May 3, 2013

How to improve social policy design

A collaborative approach to solving development problems in India. 

It is widely acknowledged that program implementation is the Achilles Heel of India’s massive social policy administration. The search for explanations has revolved around lack of outcomes focus in policy design and inadequate professionalism in implementation.

In this context, experimental researchers advocate “evidence-based” policy design, with evidence emerging from experimental research. They favour using field tests, like randomised control trials, to identify causal links between specific program designs and its desired outcomes. The works of researchers like Esther Duflo and Abhijit Banerjee at the Abdul Latif Jameel Poverty Action Lab (J-PAL) have pushed this into the center-stage of modern development economics, despite criticisms about the generalisability of its findings.


Management scholars attribute the failures to the absence of professional program management expertise within government bureaucracies. A few large private sector consulting organisations have therefore, in recent years, started to apply their standard problem-solving techniques to improving social policy implementation. An innovative integration of the two should be the way forward.

Here a distinction between policy design and implementation strategy is in order. While the policy is formulated at national or state level and can be more broadly defined, the details of the implementation strategy is context-specific. Inadequate attention to the latter has led to many excellent examples of policy design stumbling at the last mile of implementation.

Instead of top-down solutions that start with a theoretical hypothesis, highly context-specific social sector issues require an approach that begins with the problem itself. The problem should be unpacked through a process of discovering latent institutional knowledge and experimental research, with the solution emerging bottom-up from this process. This process demands close collaboration between government bureaucracies, consultants, and experimental researchers.

A public system will hire consultants to assist in program implementation. They would embed themselves within the system for a short period, undertake a comprehensive problem-solving exercise, and formalise an implementation blue-print, with its identified uncertain elements. Multiple versions of the program will then be piloted and the uncertainties resolved through iterative field experimentation.

Consider the example of a large-scale placement-linked skills development program for youth. The program consultant would formulate the implementation blueprint based on the broad policy design. Three (or four) versions of the program, varying based on, say, the method of mobilising the trainees and how the private training agency and employers interact, can be tested for a short duration as iterative pilots with tight feedback loops. Experimental researchers can help design these pilots and use the data to refine and confirm the final scalable program version. Simultaneously the consultants will help build professional implementation capacity in the government bureaucracy and exit once the scale-up stabilises.

Similarly, after a problem-solving exercise, an intervention aimed at improving student learning outcomes can be tailored around remedial instruction. But the implementation blue-print would reflect the uncertain elements of this plan – in- or after-school remediation, integration of remediation into regular classroom instruction, grouping of children, need for an additional contract teacher etc. Experimental techniques, including randomised control trials, can be used to identify which is the most effective strategy for each of these elements, for the specific context. However, instead of expensive and long-drawn pilots, three or four versions of the original blueprint, reflecting these differences, can be implemented for a short duration as iterative pilots. The iterative process will be facilitated by tight feedback loops, which would help in the continuous refinement of the original blue-print.

The biggest challenge with this approach is “missing market”. Specifically, given the constraints imposed by public sector procurements, there is a deficiency of providers who can service this market. There are three problems with this. First, public procurement rules preclude anybody other than established and experienced agencies. Second, the learning curve for management consultants, tutored in private sector problem-solving, is most often too steep. Finally, the price point for service delivery needs to be much lower than the exorbitant hours-based fees currently charged by the top-line consultants.

This means that the large and established consulting organisations are the only ones ‘officially’ eligible to bid. But unless they reinvent themselves dramatically, they are too deeply internalised with the ethos of private sector problem solving to be successful. Their social sector domain expertise and experience of public systems, which are not amenable to templates-driven problem-solving, is too limited to make them effective contributors to program implementation. In any case, their price point is so high as to make their business model a non-starter in social sector consulting.

For all the aforementioned reasons, I am not hopeful that the large consulting organisations can bridge the “missing market”. Smaller consulting firms, with requisite domain expertise, experience with public systems, and willing to offer their service at a outcomes-based price-point, stand a greater likelihood of succeeding. They are more likely to be able to provide the high-quality manpower to be embedded within the government bureaucracies for sufficiently long enough periods to make the program sustainable and also transfer professional program management skills to the public system.

There is also the important issue of identification of an appropriate entry level for the external facilitators. In India, the district, being at the cutting edge of implementation, is the right level for such problem-solving exercises. They provide a large enough canvas but a reasonably similar environment for a unified policy design and implementation strategy.

A partnership between the large private foundations that finance numerous social sector projects and organisations like the J-PAL could catalyse the development of this “missing market”. A few successful collaborations could unlock the market. Over time, the consultants would acquire the expertise to manage field experiments too and public bureaucracies would develop professional program implementation capabilities.

None of this should overlook the fact that these external participants are only enablers and the program implementation would be the responsibility of public officials.

Photo: Tatinauk

Fatal error: Uncaught Error: [] operator not supported for strings in /home/thinkpra/public_html/archives/wp-content/themes/layerswp/core/helpers/post.php:62 Stack trace: #0 /home/thinkpra/public_html/archives/wp-content/themes/layerswp/partials/content-single.php(81): layers_post_meta(4691) #1 /home/thinkpra/public_html/archives/wp-includes/template.php(724): require('/home/thinkpra/...') #2 /home/thinkpra/public_html/archives/wp-includes/template.php(671): load_template('/home/thinkpra/...', false) #3 /home/thinkpra/public_html/archives/wp-includes/general-template.php(168): locate_template(Array, true, false) #4 /home/thinkpra/public_html/archives/wp-content/themes/layerswp/single.php(20): get_template_part('partials/conten...', 'single') #5 /home/thinkpra/public_html/archives/wp-includes/template-loader.php(78): include('/home/thinkpra/...') #6 /home/thinkpra/public_html/archives/wp-blog-header.php(19): require_once('/home/thinkpra/...') #7 /home/thinkpra/public_html/archives/index.php(17): require('/home/thinkpra/... in /home/thinkpra/public_html/archives/wp-content/themes/layerswp/core/helpers/post.php on line 62