Substack

Tuesday, May 22, 2012

Empowering civil servants in using information

I have an op-ed in Mint today that advocates the use of data visualization techniques to bridge the last mile gap with effective utilization of the rich trove of information within public systems.

Monday, May 21, 2012

Do higher wages reduce supervision costs?

Freakonomics points to the example of Hungarian musician entreprenuer Gabor Varszegi who made millions by providing high-quality service in his photo developing shops in Budapest by hiring workers and paying them wages that were four times the going rate. It quotes Hal Varian on the reason behind this apparently strange wage payments,
There are very few employees per store and monitoring their behavior is very costly.  If there were only a small penalty to being fired, there would be great temptation to slack off. By paying the workers much more than they could get elsewhere, Varszegi makes it very costly for them to be fired — and reduces his monitoring costs significantly.
In other words, higher wages act as a deterrent against shirking and working sub-optimally. Consequently, the need for more rigorous monitoring is less and productivity and retention increases. However, the success of this strategy lies in the threat of being immediately fired if they shirk. Remove this threat and the model collapses.

Public bureaucracies are the best example. Salaries at the lower levels of the permanent bureaucracy in India are very high, several times higher than the going market rate for similar work in the private sector. So, the Varszegi model should have made these workers more productive and reduced supervision costs. But a very strong countervailing force, in the form of (formal and informal) security of tenure or trade union activism, works against and virtually depletes the threat of being fired. The incentives get immediately misaligned.

So can this approach work with the contract employees in the bureaucracy. Thanks to minimum wages, they too get wages that are higher than the prevailing market rates. Theoretically, removing them is easy since the contracts can be easily terminated. And this is the case when the numbers involved are smaller. However, when such contract employees swell in number and have worked for a few years, the moral hazard resurfaces. They form unions and the lines of distinction with the regular employees, atleast to the extent of their labour rights, gets blurred. The threat of being fired gets diminished.  

In other words, the moral hazard arising from being part of the regular public bureaucracy or being part of large enough contract labour force who have been working for sometime nullifies the positive effect of the higher wages. As I blogged earlier, this is yet another example of how the emergent dynamics of human responses, under certain conditions, undermines logically sound incentive systems.  

Saturday, May 19, 2012

Interpreting the US imposition of anti-dumping duties on Chinese solar imports

The United States Commerce Department has finally slapped anti-dumping tariffs of more than 31% on crystalline silicon photovoltaic cells and their modules imported from China. The decision, one of the largest in US history and certain to generate Chinese retaliation, covers one of the fastest growing categories of Chinese imports worth around $3.1 bn last year.

The Commerce Department was acting on a complaint filed by a domestic solar panel manufacturer, SolarWorld Industries America Inc. The anti-dumpring action covers not only imports of solar cells produced in China and solar modules/panels produced in China from Chinese-made solar cells, but also imports of solar modules/panels produced outside of China from solar cells produced in China.

The 'Agreement on Implementation of Article VI of the General Agreement on Tariffs and Trade 1994' outlines the conditions under which anti-dumping (AD) action can be initiated. It provides for imposition of  AD duties when it is established that exporters are selling their products at less than the "normal value" and the sales of the dumped product is causing material injury to a domestic industry that produces a like product. In other words, to attract AD action, the exporter should be dumping and this should in turn be causing injury.

The aforementioned agreement provides for three methods to calculate a product’s “normal value”. The main one is based on the price in the exporter’s domestic market. Generally, an exporter attracts AD action when the export is priced lower than the price normally charged in its own home market. When this cannot be used, two alternatives are available—the price charged by the exporter in another country, or a calculation based on the combination of the exporter’s production costs, other expenses and normal profit margins.

In the instant case, the Commerce Department calculated the 31% tariff by estimating Chinese manufacturers’ costs and then determining how far below cost the solar panels were being sold in the United States. But since China is designated a non-market economy (where the government plays such a large role in allocating land, credit and other resources that the true costs of any given product may not be apparent), the Department used manufacturing costs in Thailand as a proxy for costs in China. The Chinese are contesting this, arguing that India is a more accurate representative.

Since the Chinese are certain to appeal against this decision, sustaining the case before the WTO's dispute settlement mechanism is going to be widely watched. Of particular interest will be whether the apellate body agrees with the method adopted to establish normal value. If it is upheld by the WTO, it is certain to open the floodgates for similar action by other countries against China. India too will be watching the outcome of the appeal with interest. 

In any case, as I have blogged earlier, the provisions of the WTO’s Article III: 4 of Trade Related Investment Measures (TRIMS) and General Agreement on Tariffs and Trade (GATT) III allows for payments of subsidies to domestic producers and consumers. This exemption given to subsidizing domestic producers mean that the Chinese government's subsidies, direct and indirect, do not infringe the provisions of WTO. It is widely accepted that the Chinese firms benefit from huge subsidies by way of cash grants, raw-materials discounts, preferential loans, tax incentives and cheaper currency. In 2010 alone, Chinese Development Bank gave $30 billion in low-cost loans to top five domestic solar panel manufacturers.

Critics of the AD action argue that the cheap Chinese imports are a massive subsidy transfer to American consumers from the Chinese government. They point to the boom in demand for off-grid roof-top solar panel installations across the US. Further, they argue that Chinese companies often turn to American companies to buy the factory equipment and polysilicon they need to make solar panels. All this, it is claimed, benefits the American economy. And Mathew Kahn makes the important point that the positive externality (in terms of lowering greenhouse gas emissions by increasing the share of solar power) created by cheap solar panel imports means that they should be promoted and not discouraged with dumping.

Friday, May 18, 2012

More on the contribution of RCTs

A recent working paper of a randomized control trial (RCT) in Orissa to test the effectiveness of improved/clean cooking stoves in reducing indoor air pollution and fuel consumption found that the intervention had no impact. The study which tracked 2600 randomly selected households who were provided the improved stove, over a four year period, finds that,
While we find a meaningful reduction in smoke inhalation in the first year, there is no effect over longer time horizons. We find no evidence of improvements in lung functioning or health and there is no change in fuel consumption (and presumably greenhouse gas emissions). The difference between the laboratory and field findings appear to result from households’ revealed low valuation of the stoves. Households failed to use the stoves regularly or appropriately, did not make the necessary investments to maintain them properly, and usage rates ultimately declined further over time. 
Personally, I find nothing surprising about the study. India's social policy space is replete with hundreds of such examples of failed interventions. Logically and theoretically sound interventions, end up with disappointing results when subjected to the real world implementation test. A few days back, I posted about the failed experiment with automatic parking meters in Hyderabad.

There are two issues here. The first is a technical/technological question - Do improved or clean cooking stoves optimize on fuel consumption and reduce indoor air pollution? The second is a sociological/socio-economic question - Assuming a positive answer to the first question, how effective will be the cooking stove in realizing the two objectives when used in the real-world? More specifically, how would users and other stakeholders respond to the new stove?

Most often, when public policy is designed, the first question is scrutinized pretty rigorously while the second is assumed away. But, it is here, in the second question, that these interventions end up failing. Rightly, the working paper highlights this issue
More broadly, this study underscores the need to test environmental and health technologies in real-world settings where behavior may temper impacts, and to test them over a long enough horizon to understand how this behavioral effect evolves over time.
It is not just with technology interventions that these issues rise to the fore. Policies involving regulatory restraints, incentives, restructuring/reframing of the environment, computerization, and so on, in all sectors suffer from these implementation challenges. Stakeholders' response patterns to an intervention cannot be fully anticipated. The emergent dynamics often detracts from the desired outcomes.     

Therefore I believe that more than providing a new methodology to evaluate the impact of interventions, the more important contribution of the randomista movement to development economics may be to draw attention to the importance of human and social dynamics in determining the success of those interventions.

Thursday, May 17, 2012

Designing policy by building a distribution of results

Jim Manzi has an excellent post in The Atlantic where he examines the fundamental problem in social sciences, "How do we know that what we are doing is right?" His illustration of the complexity of drawing causal relationships in social sciences is spot-on,
We can run a clinical trial in Norfolk, Virginia, and conclude with tolerable reliability that "Vaccine X prevents disease Y." We can't conclude that if literacy program X works in Norfolk, then it will work everywhere. The real predictive rule is usually closer to something like "Literacy program X is effective for children in urban areas, and who have the following range of incomes and prior test scores, when the following alternatives are not available in the school district, and the teachers have the following qualifications, and overall economic conditions in the district are within the following range." And by the way, even this predictive rule stops working ten years from now, when different background conditions obtain in the society.
He suggests that the only way to increase the probability of success with such predictive rules that emerge from an RCT is to build as large an array of results of the same phenomenon, both from experimental and non-experimental studies, as possible and then use them to refine the original finding.
What we really need to do is to build a distribution of results of "experiments + model" in predicting the results of future experiments... We can then then compare the accuracy of such a theory this to analogous distributions of predictions made by non-experimental methods (that can vary from sophisticated regression models to newer machine learning techniques to prediction markets to the judgments of experts, and so on) for predicting the results of future experiments... 


Even if I have such a distribution of results for the predictions made by various methods, I can't ever be absolutely certain that this distribution won't suddenly change... But I think this is as close as you can get. What this demands, of course, is a lot of experiments. This is why lowering the cost per test is so critical. Not just as an efficiency measure, but because in practice in enables me to get to much more reliable predictions of the effects of my proposed interventions.
I have no problem with Manzi's suggestion to use a distribution of results to increase the predictive power of any social or public policy intervention. But it has important implications, especially given the time and resources that would have to be expended to acquire such a rich distribution of results for each issue.

Since interventions vary widely in complexity, there is a need to define some standards for collecting the distribution of results. Certain interventions like the impact of deworming or default savings accounts may not need the exhaustive collection of examples to draw fairly reliable predictions or conclusions. But certain others, like a particular type of classroom instruction model or a specific performance incentive to teachers, may need results from a wide diversity of field conditions, to establish their robustness. However, for a number of interventions, we may never be able to draw reliable enough predictive inferences even from a large distribution of results.

The collection of a number of results of all types from varied social and political environments will help tease out policy design elements that illuminate both successes and failures. Any policy tailored based on this collection of results will stand a greater likelihood of success. 

I believe that the development of an analytical framework to triage interventions and thereby more effectively use experimental and non-experimental data to design social policy interventions should be the next most important area of research in the years ahead.

Wednesday, May 16, 2012

Jaime Dimon and "regulatory capture"

Amidst all the discussion surrounding JP Morgan Chase's bet on corporate debt derivatives going terribly wrong, this from Eliot Spitzer, highlights the severe structural problems in Wall Street. The presence of JP Morgan Chairman Jamie Dimon - who lobbied aggressively to redefine the meaning of hedging under the Volcker Rule that was proposed to curb proprietary trading by deposit taking commercial institutions (banks) - on the board of the New York Federal Reserve should constitute as one of the most egregious and shameful examples of "regulatory capture",
Jamie Dimon sits on the board of the New York Federal Reserve Bank - the very organization that is supposed to oversee his bank’s financial practices, the organization that is supposed to issue all sorts of regulations that control what his bank can do, the very organization he has been lobbying to relax the rules about the bets he wants to make...
The Fed conflict is so obvious that it defies any possible rationalization or explanation. For a decade, the New York Fed has failed to pick up on any of the significant Wall Street threats:  excess leverage, subprime fraud, dangerous concentration in “too big to fail” entities.  Maybe the reason is that the board is controlled by the very voices that have been at the root of the failure. There has been not the slightest voice of protest from the board—yet it is a public organization!
By any yardstick of propriety, leave alone decency, Jamie Dimon and all others similarly placed should recuse themselves from the NY FRB Board. See this, this, this, this, this, and this about how JP Morgan's corporate debt derivatives bet went wrong. See also this on why the New York Fed is rife with conflicts of interest and incentive distortions.

Update 1 (7/7/2012)

The scandal that erupted in the aftermath of revealations that Barclays rigged its interest rate estimates used to calculate LIBOR once again spotlights attention at the rotten core of global banking. 

The London Interbank Offer Rate (LIBOR), which is a measure of the rate at which banks borrow money in the money markets, is released each day by the British Bankers’ Association (BBA), covering 10 currencies and 15 different maturities. The Libor rate is used to set rates for some $800 trillion in global financial transactions from derivatives to consumer lending. More than $10 trillion in loans to businesses and consumers have interest rates based on Libor, with the borrowers paying more or less depending on where Libor is on a particular day. Many derivative contracts are also based on it. The BBA collects the rates at which each bank says it could have borrowed money that day without putting up any security for the loan. The highest and lowest responses are thrown out, and the rest are averaged to produce a number. There is no need for any bank to show that it actually borrowed at the rate it claims.
Barclays rigged the LIBOR by manipulating its estimates to suit its interests. During the financial crisis, it low balled its estimate of borrowing rates so as to reassure everyone of its creditworthiness. At other times, when its trading positions called for lower rates, it lowballed its estimates even if it reduced the interest income received by the bank. The FT was scathing in its verdict,
The bankers involved have betrayed an important public trust – that of keeping an accurate public record of the key market rates that are used to value contracts worth trillions of dollars. They did this to make money and to conceal from the wider world their true cost of borrowing. This was market-rigging on a grand scale. It is hard to think of anything more damning – or more corrosive of the reputation of capitalism.
The always incisive Simon Johnson points to Dennis Kelleher of Better Markets, who has a devastating critique of the major banks that have been colluding to rig the LIBOR,
The Libor interest rate is based on a survey of banks like Barclays.  Those banks know what that the information they provide in that survey sets the Libor rate, which is then used to set the rate for those $800 trillion-plus transactions. What Barclays settled - and what the other banks are being investigated and sued for - is knowingly and intentionally providing false information that they knew would result in a false Libor rate being set. This is not some isolated sales practice or trading strategy. The allegations are of a massive conspiracy involving 20 or so of the biggest banks in the world manipulating one of the most important rates in the world. So this isn't about Libor - this is about Lie-More.

That seems to be the business model for the big global finance houses.  They like to call themselves "banks," but they aren't banks in any traditional sense. They are global behemoths that are not just too-big-to-fail, but also too-big-to-regulate and too-big-to-manage. Take JP Morgan Chase for example. It has a $2.35 trillion balance sheet, more than 270,000 employees worldwide, thousands of legal entities, 554 subsidiaries and, as proved by the recent trading losses in London, a CEO, CFO and management team that has no idea what is going on in their own bank.




Tuesday, May 15, 2012

Limits of technology in public policy interventions

We consistently underestimate the last mile gaps that are pervasive in society. Despite several ubiquitous examples of failures, we mistakenly continue to believe that technology fixes or regulatory diktats will help address social policy issues.   

Sample the fate of the parking ticket vending machines installed at a few locations in Hyderabad city,
The much-talked about parking meters installed in the State capital to automate the disbursement of parking tickets is proving to be futile as the vehicle-owners are not evincing any interest in utilising the facility... But due to the lack of public participation... (the company) which maintains the machines, is now forced to employ contract workers to issue the tickets... a motorist has to go and take the parking token. But none of them prefer to do it... Motorists are used to paying money and obtaining the tickets and so they are reluctant to walk down to the parking meters.
It is obvious that the opportunity cost of walking across to the parking meter, dropping the coins and generating a parking ticket is too high. There are two elements to this cost. One is the deeply internalized behavioural inertia against this additional exercise, especially when motorists are so used to getting tickets from parking attendants.

Two is the minimal cost associated with not generating the parking ticket. Without strict supervision, motorists would simply drive off without making the payment. However, if supervisors are appointed and monitoring is made rigorous, then (given the numbers required to supervise even small stretches) the cost becomes prohibitive and defeats the purpose of installing such devices. We could as well have a more efficiently run outsourced parking attendants (equipped with electronic billing machines) based system. And there is always the danger that the supervisors will themselves start collecting illegal parking rents.

Lowering the opportunity cost would involve inculcating a sense of civic responsibility among motorists and also increasing the cost of non-compliance (by prohibitive fines or vehicle tow-aways). But both these solutions run into problems when examined from the lens of real-world implementation. The first takes a long-period of continuous use of such parking meters, apart from issues related to general socio-economic development. The second poses significant enforcement challenges and compliance costs.

So, at this stage of the country's development, except for a few locations (where civic responsiveness is likely to be higher), the traditional attendants based parking regulation appear the more realistic proposition.The availability of cheap labour only serves to increase its attractiveness.