30 August 2016

Social Policy Bonds as a meta-system

I haven't read The Moral Economy: Why good incentives are no substitute for good citizens, by Samuel Bowles, but I have read reviews. Mr Bowles points out that, in some circumstances, monetary incentives alone cannot make people behave as we should wish and can even encourage perverse behaviour. This echoes the work of Professor Bruno Frey who found that monetary incentives can undermine our willingness to do the right things for ethical and moral reasons. People perform valuable social or environmental services not only for monetary gain, but also because they enjoy doing them for their own sake, because they believe them to be the morally right things to do, or because they believe that their actions will advance some cause to which they are committed. These ‘intrinsic’ motives are qualitatively different from external, monetary incentives, and offering monetary rewards might ‘crowd out’ or undermine these less mercenary and more civic-minded motivations. Crowding out internal motivation can occur, writes Prof Frey, because, monetary incentives can undermine people’s feelings of self-determination and self-esteem. Also, when external incentives are supplied, the ‘person acting on the basis of his or her intrinsic motivation is deprived of the chance to exhibit this intrinsic motivation to other persons.’

Not mentioned by Frey, but also plausible is that financial incentives can undermine the cognitive outlook that sees socially and environmentally beneficial services as worthwhile in their own right, rather than as a cost for which compensation and payments must be paid by taxpayers.

What do these findings mean for Social Policy Bonds, which at first sight seem to be entirely dependent on monetary incentives that will encourage achievement of socially desirable goals? First, it's important to note that, as Frey points out, the crowding-out effects are not always significant. In markets, which are based on relationships amongst essentially self-interested strangers, financial incentives as exhibited through the price effect do work as classical economics predicts. That is, they work to increase supply. And when (as they would be under a Social Policy Bond regime) external rewards are seen as correlated with civic duty rather than an attempt to ‘buy’ one’s civic performance, they may well support, rather than undermine, moral and other intrinsic motivations. This would be especially true if, partly because of the role that a bond regime could play in raising taxes on less socially valuable forms of wealth accumulation.

Second: Social Policy Bonds are not merely a system by which monetary incentives are funneled into the most efficient providers of public goods and services, but a 'meta-system' that motivates bondholders to find the best ways of encouraging socially beneficial behaviour - whether these be monetary or not. A bond regime could give bondholders incentives to explore further the insights of Mr Bowles and Prof Frey, looking in detail at the relationships between financial incentives and civic performance. They could use this knowledge to minimise the costs of achieving targeted objectives by, for example, finding out when monetary incentives are least likely to supplant the intrinsic motivations of people who help achieve objectives, and concentrating their use in those circumstances.

Social Policy Bonds, then, are not a relatively crude financial instrument that rewards payment for performance in a narrow sense (that would be Social Impact Bonds), but rather a way of rewarding people for encouraging socially beneficial behaviour, however they do so.

27 August 2016

Make them tradeable

Quite a bit going on with Social Impact Bonds, as is apparent from their Wikipedia page and the almost daily Social Impact Bond newsletter; one recent issue highlighting Japan's interest in the concept. I do have reservations about SIBs, which I have written about here and here, though they do seem to have been inspired by my early work on Social Policy Bonds. One weakness, in my view, is that because they are not tradeable, SIBs are more subject to gaming and manipulation than Social Policy Bonds. This could become a bigger problem if the SIB concept becomes so widely adopted that they avoid public scrutiny. There is a long and sorry history of (presumably) well-intentioned measures, ostensibly taken to boost efficiency in the public sector, ending up as subsidies to the already wealthy and powerful. See, for instance, this piece about the UK's Private Finance Initiative. This is corporate welfare wearing a thin disguise.

So, given my doubts about SIBs, do I have anything positive to say about them? Yes:

First, is that they aim to reward better performance in the provision of social services. True, their lack of tradeability drastically narrows the range of such services and severely restricts their applicability over space and, especially, time. Still, they do give incentives to service providers to do a better job, given such limitations: something that is more revolutionary than it should be, but nevertheless a definite step forward.

Second, whatever their weaknesses, SIBs might be an improvement in policy areas that are particularly poorly served by existing interventions. Such unglamorous policy areas as, for instance, provision of services to the mentally unwell or to newly-released prisoners to help prevent them from recidivism.

And third, of course, SIBs might serve as a helpful or necessary transitional step toward Social Policy Bonds. SIBs allow the bonds' issuers, whether public- or private-sector a greater degree of control than Social Policy Bonds over who shall go about achieving the bonds' objectives and (less directly) how they shall do so. Government and other backers of bonds are, understandably, reluctant to relinquish their control over these levers of power. But Social Policy Bonds could be introduced gradually, and government agencies - if they are truly efficient - need not fear their introduction. Exposure to competition from others motivated to achieve society's goals, as targeted by a Social Policy Bond regime, would stimulate the exploration and implementation of diverse, adaptive solutions to national and even global problems. If it takes a decade or two's experimentation with SIBs to get there, it's a worthwhile journey. My hope is that SIBs' weaknesses and the greater scope they give for manipulation do not falsely discredit the Social Policy Bond idea in the eyes of the public.  

For more about a transition to a Social Policy Bond regime, see chapter 3 of my book.

22 August 2016

Alternative data and metrics for a bond regime

The Economist writes about 'alternative data':
The growth of small, low-cost satellites and machine learning means companies can quickly and cheaply parse millions of satellite images a day. A common trick is to analyse photos of car parks outside big-box retailers such as Walmart to get a sense of daily revenues. A Chicago-based data firm, RS Metrics, sells estimates on the productivity of factories by tracking the number of lorries parked outside. The watchers, the 'Economist', 20 August
Mindful of Campbell's Law, I've always thought that each Social Policy Bond issue should target a set of metrics, each of which has to fall within a specified range before the bonds can be redeemed. Also, that, if we target a broad national goal (universal literacy, say) the bond's issuers could stipulate that the bonds would be redeemed on the basis of data from a random sample of people of the country, so as to minimize the risks of manipulation. Alternative data could contribute to robust combinations of metrics for the purposes of a bond regime, especially those covering countries where official data is scarce or unreliable. As the article says: "Investors are particularly keen for firms to study pictures that yield rare data on, say, steel production in China or Russia, where official data can be patchy." It's not difficult to imagine scenarios in which alternative data could play a possibly indispensable role in the monitoring of progress toward environmental goals, or goals such as disaster prevention or conflict reduction.

20 August 2016

Poverty: the need for diverse, adaptive approaches

The Economist writes about US President Clinton's 1996 welfare reform package. Under the sub-head Blockheads:
Challenged to reduce the number of people receiving welfare, many states merely shifted people onto disability insurance instead, declared victory and sent the bill to Congress.... How might the reform be reformed? Most vitally, by concentrating attention and resources on those 1.5m families at the very bottom. Since this is the hardest group to reach, the federal government should use its money to encourage states to find new ways to help them. A patchy record at 20, the 'Economist', 20 August
Quite so. Clarity about aims is an essential and inescapable first step in implementing a Social Policy Bond regime. Unfortunately, policymakers under the current system can get away - or get rewarded for - with shifting people from 'receiving welfare' to receiving 'disability insurance'.

The article continues:
A useful model is “Race to the Top”, an education initiative from the Obama administration which rewards states that achieve improvements with extra money, in the hope that others will copy their success. There are plenty of policies worth experimenting with: expanding tax credits for those without children, extra government help with finding a job and even public make-work schemes. But this must be experimentation with the right purpose—helping the poorest into work rather than simply cutting welfare rolls.
True: the aim is not to cut the number of people on welfare. But I question whether raising the number of poor people in work is exactly what we want to achieve. I would think our over-arching goal is to eradicate poverty over a sustained period. Increasing employment among those currently poor may be one way of doing that, but we should not assume that it is the most efficient way. Nor the most compassionate: it's not difficult to think of people for whom employment would be less helpful than other interventions. For instance, a struggling single parent of several small children could benefit more from, for instance, help with childcare. Society as a whole would benefit more, in the long term, by improving the children's education and healthcare or their physical and social environment, or improving the parent's access to information about how best to nurture children. If parents were compelled to work, the benefits of a higher household income could be outweighed by the negative effects on the children.

As Barbara Ehrenreich put it:
The "working poor" ...  neglect their own children so that the children of others will be cared for; they live in substandard housing so that other homes will be shiny and perfect; they endure privation so that inflation will be low and stock prices high. To be a member of the working poor is to be an anonymous donor, a nameless benefactor, to everyone. Source
Social Policy Bonds targeting poverty could encourage the exploration and implementation of whichever approaches would best suit the varied and ever-changing circumstances of a population. Every one of those 1.5 million worst-off families referred to in the first excerpt above will face different challenges. Employment will be a solution for some, but not all. Social Policy Bonds would motivate people to find the diverse, adapative solutions that extreme poverty and many other social and environmental problems require if they are to be solved, rather than merely disguised.

14 August 2016

Social Policy Bonds: state of play

I don't think any Social Policy Bonds have yet been issued, despite their having been in the public arena for 28 years. That said, more national and local entities are issuing or considering Social Impact Bonds, the non-tradable variant of Social Policy Bonds. This wikipedia page summarises the history and current deployment of SIBs. I'm pleased to say that they are becoming more widespread. Countries in which they have been issued include the UK, Australia, the US, and they are also being considered in Brazil, Israel and New Zealand. Dan Corry of New Philanthropy Capital in London summarises the current (9 August) state of play with SIBs in the UK here, while Patrick Young, in Australia puts out the Daily SIB Newsletter.

I do have reservations about SIBs, which I have expressed here, here, and in several blog posts (search this page for Social Impact Bonds). Perhaps necessarily, they are narrow in scope and, in my view, will be prone to manipulation and gaming, especially if they become so commonplace that they escape public scrutiny. Because of their limitations they are also, as I expected, relatively costly to administer, as Alliance 54 reported in July: 'SIBs are gaining traction with 57 models operating, but they have proven complicated and costly to design and implement.' I haven't been consulted about, and have no involvement in, any of these SIB issues. My hope is that SIBs, will advance, rather than discredit, the Social Policy Bond concept.

07 August 2016

Health funding needs to be more impartial

Priorities for healthcare funding are heavily influenced by factors other than the ratio of benefit to cost.
[A] review by Cancer Australia (pdf) showed that between 2006 and 2011, breast, colon and prostate cancer all received funding greater than their proportional toll on society — measured in years of healthy life lost. By the same measure, research into lung cancer, along with lymphoma, pancreatic cancer and cancer of the brain were underfunded. Lung cancer research underfunded compared to societal impact, ABC news, 12 August 2015
In this graph, taken from the above link, 'DALYs' on the vertical axis means disability adjusted life years lost to the type of cancer on the horizontal axis. Funding is given in millions of Australian dollars.
 Research funding for different types of cancer compared to healthy life lost
Some disparities are striking: "Lung cancer, which takes the heaviest toll on years of healthy life, received less than a quarter of the funding given to breast and colon cancer research."

Governments have to make their resource allocation decisions on the basis of data that are necessarily incomplete and constantly changing. So, by default, health expenditure is influenced by groups of medical specialists with little incentive or capacity to see improvements in the general health of the nation as an objective. As a result, funding of these specialities depends to a great and varying extent, on the strength of their lobby groups or on their public profile rather than on what would best meet the needs of society.

The problem is the same sort of top-down, one-size-fits-all, fossilised systems of funding that bedevil other (well-meaning) attempts by government, or any large organisation, to keep track of multiple variables across any but the smallest geographic area. In health, as in education, housing, crime prevention, or environmental pollution we need diverse, adaptive approaches to solving our problems. Society is just too complex now for simple approaches to work effectively, except in those increasingly rare cases were cause and effect can be readily identified and relatively stable over time and space.

The Social Policy Bond principle can be applied to health. Essentially, under a bond regime, government would target for improvement the health of the entire population as measured by (probably) DALYs in combination with other measures. Resources woudl then be allocated impartially according to where they will yield the most benefit per dollar spent. Any target could be long term: if it were several decades bondholders would have an incentive to investigate numerous approaches, including preventive measures, research and education, on a dynamic basis and always with an eye to cost-effectiveness. For more on this, see my brief piece on Health Bonds.

03 August 2016

Government of the people, by the rich, for the rich

Elizabeth Drew quotes John Tanner, a former US congressman, explaining extremist attitudes in US primary elections:
If a Republican strays from the far ideological right ... they put themselves in political peril. They know better—some of them—but it’s not worth the political fallout to wander into the sensible center and try to sit down and work something out. No one will do what they all know has to be done to keep the country from going adrift. (My emphasis.) American democracy betrayed, Elizabeth Drew, 'New York Review of Books', dated 18 August
I've always maintained that there is more consensus about our social goals than there is on how to achieve them. Even so, as Mr Tanner indicates, there is also wide agreement over what needs to be done: that is, over activities. And yet the way we conduct our politics makes us incapable even of undertaking those activities. Divisiveness, which in its less virulent form may on balance have served us well over history, has become toxic. In the US particularly, it has become self-entrenching, through cynical manipulation of constituency boundaries, as in this example given by Ms Drew in the same article:
[A] a tiny peninsula was added at the last minute to a new district in northeast Ohio not because it contained certain residents but because it was the site of a large manufacturing company that could produce campaign contributions.
To put it bluntly, our system has been hijacked by experts who have little interest in the well-being of broader society. The result is becoming clear to us all: a wide and widening gap between government and big business on the one hand, and ordinary people and small business on the other. The political systems, not only of the US, but in all countries, have become too complex for ordinary people to follow. The dangers are becoming apparent: growing cynicism, extremism and nihilism.

To close the gap between people and their supposed representatives, I offer Social Policy Bonds. They aim to inject the market's efficiencies into the achievement of our social and environmental goals. But perhaps even more important, they demand clarity of these goals. Meaningful, explicit goals, which all of us can understand. Goals such as universal literacy, improved health, reduced crime and pollution and, at an international level, peace. With greater comprehension of our politics, we'll see more public engagement, and hence more buy-in: an end in itself, as well as a means towards more efficient ways of solving our social problems. What we have now is a travesty: government of the people, by the rich, for the rich. I don't think it's sustainable. By switching deliberately and slowly towards something like a Social Policy Bond regime we could ensure that, instead of a likely collapse followed by anarchy or tyranny, we'd see politics geared towards the well-being of all society and not, as now, those who are powerful enough to manipulate the system.