Why do we gloss over state capability deficiencies?

Why do we consistently under-estimate state capability deficit and see technology and other innovations as the solution to public policy problems?

Consider three examples. One, using GIS to improve property tax collection. Two, technology interventions (GIS mapping, SCADA, smart meters and smart grids, etc) or financial engineering of state balance sheets to reduce distribution losses in their electricity distribution companies. Three, state-of-the-art regulation and processes to improve the effectiveness of regulatory institutions. All have been tried ad-infinitum not just in India, but across the world, with minimal success. 

It is not to say that these are not useful, but just that the underlying problems are fundamentally about weak state capability (at local government, public sector unit, and state/central government levels respectively above), and without atleast partially addressing them, the fixes suggested will hardly make a dent on the problem. In all these cases, policy makers over-estimate the contributions of technology interventions and process re-engineering in delivering the desired policy objectives. In fact, there are atleast six distinct biases that nudge us into an unqualified embrace of such interventions.

1. The desire for the tangible and conclusive. It has become part of the social internalization that everyone now views a policy intervention in terms of norms and components, clearly defined processes, time-lines, and a list of outcomes, all of which should form the basis for scaling up. Even when operational flexibility is afforded, the overall architecture is generally self-enclosed.

Unfortunately, most such policy interventions are transactional, requiring continuous engagement by officials at the cutting-edge with other human stakeholders. Such engagements, the quality of which is critical, cannot be prescriptive and decreed into implementation. They require capable and engaged individuals, who are sufficiently empowered (with resources and operational freedom), to discharge their responsibilities in an environment which allows them the freedom to do so.

Given that all these assumptions are questionable in the median case, the scaling-up challenge becomes simply humongous. We need less prescriptive and uniform approaches to dealing with the problem. But such approaches involve providing considerable program design and implementation autonomy with attendant delegation of responsibilities, which would make monitoring far more difficult (in fact, the current type of monitoring pretty much impossible) and outcomes less certain. In other words, this is an altogether different program design, implementation and monitoring paradigm. It would need greater tolerance for failures and turmoil and adoption of more dynamic program management approaches.

2. The partial equilibrium bias. In our problem solving moments, even when we are most logical, we rarely go beyond the first order problems. Accordingly, once we can do a GIS mapping of all the properties, the tax administration is reduced to a simple process of tagging and matching. This assumes that it is easier (than doing the same manually) to do GIS mapping and generate analytics, and then act on them to expand tax base and improve collection efficiency.

What if the implementation process of GIS mapping itself can be interminable (how many cities have completed even one iteration of city-wide GIS property mapping)? What if the process of constant updation of the GIS database can be a challenge (this assumes that the Town Planning guys have all the information available on the changes made to houses and new construction)? What if getting the tax collectors to act on the analytics may prove insurmountable (do we seriously believe that the bill collectors in their area or the Commissioner for the city as a whole already does not know about who are the big defaulters)? These, and many more, second and third order issues remain far from out thoughts when we support such interventions.

The neatness and simplicity of the partial equilibrium of the interventions, and the false sense of comfort that it provides, blinds us to the general equilibrium dynamics that are invariably generated and should necessarily be overcome for any such intervention to succeed.

3. Convenience bias. All of us are primed towards embracing something that appeals as neat and simple, and one which increases our convenience. For sure, GIS mapping and regulatory reforms appear very neat and simple, and are better than the existing systems to tackle the problem being addressed. It has always been the case that we demand "more and better, even when less is enough".

4. Optimism bias. It is always the danger that project teams under-estimate the magnitude of the task entrusted and see the road ahead with more optimism than it should merit. Accordingly, officials who champion a technology or process intervention are instinctively likely to under-estimate the problems and over-estimate their chances of success. 

5. Doing-something-new (or innovation) bias. When something is persistently wrong or a failure, we tend to over-react and assume that the existing design and processes have failed and we need to adopt something new. We have deeply internalized that failures are due to lack of innovation with design, process, and technology. Very rarely do we step-back to see whether the original design and processes themselves were rigorously implemented or not. It is very comforting to rationalize away failures by blaming it on the design and other extraneous factors, rather than questioning our implementation capability.

6. Best-practice bias. Occasionally, amidst the gloom surrounding the implementation canvas, we see bright-spots and embrace them as best-practice models to be transplanted across the world. Accordingly, the best-practice regulatory architecture is transplanted into a system which neither has the resources nor the pre-requisite environmental conditions for the effective implementation of the best-practice model. Little do we try to examine whether the bright spot was due to the extraordinary personal initiative of a committed individual or group of individuals or due to some systemically built (and therefore replicable) capacity. 

7. Finally, the illusion of control bias. No policy maker or political leader wants to face up to the reality that their primary implementation instrument, the bureaucracy, in its present condition, is just not capable enough to implement the proposed intervention, leave aside achieve the desired outcomes. Once this assumption is shaken, it can be a very unsettling process for bureaucrats and politicians to craft a policy intervention and its implementation plan. In fact, unlike earlier, when you only had to design a policy intervention (a best-practice model), now you have to also craft an implementation plan. Worse-still, you need to tailor the policy intervention, conditional on your implementation capability. Apart from being personally unsettling, the challenge associated is, in any case, now far more complex than before.

It is more likely that all of them bind in varying degrees nudging policy makers and political leaders to under-estimate state capability problems.