22 April 2017

Reinforcement learning and the public sector

Will Knight writes about reinforcement learning:
Reinforcement learning is a way of making a computer learn through experience to make a series of decisions that yield positive outcomes—even without any prior knowledge of how its actions will affect its immediate environment. A software-based tutor, for example, would alter its activities in response to how students perform on tests after using it. Reinforcement Learning, Will Knight, 'MIT Technology Review', March/April
It's the same approach practised by the computer called DeepMind, which 'mastered the impossibly complex board game Go and beat one of the best human players in the world in a high-profile match last year.' It's also, at least in theory, the basis of our economic system: the learning process that is supposed to go on through the creative destruction of business enterprises that fail to adapt to changing circumstances. As applied to driverless cars, 'the software governing the cars’ behavior wasn’t programmed in the conventional sense at all.' The software learns through trial and error.

Large corporations and their friends in government have seen to it that competitive markets play a lesser role than they should in determining our economic prospects. Corporations succeed today not so much through competing in markets, but more by manipulating the legislative and regulatory environment, and participating in effective cartels with other powerful players, both private- and public-sector. But the principle - that of survival of the most adapted, and the most quick to adapt - has generated enormous material wealth, and continues to operate, though more to the benefit of the already-wealthy than to ordinary people.

With its record of success, why isn't reinforcement learning, or competition to find the best approaches, allowed to operate in the public sector. That is, why don't we allow competition to solve our social problems? It's partly for  historical reasons. When governments began to take an interest in the welfare of ordinary people, solutions to their basic problems were fairly easy to identify. Requirements for things like sanitation, elementary education, a police force, fire engines and hospitals were - and are - difficult to argue against. But as society has grown more complex so too have our social and environmental problems. How, for example, do we go about reducing crime, improving our (already relatively high) levels of physical health, improving our mental well-being, or ending war? No obvious solutions leap to mind.

Which is where Social Policy Bonds could enter the picture. Rather than let public-sector organisations have a monopoly on trying to deal with these problems, a bond regime would, in effect, contract out finding the best solutions to the private sector. A goal such as eliminating war is going to need the exploration, deployment and refinement of a multitude of potential approaches. More than that, though, it's going to need to reward the most effective of these approaches and to terminate the useless ones. There are well-meaning organisations working to end war, but the people working for them are not rewarded for their success. There are no built-in incentives for the organisations to find optimal solutions, nor for inefficient organisations to be dissolved if their efforts prove futile or counterproductive. The result is that the challenges humanity faces, including environmental calamities or nuclear proliferation, are nowhere near being effectively met. But, unlike computers learning how to drive cars, or businesses operating in competitive markets, our failed approaches and ineffectual organisations aren't terminated. Sometimes, in fact, our politicians pump even more money into them.

Social Policy Bonds would change that. They would channel the market's incentives and efficiencies into the discovery and implementation of diverse, adaptive and, above all, efficient approaches to our social and environmental problems, including those, like war, that many of us have concluded have no solution. Our current system places responsibility for the solution of our problems in the hands of large, usually monopolistic, organisations that face little competition and have no incentive to try diverse approaches. There's certainly no reinforcement learning. It is because of its ability to stimulate diverse, adaptive approaches that a Social Policy Bond regime could succeed where existing policies have failed.

15 April 2017

What are governments for?

George Monbiot quotes from Kate Raworth's Doughnut Economics, saying that:
[E]conomics in the 20th Century “lost the desire to articulate its goals.” It aspired to be a science of human behaviour: a science based on a deeply flawed portrait of humanity. The dominant model – “rational economic man”, self-interested, isolated, calculating – says more about the nature of economists than it does about other humans. The loss of an explicit objective allowed the discipline to be captured by a proxy goal: endless growth. Circle of life, George Monbiot, 13 April
I've inveighed for years against the de facto goal of governments everywhere which, in the absence of clear, explicit goals, has become economic growth, expressed as the rate of increase of Gross Domestic Product, or GDP per capita. The grievous effects of this mistargeting on income and wealth distribution, the environment, and much else, are now becoming apparent. More recently, I've asked What are economists for? Governments as well as economists have lost their way. As Raworth says, they do not think in terms of explicit goals that are meaningful to ordinary people. They are pre-occupied with process and complexity:
The Dodd-Frank bill, like Obamacare, is tyranny by complexity. Consider the Glass-Steagall Act, at 37 pages in length, and the 2,319-page monstrosity of the Dodd-Frank Wall Street Reform and Consumer Protection Act. Charles Hugh Smith, 6 April
The only people who follow and understand our policymaking system are those politicians, bureaucrats, academics, lawyers and lobbyists who receive monetary rewards, sometimes vast, for doing so. The 'rational economic man' just happens to be the person who, in our sad attempts to purchase things that used to be supplied by the commons, maximises economic activity: that is, GDP. I think most people now see the flaws in targeting GDP as if it were an end in itself, but we are less united over what to do about it. My suggestion is that we start to express our policy goals in terms of broad, verifiable, explicit outcomes that are meaningful to ordinary people. Social Policy Bonds could do this, and inject the market's incentives and efficiencies into the achievement of those outcomes.

Under a Social Policy Bond regime, governments could still do what they are good at: raising revenue to achieve our social and environmental goals. And, though it would be a departure for most of them, they could learn to articulate these goals, as expressed and debated by the people they are supposed to represent. They could even, through government-financed bodies, help achieve these goals, but only if, through the Social Policy Bond mechanism, bondholders have confidence in their ability to do so more efficiently than other investors in the bonds. The current model which as Raworth points out, is dominant - that of 'rational economic man' - assumes and entrenches a paradigm that undervalues and often conflicts with community, the environment, small enterprises, and ordinary people's mental well-being. The short-term interests, as measured by accountants, of large organisations, both public- and private-sector, are privileged at the expense of the the rest of us. Social Policy Bonds would change that, right from the start, by posing the question that we cannot continue to evade: what are governments for?

06 April 2017

The European Union: just like every other organisation

From Nigel Farage's recent speech to the European Union Parliament:
Eighty-five percent of the global economy is outside the EU. if you wish to have no deal [with the UK] it is not us that will be hurt. ... A return to tariffs will risk the jobs of hundreds of thousands of people living in the EU. What you are saying is that you want to put the interests of the European Union above that of your citizens and your companies. Nigel Farage, 5 April
 Earlier in the same speech Mr Farage also pointed out that in 1973 the UK voted to stay in the European Economic Community; not a political union. The speech neatly illustrates two principles which I deem axiomatic:
  • Every organisation, be it a church, trade union, university, government or whatever, will always seek to overplay its hand. 
  • Every organisation will, sooner or later, forget its founding ideals and its stated objectives, and devote its energies to self-perpetuation.
In the private sector there is, or should be, the discipline of the market and competition, though we increasingly see larger companies winning out over small businesses not through fair compeition but via manipulation or subversion of the market.

Social Policy Bonds would lead to a new type of organisation: one whose goals would exactly coincide with those of society. That's one feature that would differentiate them from all other organisations in history. The other is that the goals of the organisation would either themselves be, or be inextricably linked to, society's well-being. Every activity of a coalition of holders of Social Policy Bonds would be undertaken with exactly one objective in mind: to maximise the efficiency with which society's targeted social or environmental goal is achieved. The activities, structure and composition of the organisation would be entirely subordinated to that one objective.

The difference between a Social Policy Bond regime, and any other current policymaking system, is stark. Under a bond regime governments would do what they do best (democratic governments anyway): that is, articulate society's goals and raise the revenue for their achievement. Then they'd, in effect, contract out the achievement of these goals to new organisations whose interests are exactly those of society. Their focus, and their striving for efficiency, are built into the Social Policy Bond mechanism. All society would benefit.

My reasons for voting for Brexit are given here. For more on Social Policy Bonds see the overview papers linked to here.

28 March 2017

The corruption of science

John Michael Greer writes:
Especially but not only in those branches of science concerned with medicine, pharmacology, and nutrition, the prostitution of the scientific process by business interests has become an open scandal. When a scientist gets behind a podium and makes a statement about the safety or efficacy of a drug, a medical treatment, or what have you, the first question asked by an ever-increasing number of people outside the scientific community these days is “Who’s paying him?” .... These days, in any field where science comes into contact with serious money, scientific studies are increasingly just another dimension of marketing. Dark Age America, John Michael Greer, 2016
It is a scandal, and it's no wonder that we are increasingly cynical about scientists' pronouncements on climate change, diet, or virtually anything else. Science, like politics, has become removed from the concerns of the people it's supposed to serve. The gap between on the one hand government and big business, and on the other ordinary people, grows ever wider. Without deliberate action that's unlikely to change, as many of the reason for the gap are self-reinforcing: more funding means more can be spent on advertising, and lobbying for still more funding and a more favourable - to corporate interests - regulatory environment. Ordinary people don't come into it. The result? Skepticism, cynicism, alienation, and despair. Not all the reasons for the gap can be ascribed to self-interest or malevolence, but it remains an urgent concern. 

My suggestion is that we start to express policy goals in terms of outcomes that are meaningful to ordinary people. Broad, long-term goals such as the improved health of a country's citizens, should guide the activities of all the agents, public- and private-sector that have anything to do with health: not only big pharma, but also the medical profession, agribusiness, retailers, schools, town planners, architects and so on. We cannot solve our health problems simply by spending more on organisations whose sole (stated) interests have to do with health. And just as health is not just a matter for drug companies alone, so crime is not just a matter for the police and the corrections industry. And so on.

Which is why I suggest we issue Social Policy Bonds that would not only target broad, long-term social goals, but also inject the market's incentives and efficiencies into their achievement. Social and environmental goals are more easily understood and more meaningful than the narrow, short-term goals of the bodies that are supposed to be achieving them but end up, as we see in science, spending much of their energy trying to justify their survival and expansion. With more comprehensible, explicit, policy goals, more people would engage with the policymaking process. Our views might be debated and over-ruled, but we'd have a greater understanding of the inevitable trade-offs that resource allocation entails. Importantly too, we'd have buy-in to policies that are not only currently remote from us, but also someitmes conflict with our interests. Most important of all, the goals of a Social Policy Bond regime would be understood by, and linked to, the concerns of all of us. It's a sad indictment that we cannot say that of our current system.

18 March 2017

Social Policy Bonds: the acceptable face of capitalism

Social Policy Bonds have many advantages over conventional policymaking. Mostly, they would be more efficient, as investors would have incentives to achieve society's social and environmental goals quickly and cost-effectively. Clarity and transparency of policy goals are further advantages, arising from the bonds' focus on, and explicit targeting of outcomes rather than the supposed means of achieving them. The question of which goals should be targeted is one that most governments evade or obscure, leaving a vague goal, most commonly 'economic growth', expressed as GDP per capita, to fill the vacuum. For much of history, that goal would have been strongly correlated with the improved well-being of a country's population. Nowadays, though, that correlation appears to be broken, with most of the gains in national wealth and income accruing to a minuscule number of people dubbed 'the elite'. In the US, for instance, the most recent data show that top 1% families captured 52% of total real income growth per family from 2009-2015 (source, pdf). 

Under a Social Policy Bond regime, we could ask whether society sees such inequality as a problem to be targeted directly or, rather, whether our main income goal is the alleviation of poverty. Even if inequality were seen as a problem in itself, it might not be necessary to target it for reduction. This is because Social Policy Bonds would be a means of acquiring wealth with which private gain is strongly correlated with public benefit. Many bondholders would be rich and, if their bonds were redeemed early, they would become richer. But this would be a socially acceptable way of acquiring wealth. Bondholders would become richer only by efficiently and quickly achieving society's targeted social goals. This means of accumulating wealth would allow other, less socially beneficial ways – inheritance, say - to be taxed more heavily. It might even transpire that investors would find achieving society's social goals will be easier than the sort of ethically questionable ways of acquiring wealth that constitute the ever more unacceptable face of capitalism. Social Policy Bonds: a new means of acquiring wealth that actually leads to a more cohesive society? Call them 'the acceptable face of capitalism'.

12 March 2017

Fossil fuel subsidies

Two brief excerpts from the International Energy Agency's publication, World Energy Outlook 2016, highlight what happens when there's no guiding strategic vision for our planet's future:
The value of fossil-fuel consumption subsidies dropped in 2015 to $325 billion, from almost $500 billion the previous year. ... Subsidies to renewables are around $150 billion today. World Energy Outlook 2016, Executive Summary; link to pdf here
The drop owes something to what the publication calls a 'subsidy reform process' but also to lower fossil-fuel prices. As well as these billions, there are, of course, subsidies for the exploration, development and production of fossil fuels. These are difficult to assess because of the “lack of transparency in government and company accounts, and limited information on off-budget subsidies to producers" source. It's easy to see how these subsidies arose, and why they persist. What is more difficult to discern is how our policymaking systems can do anything to address them. Large corporations have immense political power, partly funded by the very subsidies whose withdrawal they can effectively resist. The multitude of smaller interests, those of ordinary people and the environment, have little weight in our policymaking system.

There's nothing new in such an analysis, but what I can offer is a means by which we can aggregate and represent the interests of the vast majority of human beings - now and in the future - who don't benefit from energy subsidies and who suffer from the environmental depredations they cause. We could issue Social Policy Bonds that target for improvement the health of humans, plants and animals. Backed by national governments, the bonds would channel the market's incentives and efficiencies into improving the environmental well-being of the entire planet. Large corporations could still make money doing what they do now, but they'd find it less profitable. They might instead re-focus their resources into more environmentally beneficial areas. The way the bonds work means that bondholders would pick the lowest-hanging fruit first. In other words, they would maximise the environmental improvement per dollar spent.

But perhaps most important of all, Social Policy Bonds would clarify what we, as a species, wish to see. The current arrangement, the one that subordinates mankind and the planet to the narrow, short-term interests of large corporations, came about without consulting the rest of us. It thrives because the policymaking process is opaque to ordinary people, concentrating as it does on legalisms, institutional structures and, as we see with energy subsidies, obscure funding arrangements. A Social Policy Bond regime would change that. Because the bonds' starting point would be social and environmental outcomes, the public could participate in the policymaking process. So, if we wanted to donate billions of our taxes to large energy corporations we could still do so under a bond regime. I doubt that we'd choose to do that, but if we did we'd be doing so with our eyes open.

For more about environmental applications of the Social Policy Bond principle see here.

02 March 2017

How we decide

How we decide whom we shall vote for: Andrew Tyndall, a media analyst based in the US, monitors American broadcast news media at his website the Tyndall Report. He found that the overwhelming majority of the time spent by the US media in coverage of the run-up to last November's presidential election was spent on coverage of (1) the candidates and (2) polling.What about policy issues?
With just two weeks to go, issues coverage this year has been virtually non-existent. Of the 32 minutes total, terrorism (17 mins) and foreign policy (7 mins) towards the Middle East (Israel-ISIS-Syria-Iraq) have attracted some attention. Gay rights, immigration and policing have been mentioned in passing. No trade, no healthcare, no climate change, no drugs, no poverty, no guns, no infrastructure, no deficits. To the extent that these issues have been mentioned, it has been on the candidates' terms, not on the networks' initiative. Andrew Tyndall, 25 October 2016
From my viewpoint, it doesn't really matter whether we, the public, are more interested in personalities and opinion polls than issues, or whether the almost total absence of policy discussion reflects the preferences of the mainstream media. The important point is that we do not, as a society, debate the outcomes we want to see. Policymaking in all the democracies, not just the US, has become so obscurantist as to become inaccessible to all except those who are paid to follow it. All that's left for the rest of us is to discuss the personality flaws of the candidates, and how people react to them. It's not an edifying spectacle, nor is it one that makes for social cohesion. Powerful interests, with much at stake and resources to match, follow policymaking and influence it in their favour. We have, then, on the one hand government and big business (which these days includes much of academia) and on the other, ordinary people and small businesses. The gap between the two is already wide, and appears to be growing wider.

Social Policy Bonds could change that. They are focused entirely on the social and environmental outcomes we wish to see. 'We' meaning all society; not just policy wonks, ideologues and public- and private-sector interest groups. Social outcomes, under a bond regime, would be agreed and targeted for improvement. There's more consensus over such outcomes and the direction in which we want to move, than there is about the supposed ways of achieving them, or who we think will be better at achieving them - even when outcomes are actually under discussion which, as we see, is very seldom. And there's certainly more comprehension about what outcomes are than about the arcana that are such a feature of our current policymaking process. Of course, even under a bond regime, there would be disagreements about the priorities people would give to different goals but, importantly, even those whose strong opinions would be over-ridden, would have been consulted. More generally, with greater public participation in the policymaking process, there would be more buy-in - an essential feature that is largely absent from today's politics.

"How we decide" today is a risible process; one that is divisive and distracting. As a result "what we decide" is far removed from the concerns of ordinary people. Social Policy Bonds targeting broad, meaningful goals, would be a big improvement.

20 February 2017

Dealing with the unknown

Speaking ahead of an address to the Munich Security Conference, [Bill Gates,] the richest man in the world said that while governments are concerned with the proliferation of nuclear and chemical weapons, they are overlooking the threat of biological warfare. Bioterrorism could kill more people than nuclear war, Bill Gates to warn world leaders, 'The Telegraph', 17 February
One of the strengths of the Social Policy Bond principle is that we do not need to specify in advance the exact nature of the problem we want to solve. Many potential problems could arise from our denser, more linked, populations and higher technology. These are difficult even for a well-resourced government, or indeed any single big organisation, to anticipate, let alone do much to forestall. Hurricanes, tsunamis or pandemics are only a subset of a range of and disasters that threaten mankind. Others include the risks arising from new biological advances or scientific experiments that concentrate energy, or natural disasters such as asteroid impacts or volcanic super-eruptions. These threats are in addition to the 'known unknowns' of more widely understood catastrophes. Social Policy Bonds are versatile: depending on society's wishes, and the views of the bonds' backers, bonds could target any type of disaster that, say results in the deaths of 10 000 people within one week or its occurrence, however caused and in whatever part of the world.

Holders of such Disaster Prevention Bonds, would be able to redeem them only after a sustained period of, say, ten years, during which no such disaster has occurred. They would have incentives to anticipate potential disasters and to work to prevent them and mitigate their effects on human life. Today's disaster prevention policies are mostly carried out by bodies anticipating 'known' types of disaster, and mitigation strategies are often merely reactive. As well, few have the types of incentives that Disaster Prevention Bonds would put in place, that would reward the sustained absence of a disaster.

08 February 2017

Consensus and buy-in

From the current issue of the Economist:
Millennials are accustomed to tailoring their world to their preferences, customising the music they listen to and the news they consume. A system that demands they vote for an all-or-nothing bundle of election promises looks uninviting by comparison. Not turning out, 'Economist', 4 February
And not only millennials. There's palpable disillusion, cynicism and despair about where our political systems are taking us. There's consensus that they are not working, but little consensus about how to change them. In such circumstances, we appear to be drifting toward authoritarianism.

My suggestion? Instead of taking as our starting point the political party and its membership, let's focus on the social and environmental outcomes that we'd like to see. There's certainly more consensus over the in which direction we'd like to see broad outcomes move than there is about the supposed means of effecting that movement. I would say there's near total agreement amongst the global electorate that, for example, nuclear war is something we don't want. And while we do have organizations at every level trying to defuse conflict, they do not have the command of resources nor the expertise to look at the global interest over a period of decades. And unlike the 'defence' contractors or the ideological fanatics, neither do their rewards correlate with their success or otherwise in achieving their organizations' objectives. More fundamentally, because we take our current policymaking system as a given, the wishes of the vast majority of human beings have to be channelled through our over-worked, or entirely self-interested, or corrupt politicians and bureaucracies. Few people can afford to spend their time and expertise in working full-time to ensure nuclear peace. When the issue does exercise the public imagination, it invariably becomes a forum for energy-sapping, ideologically-based squabbling about partisanship and motives.

A Social Policy Bond regime targeting sustained nuclear peace would transcend party political differences. It would generate a motivated coalition of interests devoted to achieving that goal - which would be exactly society's goal, as laid down in the redemption terms of the bond. That coalition would be of diverse, changing composition and structure, but its goal of nuclear peace would not be subject to the whims and caprices of faddish ideology or party politics.

The same would apply to less lofty goals: there would be little debate about, for instance, the direction in which we'd want to see a nation's health go. We could debate definitions, targets and priorities but, unlike our current arcane policymaking system, these debates would be relatively easy to follow, so we should have greater public participation and hence greater public buy-in - an asset of crucial importance and one that's almost absent from today's politics.

For more about applying the Social Policy Bond concept to nuclear peace, see here. For health, see here. If you would like to consider supporting my work through patreon, please click here.