‘Learning’ in public policymaking
We seem to have a scenario where fallible
technology is being pushed through by the government
In theory, the recent policy announcement regarding the
direct cash transfers is a great example of implementing a learning
agenda that would follow a sequence of steps—an ambitious technology
platform is used to announce a roll-out, the government chooses pilot
districts and then proposes to roll out the programme to the rest of the
country. The reality however, is far less encouraging, with the policy
announcement seemingly rushed through with an eye on the 2014 elections.
There is absolutely nothing wrong in governments attempting to win
elections on the back of effective policymaking and to suggest otherwise
or to label cash transfers as ‘bribes’ is just plain stupid. However,
unfortunately we seem to have a scenario where fallible technology is
being pushed through by a government that appears not only muddled
conceptually, but also one that has not thought through the entire
implementation chain.
A genuine attempt at implementing a systematic learning
approach would have taken care of most of these fears and also provided a
platform for genuine improvisation in the use of technology in ensuring
security and reliability at scale. The example of MGNREGA, often quoted
by the minister for rural development, Jairam
Ramesh, as another instance where the government followed a
phased roll-out, starting with 200 districts in the country, is not an
encouraging precedent. Seven years into the scheme, the MoRD’s Sameeksha
report (2012) reveals that the average person-days per employed
household has crossed 50 only once inspite of widespread distress in
rural areas. Also, ensuring full and timely payments continues to be a
challenge, demonstrating how last mile connectivity issues are yet to be
sorted out in the implementation of government schemes.
The question that arises then is the following—what needs
to be done to institute a learning cycle for policymaking? This is a
question relevant for almost all kinds of policy decisions and
legislations. I have, in previous columns, highlighted the importance of
a monitoring and evaluation framework for organizational learning. A
combination of concurrent process evaluations and ex-post evaluations
can yield significant dividends for public agencies. A recent paper by
Lant Pritchett, Salimah Samji and Jeffrey Hammer (Pritchett et al) for UNU-WIDER, draws on existing
evaluation techniques and calls for organizations to actively seek
‘experiential learning’. This would be a structured system of iterative
learning through design variations in implementation.
It is easy to see how this can be used in the case of
cash transfers. Various questions regarding the efficacy of alternate
implementation strategies can be tested and lessons incorporated in the
programme design. For instance, it is unclear at this point how cash (or
benefits, as the government now wants to call them) will actually reach
the hands of the poor. How should banking correspondents be
incentivized and regulated so that they may succeed where post-offices
have failed before? The ministry of finance could work with its
counterparts in the states and with banks to initiate pilots that test
design variations. This will of course require greater coordination with
a wider set of stakeholders at the design stage, but will yield a rich
evidence-base that would immediately place direct cash transfers on a
firmer footing. Demonstrating that the government is prepared to discard
what does not work and enhance what seems to work could even convert
some of the critics!
Could this also have been an opportunity to experiment
with mobile money (through cellphones) at a large scale? After all, the
kind of transfers that the government is talking of currently could
benefit immensely from a system that integrates the Aadhar with a mobile
money transfer infrastructure on the lines of Kenya’s highly successful
M-Pesa—and
enables monetary transactions with the local PDS outlet or kirana
shop. There hasn’t been much mention of the 600 million cellphone
subscriptions in this context.