Algorithms and (in)justice

It’s been one of those weeks. One of those years, actually – David Bowie *and* Leonard Cohen. Listening to ‘Democracy‘ as I write this.

Still, I have managed to read Cathy O’Neil’s excellent Weapons of Math Destruction, about the devastation algorithms in the hands of the powerful are wreaking on the social fabric. “Big data processes codify the past. They do not invent the future. Doing that requires moral imagination, and that’s something only humans can provide. Sometimes that will mea putting fairness ahead of profit.”

The book’s chapters explore different contexts in which algorithms are crunching big data, sucked out of all of our recorded behaviours, to take the human judgement out of decision-taking, whether that’s employing people, insuring them or giving them a loan,  sentencing them in court (America, friends), ranking universities and colleges, ranking and firing teachers….. in fact, the scope of algorithmic power is increasing rapidly. The problems boil down to two very fundamental points.

One is that often the data on a particular behaviour or characteristic is not observed, or unobservable – dedication to work, say, or trustworthiness. So proxies have to be used. Past health records? Postcode? But this encodes unfairness against individuals, those who are reliable even though living on a bad estate, and does so automatically with no transparancey and no redress.

The other is that there is a self-reinforcing dynamic in the use of algorithms. Take the example of the US News US college ranking. Students will aim to get into those with a high ranking, so they have to do more of whatever it takes to get a high ranking, and that will bring them more students, and more chance of improving their ranking. Too bad that the ranking depends on specific numbers: SAT scores of incoming freshmen, graduation rates and so on. These seemed perfectly sensible, but when the rankings they feed into are the only thing that potential students look at, institutions cheat and game to improve these metrics. This is the adverse effect of target setting on addictive crystal meth. Destructive feedback loops are inevitable, O’Neil points out, whenever numerical proxies are used for the criteria of interest, and the algorithm is a black box with no humans intervening in the feedback loops.

The book is particularly strong on the way apparently objective scoring systems are embedding social and economic disadvantage. When the police look at big data to decide which areas to police more harshly, the evidence of past arrests takes them to poor areas. A negative feedback loop – they are there more, they arrest more people for minor misdemeanours, the data confirms the area as more crime-ridden. “We criminalize poverty, believing all the while that our tools are not only scientific but fair.” Credit scoring algorithms, those evaluating teachers using inadequate underlying models, ad sales targetting the vulnerable – the world of big data and algos is devastating the lives of people on low incomes. Life has always been unfair. It is now unfair at lightning speed and wearing a cloak of spurious scientific accuracy.

O’Neil argues that legal restraints are needed on the use of algorithmic decision-making by both government agencies and the private sector. The market will not be able to end this arms race, or even want to as it is profitable.

This is a question of justice, she argues. The book is vague on specifics, calling for transparency as to what goes in to the black boxes and a regulatory system. I don’t know how that might work. I do know that until we get effective regulation, those using big data – including especially the titans like Facebook and Google – have a special responsibility to consider the consequences.


Bank of England independence from Nigel Lawson

This is a bit late but it has been in the back of my mind this week that Lord Lawson used to be keen on isolating central banks from political pressure. I remembered right. He made the case to a sceptical Margaret Thatcher, arguing it would give UK monetary policy an anchor that was necessary if the country was not to commit to an external anchor such as ERM membership. Here is Nigel Lawson in 1992, in his autobiography The View from Number 11:


Nigel Lawson in 2016, about Mark Carney – a change of tune about the propriety of political pressure on an independent central bank, then:untitled

The trade-investment-service-intellectual property nexus

I’ve managed to resist reviewing Richard Baldwin’s new book The Great Convergence: information technology, trade and the new globalization until now, and it has taken serious self-restraint as the book is so relevant to (among other things) the Brexit debate. I would for one thing force every Cabinet member to read it and not allow them to keep their jobs unless they could pass an exam based on it. Anyway, the book’s published on 14th November and now it’s November my self-denying ordinance can end.

The Great Convergence offers a compelling framework for thinking about how trade is organized and why and how it benefits whom. The first part is a historical overview of trade leading up to the first, the Old Globalization or the 19th century. This phenomenon, due to steam power reducing trading costs, industrialization and a context of relative global peace led to the Great Divergence: the major economies of Asia, which had been richer than the West, fell behind, dramatically so over the course of two centuries. The New Globalization, since the 1980s, driven by the new information and communication technologies, has taken the rich countries’ share of global output back to its 1914 level in little over two decades. China is the standout story, going from uncompetitive in 1970 to 2nd biggest in the world by 2010, but other rapidly industrializing nations in the New Globalization are Korea, India, Poland, Indonesia and Thailand (ie. a different group from the notorious BRICs).

However, as the book goes on to document, the New Globalization is a completely different kind. Trade over distance has three costs: the costs of moving goods, ideas and people. When moving goods got cheap, the first explosion of trade occurred, but ideas were costly to move so the innovations of the industrial revolution were not easily exported. The Old Globalization was the result of low shipping costs and high communication costs. ICTs have reduced the latter significantly, so industrial competitiveness is defined in terms of production networks, interlinked supply chains, that cross national borders. Knowledge has been offshored, and the rapid growth in a few previously poorer countries has come about because of their geographical location, close enough to G7 industrial centres that managers can travel there, sharing knowledge within the confines of the production network.

This means the New Globalization happens at the level of stages of production and occupations. This makes it harder to predict who will be affected – which jobs will be offshored, which areas most affected. “Nations are no longer the only natural unit of analysis”. Much of the book describes a new data set making it possible for economists to begin to explore the ‘value added’ pattern of trade created by the switch from trading finished goods toward trading components in global production chains. The picture is going to be utterly different – the famous example being the iPhone which is sourced conventionally as a Chinese export to the US but where the value added is concentrated in the American business and the Chinese import a lot of the components they assemble and re-export with not much value added at that stage.

This is one insight the Brexiteers need to appreciate, although the Nissan letter suggests at least some members of the government realise the signficance. British businesses are woven into supply chains with our near neighbours: we aren’t importing prosecco and salami so much as gear boxes. Brexit threatens to tear apart these links. If the cost appears to be too high, the multinationals at the head of the supply chains will relocate chunks of their production networks, and won’t care if they’re exporting gear boxes to the Czech Republic rather than Britain.

The book adds: “Twenty-first century supply chains involve the whole trade-investment-service-intellectual property nexus, since bringing high quality, competitively priced goods to customers in a timely manner requires international coordination of production facilities via the continuous two-way flow of goods, people, ideas and investments. Threats to any of these flows become barriers to global value chain participation…” Baldwin adds that the movement of people is still a binding constraint on globalization, and face-to-face communication – and so distance – remain important. He argues that the improving quality of telepresence is changing this, but I think that remains to be seen.

Ultimately, trade policy today is not just about trade nor about nations. It involves deploying the nation’s productive resources through overseas connections. This is why 90% of the economics profession thought, and thinks, Brexit so damaging, and the idea that the UK has more economic self-determination outside the EU a delusion. The Great Convergence is not about Brexit – it ranges far wider. I can’t imagine a better and more accessible analysis of trade and globalization in the digital era.


(20 November: minor typos corrected)

Still more upcoming books

It’s always a pleasure to read the new publishers’ catalogues. Apart from the lovely pictures, one always has the frisson of thinking about all the interesting books there isn’t going to be time to read but maybe mental osmosis from looking at the covers and blurbs works a little….

Anyway, the Yale University Press catalogue for Spring and Summer 2017 just landed. The economics/tech books on offer promise a rather gloomy view of the world.

I’m looking forward to (and will definitely read rather than osmose) Stephen King’s Grave New World: the end of globalisation and the return of the economics of conflict. The title says it’s even gloomier than his last book, xxxx. There are two interesting looking titles on environmental economics. One is my Natural Capital Committee colleague Dieter Helm’s Burnout: The endgame for fossile fuels. The other is Benjamin Barber’s Cool Cities: the urban fix for global warming.

The other title here I’m very much looking forward to reading is Zeynep Tufekci’s Twitter and Teargas: the power and fragility of networked protest. Remember the Arab Spring, the hope of social media facilitated liberation movements? It seems an aeon ago. Now we have authoritarian surveillance and Trumpism and the Alt-Right.


Wanting to change

Anybody who reads Duncan Green’s excellent blog, From Poverty to Power, won’t be entirely surprised by the approach he takes in his equally excellent new book, How Change Happens. It is based on two pillars. One is Amartya Sen’s capabilities approach to human development (‘the freedoms to do and to be’), and I’ve always thought that when you appreciate its ethical and practical merits, it’s hard to take any other approach. The other is the need for systems thinking when it comes to considering economic policies or other interventions – in any context, really, but certainly in the case of development.

“Change in complex systems occurs in slow steady processes such as demographic shifts, and in sudden, unforseeable jumps,” Green writes. Mostly, change is extremely, painfully slow. It turns out to be impossible to do one thing because another, linked thing gets in the way. Events and crises open the way for the big shifts – being an economist, I think of this in terms of what it takes to move a co-operative game to a new focal point. But even then, the direction of the jump is contingent, messy, unpredictable. It anyway depends on the prevailing climate of ideas and norms – so part of the challenge is to be ready to take advantage of a crisis by having done all the contextual spade work, all the while getting on with the day job of trying to bring about incremental changes in the previous state of affairs.

Needless to say, this does not make for a concise ten-point plan in the final chapter (although it does try to sum up the whole in a ‘power and systems approach’ in the final few pages). The book has some interesting practical ideas, however. I like the principle of looking for ‘positive deviance’ – look for examples of people or activities that succeed against enormous odds, for outliers, and use them as ‘social proof’ so others copy whatever it is. This is exactly the way new technological innovations spread: the ideas are there, a few people try, and others imitate them. There are loads of examples of advocacy and development organisations and initiatives that have been able to implement responsive, adaptable changes (many of these brought Tim Harford’s Adapt to mind). Other suggestions are harder to see how to implement. The book argues that principled leadership matters. I agree. But where is it? How do donors encourage it?

Green concludes that many organisations in the aid world, including his own, need to move away from linear thinking and get wiser to context and the whole complex environment (actual and political) in which they operate. I hope they follow his advice and this book is certainly well worth anyone working in this world reading. The one element missing, though, seems to be the meta-analysis of the development agency ecosystem itself, and the prevailing ideas. For example, how do you get social innovation akin to technological innovation in a world of impact assessment and RCTs? Or indeed combine fleetness of foot with a genuine need to understand ‘what works’? Understanding one’s own cognitive biases or limitations is a tall order. What’s more, the aid world has incentive structures built in that will discourage change. In a variation on the old lightbulb joke (How many psyhologists does it take to change a lightbulb? Only one, but the lightbulb has to really want to change), how change happens is that a lot of people have to want change to happen.

Anyway, there’s no excuse for not reading the book, as it’s also published as an open access pdf. I hope lots of activists read and digest and change their approach, but suspect it will prove difficult for many.