Blog / DevLearn Digest 11: A Sceptical Perspective on MSD for Employment
31 January 2024

DevLearn Digest 11: A Sceptical Perspective on MSD for Employment

DevLearn Digest 11: A Sceptical Perspective on MSD for Employment

Hello everyone,

This Digest examines the challenge of using market systems development to generate employment – something many of you are probably wrestling with.

Before we dive in, the next DevLearn online training courses in May are now open for booking. You can take courses on market systems development, monitoring, evaluation and learning, or both. There’s an early bird discount of 10% before the end of February, and our courses tend to get booked up well in advance – so if you’re interested in applying, please do so! If not, we’d be very grateful if you could share it with a friend or colleague who might be interested.

Now, on with the Digest!

For a long time, the market systems development (MSD) community was dominated by rural agriculture. Over 75% of the BEAM Exchange’s evidence map, for example, is drawn from programmes working in the agricultural sector. Employment creation was mostly a secondary benefit, primarily consisting of seasonal agricultural labour.

Now a burst of enthusiasm for employment is sweeping through the MSD community, seeking to elevate job creation to a core MSD objective. This is inspired by new programmes (such as LIWAY in Ethiopia and EYE in Kosovo), guidance (with Rough Guides to Youth Employment and Measuring Job Quality) and communities of practice. The theoretical case is undeniable. Most people would agree that providing high quality jobs to young people is important, and readers of this mailing list hopefully appreciate the value of the MSD approach.

This newsletter aims to introduce a dose of scepticism into the debate. I’m certainly not saying that MSD for Employment (MSD4E, for short) is unworkable, or even that it is unlikely to succeed. However, I think that it faces significant and under-appreciated challenges, and the jury is still out as to whether it can succeed at scale.

The key challenge is that successful MSD programmes typically build on the resources and assets of their target group, who are closely linked to essential supporting services. A farmer, for example, typically owns the land that they are farming, and the equipment that they use to farm it. They have linkages to input suppliers, markets, suppliers and other farmers, and a lifetime of farming expertise to draw from. Of course, plenty of things are not working well – this is why MSD interventions are needed. However, there is a partially functioning system that enables textbook MSD interventions – access to finance, seedlings, and other supporting services – to have a rapid impact at scale.

In MSD4E, by contrast, the target group have fewer assets. If you’re looking for a job, you probably don’t own a factory. Of course, job-seekers bring enthusiasm, technical skills, and soft skills. But, as you might remember from nervous job interviews at the start of your own career, enthusiasm only gets you so far. Especially in a context of high levels of underemployment, there are few opportunities for minor innovations that can deliver big results.

Moreover, employment markets incorporate fewer important support markets for programmes to leverage. Think back again to your job-seeking days. You probably used a jobs board, and maybe visited a career councillor. But the depth and importance of these relationships are not comparable to agricultural production. A smallholder farmer spends far more on agricultural inputs than a job-seeker does on job-seeking. Moreover, many of the employment support functions are managed by the public sector, which provides education, training, subsidised childcare and job-matching in many wealthy countries. There’s a good reason for this – the user of these services can’t afford to pay.

None of this is to say that MSD4E is doomed to failure. I’m personally excited about jobtech, which offer access to a global market worth billions of dollars. Working with self-employment rather than waged employment addresses some of the above concerns, though seems similar to the Business Development Services movement that pre-dated MSD, from the ancient days of 2001. And working with the public sector is an effective strategy in some countries, like Ethiopia, with a relatively capable government.

But I think the above challenges will significantly hinder MSD4E in reaching scale. They block the pathways that traditional MSD programmes have used to quickly reach hundreds of thousands of users. And without scale, programmes that spend millions – or tens of millions – or sometimes even hundreds of millions – of dollars will struggle to justify themselves. Indeed, many of the examples I’ve seen of MSD4E have grappled with exactly that problem.

Perhaps the underlying issue for MSD4E is a lack of a clear theory of change. Despite a proliferation of intervention-level results chains, MSD4E is agnostic on key questions underpinning any employment programme. What is the role of industrial policy in driving structural transformation? How should we think about foreign direct investment? Should we prioritise micro-enterprises or formal jobs in modern, large companies? Are agricultural jobs a poverty trap? Is it true that skills development is ineffective? Context always matters, and I understand the inclination to leave each programme to find their own path. But a good framework for analysis, design and implementation cannot be completely open; it should guide projects towards a strong, evidence-based answer. 

The overall principles of the MSD approach are always relevant. Be humble, iterative, and thoughtful. Seek partnerships with local actors, and think about sustainability. Understand the context you are working in. Few people would argue with that. But the challenges of employment generation are real, and we believe that there is more work to do to ensure that the tools and frameworks from MSD can be fruitfully applied.

That’s it from us for this DevLearn Digest! As always, we would be keen to hear your thoughts and responses, and suggestions for future topics of interest,

Adam and the DevLearn team