picture was one of inertia, with programmes implementing pretty standard sets of interventions without much conviction that they were altering people’s lives for the better.
  • Second, the field was struggling to tell a compelling story to policy makers about what it was trying to achieve. The slightly high-faluting objectives often ascribed to law and justice programme – from promoting the rule of law to more efficient economies to empowering the poor – all seemed rather a big ask for a set of technical assistance programmes. It’s hard to make a convincing value for money case if you haven’t settled on what value you’re selling.

This did not mean that all the programming we reviewed was poor. We spoke to many capable and committed practitioners, and saw many programmes that were genuinely innovative. In Indonesia, we were impressed by partnerships with CSOs and the Supreme Court on justice sector reform, and a really useful and grounded initiative to help women heads of households in poor communities obtain legal documents and thereby access public services. In Cambodia, Australian programmes had improved health services in prisons and were undertaking valuable pilots on crime prevention and community safety.

But the success stories often seemed somewhat fortuitous – a result of thoughtful TA providers being in the right place and time to spot a problem and do something about it.

The primary objectives of the programmes – the large institutional reform pieces – were largely unsuccessful on their own terms.

We found that security and justice institutions seemed to have an ability to absorb large amounts of training, equipment and technical support – not to mention attendance at regional conferences on security and justice – without changing their behaviours in any fundamental way.

Our conclusion was that there was something fundamentally wrong with the way we were approaching TA and capacity building. In short, we were offering technical solutions to problems that were not fundamentally technical in nature. And, insofar as we were building any capacity, we were doing so in contexts where the lack of capacity was not the binding constraint on institutional performance.

At the time of the 2012 evaluation, the thinking in the field was that we should move towards sector-wide approaches to law and justice. This reflected the thinking on aid effectiveness at the time, and the obvious fact that a justice sector is a chain of institutions, from police investigators through to corrections. Strengthening just one link will not make the chain any stronger.

But attempts to shepherd security and justice institutions into sector working groups and invite them to formulate ambitious sector-wide reform strategies produced, at best, long shopping lists of equipment and capital projects that donors were invited to pay for.

We concluded that there were a number of reasons why attempts to build capacity at agency and sector level were failing to get much traction.

One is that formal security and justice institutions themselves had very little incentive to reform. Inefficiencies within and between institutions create rent-seeking opportunities of various kinds. In many of the countries we visited, jobs as police investigators, prosecutors and judges were being sold to their incumbents for prices that reflected the rent-seeking opportunities they offered. However much the incumbents might personally be committed to the rule of law, they had to recoup their investment – and by extension, turn a blind eye to colleagues doing the same.

We also suspected that security and justice institutions were hemmed in by political red lines dictating what they could and couldn’t do. For example, in Cambodia at the time, there was a speculative bubble in real estate, and political elites were seizing land from poor people on the fringes of rapidly growing cities. There was no question of the judiciary acting as a check on this predatory behaviour – in fact, we were told that senior judges met with the powerful Interior Minister every Tuesday to discuss their weekly docket.

The law and justice practitioners we met were not naïve about these dynamics. Once in-country enough time, they had a fair idea about the informal rules of the game and the limits these placed on reform efforts. But these were concerns to be discussed in the evening over a bottle of wine, and set aside during the day so that they could get on with the jobs they were being paid for.

In fact, we also got that kind of double think from senior counterparts, many of whom were educated in London or Sydney or Auckland. They could talk eloquently about Robert Peel, the elegance of the common law or the rehabilitative role of a modern corrections system. But they were also adept at operating within the informal institutions – if they weren’t, they wouldn’t be in their jobs.

In summary, we began to suspect that the field of security and justice assistance was caught up in a powerful set of dynamics where security and justice institutions pretended to reform and we pretended to believe them.

Why is technical assistance so problematic?

Over the past few years, there has been a surge in interest in the political economy of development assistance. That has given us a more sophisticated conceptual vocabulary for understanding these dynamics.

Technical assistance in particular has come under intense scrutiny. There are a number of well-documented problems with common TA practice.

  • It tends to be supply-driven, pursuing donor rather than local priorities.
  • There is the perennial danger of capacity substitution – a product of foreign experts under pressure to get the job done.
  • There is the pernicious problem of the per diem culture, that can have a debilitating effect on national institutions.
  • And we are beginning to suspect that whole set up of TA, with its more or less explicit assumptions about expatriate expertise and local ignorance, is in itself disempowering, working against institutional learning.

But at the most fundamental level, I think we have come to recognise that traditional TA practice rests on a faulty theory of change.

We have known for some time that institutional templates don’t easily transplant across national boundaries, and particularly not across radically different economic and social contexts. For a decade or so, we have said that we are interested in ‘best fit’, rather than ‘best practice’.

But ‘best practice’ is a hard trap to avoid – our whole claim to expertise as TA providers comes from our knowledge of how institutions work in our home contexts. And law and justice practitioners in particular bring a strong normative outlook to their work – a powerful sense of the rules and principles that should underpin a justice system. This is deep in our training as lawyers, court officials, cops or corrections officers. We tend to be quite confronted when we encounter variance with that normative framework. In the words of one AFP official we interviewed back in 2012, this leads use to treat law and justice institutions in developing countries “as broken versions of the ones back home.”

The problem with institutional templating is not just that context matters. It is also a misunderstanding of how institutions change.

I find it useful to think of institutions as an ecosystem, populated by competing interests that have managed to achieve some kind of equilibrium. That equilibrium sets the boundaries of what is possible. A capacity constraint is not just an absence of technical knowledge and material resources, or a poorly designed structure or business process. It is a boundary created by a certain configuration of interests.

Of course, institutions can and do change all the time, as the underlying interests wax and wane and the compromises are renegotiated. But the shape of the change emerges from within the system, through this process of renegotiation – not from institutional designs offered by external advisers. It is more useful to think of institutions as biological systems, that grow and adapt, than as mechanical systems that can be re-engineered.

It’s true that there are often professionals within law and justice institutions in developing countries trying to do a better job in difficult circumstances, who are open to external assistance. They offer patches of firm ground in the swamp where a TA programme may be able to get some purchase. But the boundaries of how much change is possible are always set by the wider political context.

New thinking on TA?

So is there a way through the swamp? You’ll all be aware of the new thinking about TA and capacity building that has emerged in recent years. It goes by various names – Doing Development Differently, Thinking and Working Politically, Problem-Driven Iterative Adaptation (PDIA) – but they all proceed from similar premises.

  • They draw on the tools of political economy analysis to understand the interests and levers of power in any given system, to help identify scope for change.
  • They advise making small bets on change processes, to get some skin in the game and learn your way around the institutional and political landscape.
  • They involve identifying issues or challenges that matter to key stakeholders in the system, where there is an actual or potential constituency of change.
  • And rather than designing interventions with pre-set activities and outputs, they involve trying things out and then learning and adapting, while gently pushing on the boundaries of what change is possible.

In some ways, this is a radical re-conceptualisation of TA. It involves reimagining the role of the TA provider: not as the expert who knows all the answers, but as a critical friend who can help broker solutions and facilitate locally-led change processes.

Yet in other ways, I think it is what the most effective TA providers have always done. I said before that some of the best programming we saw seemed to have come about by thoughtful TA providers being in the right place and the right time to make things happen. I think these were people with an intuitive understanding that this was the right way to work and the soft skills to succeed.

So, do we know that this approach works? The answer to that is: yes and no. Yes, there is a growing body of well-documented case studies. There is no question that there are moments when external TA providers can catalyse important change. But we don’t know how often these moments arise. We don’t know whether it requires exceptionally gifted people, or whether it is a technique that anybody can learn.

And most importantly, we don’t know whether these approaches can be implemented at scale. Personally, I think that PDIA calls for small bets. The idea of large-scale PDIA strikes me as a contradiction in terms.

How much is it really happening?

How far have we got with retooling our practice along these lines? If you’ve been reading programme design documents in recent years, you might think that we’re now all about problem solving and adaptation. The reality is, I suspect, a lot less impressive, although I’ll be interested to hear others’ views on that. I think there’s a lot of lip service involved. PDIA risks becoming a rhetorical wrapper, or a small, speculative element tacked onto the side of traditional programmes. At its worse, it can be an excuse for muddling through. We come across that from time to time in ICAI: poorly conceived programmes that have to be redesigned mid-course, rather optimistically offered to us as examples of adaptive management.

There are a number of factors working against adaptive approaches and locking us into traditional TA.

  • Programmes are under pressure to deliver results quickly and at scale. They aren’t given the freedom to engage in extensive, exploratory design processes and come up with creative approaches.
  • There is also growing pressure to hold contractors accountable for their work. This includes moving to payment-by-results contracts, which lock them into predefined activities and outputs. PDIA requires us to think about accountability relationships within aid delivery in a completely different way, and we’re still working that through.

Having said that, I do think that Australian aid is inherently more flexible than UK aid – mainly because DFAT never took to logframes and value for money with the same enthusiasm as DFID.

Moving from institutions to justice

I turn now more briefly to the larger question of the role of law and justice in the development project, and how we can make the case to policy makers for spending in this area, given heightened competition for a shrinking aid budget.

One of the curious features of law and justice assistance is that there are too many reasons for providing it. Let me offer you five common ones.

  1. The first objective is stabilisation of post-conflict or fragile countries, including strengthening police and ensuring democratic control over security services.
  2. A second objective is community-level justice. Drawing on Amartya Sen’s writing on justice as intrinsic to development, the focus is on reducing insecurity and injustice in the lives of poor people, in areas such as land rights, violence against women and children, and crime and violence in slum settings.
  3. Third, there is the pursuit of social justice through law, which is known as legal empowerment. This is a more explicitly political agenda, based around civil society activism. Historically, the legal system is more often a tool of oppression than liberation, but its form creates the potential for it to be used in defence of the poor through public interest litigation.
  4. The fourth area is law and economics. For new institutional economists, the enforcement of contracts and property rights is essential to the efficient functioning of the economy, creating the certainty needed by investors.
  5. Finally, there is the rule of law and democracy. Because the rule of law is a basic constitutional principle in our own society, we sometimes suggest that strengthening legal institutions is a means of promoting democracy. This idea had a resurgence in the UK with David Cameron’s Golden Thread.

My view is that these very different objectives have been unhelpfully packaged together into the idea of law and justice as a service sector. If we strengthen the service providers, we will get more security, more just communities, more efficient markets and so on – in the same way that strengthening health services improves public health.

This thinking was explicit in the UK. A 2009 White Paper on international development included a commitment to treating security and justice “as a basic service, on a par with health or education”.

There are two major problems with thinking of law and justice as a service sector. One is that it leads easily into the trap of seeing the law and justice institutions themselves as the principal beneficiaries of external support, rather than the poor – which emerged as a significant concern in both the Australian and UK portfolios.

The other is that law and justice is not analogous to other public services. Security and justice does not flow automatically from stronger security and justice institutions. Improvements in personal security, for example, may come about through many other factors: the economic empowerment of women, better urban planning, proper sanitation in villages or gradual changes in underlying social norms. If the police force has been established to serve the interests of the regime, rather than the public, building its capacity may not be to the benefit of the poor. As one NGO expert put it to us, “you don’t necessarily get security by doing security”.

When it comes to law and economics, there is ample evidence to show that an independent judiciary is more likely to be the result of an economy reaching a certain level of sophistication, rather than its cause. In many rapidly growing Asian countries, the courts are neither independent nor particularly important actors in the economy. There is no reason to believe that investing in the judiciary is a useful way of kickstarting growth.

If we don’t think of security and justice as a service sector, how should we think of it?

My suggestion is that we stop defining our objectives in terms of improving security and justice, which is a vague and slightly grandiose notion. Instead, we focus our attention on addressing insecurity and injustice.  For development practitioners interested in poverty reduction, this means assessing, in any given context, which forms of insecurity or injustice are most harmful to the poor or to the country’s development objectives, and applying a problem-solving lens to those challenges. In problem-solving mode, capacity building of formal law and justice institutions is only one possible tool, and certainly not the default option. But it could well be part of the mix. In the UK, there is some impressive programming on violence against women and girls that combines a mainstreaming approach across education, health and WASH programming with innovative approaches to changing social norms and empowering adolescent girls, plus legal initiatives to promote and enforce laws against child marriage.

For other Australian government partners whose mandate is law and justice, you are I guess the proverbial man holding the hammer, but you can certainly be an informed and thoughtful wielder of that hammer. As well as supporting DFAT programmes where there’s a need, I would suggest that there are two legitimate and important spheres of action.

First, Pacific countries need a minimum of basic law and justice institutions to be stable and resilient. This includes core government legal offices, as well as courts and police. Small countries find these functions hard to resource, and Australia has an interest in preventing resource gaps from turning into vulnerabilities. Australian government agencies can help build and maintain core security and justice functions, including through secondments and over-the-horizon support. I don’t think this work can ever be sustainable, but it’s a long-term commitment that Australia needs to make.

Second, an increasing share of global development aid in the next decade will go into global and regional public goods – that is, tackling challenges that require cooperation with developing countries, to the benefit of both parties. These are issues such as fighting organised crime, terrorism, money laundering, people trafficking and smuggling, and tax evasion. As international climate action accelerates, there will be new challenges around regulating emissions, protecting fragile ecosystems and enforcing international agreements to protect the oceans. Many of these areas will require international legal cooperation, from the harmonisation of legislation to sophisticated multi-country law enforcement. These are all areas where Australian expertise and Australian aid dollars can be put to good effect.

So, I think there is a legitimate and necessary role for security and justice support in the aid programme. But headline message that I would like to leave you with is: let’s retire the idea that investing in building the capacity of security and justice service providers is the primary means of increasing security and justice for poor people.