Experts and values

What is the scientific method? How does science advance? Why did modern science emerge when it did about 300 years ago? Will it continue? All questions addressed by Michael Strevens’ interesting and enjoyable The Knowledge Machine: How an Unreasonable Idea Created Modern Science.

The book begins by setting out two competing theories about the scientific method: Popper’s claim that scientific knowledge progresses through falsfication of hypotheses and Kuhn’s even more famous claim that it occurs through revolutionary changes in paradigm. These are incompatible explanations – if falsfication happened, how could scientists hold on to an incorrect paradigm – and both wrong, the book argues. The falsfication hypothesis not states that ‘facts’ are only tentatively true, always vulnerable to falsification – so we can be no more confident that the sun will rise today (Hume’s induction problem) than that a human born today will live forever. The paradigm revolution hypothesis seems to rule out progress or at least doesn’t account for why the scientific community jumps from one paradigm to another with better predictive power. Strevens is even more critical, though, of the radical subjectivists who argue that there is no scientific method, that science is fundamentally all about the status of scientists, how they exercise power within the scientific institutions so that certain ideas come to dominate.

It is not that scientists are not subjective, he says; of course they have personal priors in evaluating evidence. However, the key point is that in the official public discourse in scientific publications and seminars is that there is an iron rule: only empirical evidence with predictive power counts. The evidence will often be microscopic or concern minute differences in measurement: much scientific endeavour is super-dull, detailed collection of measurement. Scientists have to be trained to have a high tolerance for this painstaking work, because it is the only way progress happens – much as in their private lives they are fascinated by the big philosophical theories or the beauty of the universe. In scientific debate, “Victory does not come through smooth rhetoric, metaphysical inquiry, moralizing, or any other sort of sweet talking or big thinking. To win, players must front up with meticulous observations.” The subjectivists, the book argues, overlook this actual scientific method – which was cemented as the way to do science by Newton.

Is this immensely successful scientific method secure? The book ends with a section on the need to retrain every genertion of scientists in this approach – dull meticulousness in disciplinary debate whatever their private philosophical flights of fancy. And on the need to sustain the institution of modern science in the face of the challenges facing humanity today, from covid to climate.

It’s a very enjoyable book, with plenty of stories of science past and present. The argument certainly persuaded me. Economics differs substantially in that the social processes we investigate do not generate data in the same way as physical or biological ones, but the ideal method is the same: of course we bring our own values to evaluating evidence but it is the evidence that wins the official disciplinary debate. We cannot ultimately be objective and impartial but we can and should set objectivity as the standard for evaluating claims. So I recognised the book’s description of the iron rule as the way researchers (try to) operate. And similarly, I reject the subjectivism of those who say economics should be pluralistic – can only be so – because there is no way of testing claims against each other.

419TbK7hZPL._SX323_BO1,204,203,200_My other reading this week has been Behold, America by Sarah Churchwell. Having been comforted by the arrival or President Biden, the book got me worried all over again about the long and strong roots of American fascism – shocking and startling photos of the swastika flying in Madison Square garden rallies. The book analyses the histories of the slogans America First and the American Dream – tracing the more-or-less constancy of the first (white supremacy) and the evolution of the second (from an evocation of equality and civic participation to a claim about material prosperity) through newspapers and other records. Along with this evidence about the persistence of populism, perhaps we should stay worried.



Burning platforms

In 2011 the CEO of Nokia, Stephen Elop sent a memo to staff comparing the business to a burning platform from which it was essential to jump in order to change course. He was referring to the disruption of Nokia’s revenues and profitability by Apple and Android smartphones, and he was right. Rebecca Henderson’s Reimagining Capitalism in a World on Fire is essentially arguing that most of the world’s businesses are on burning platforms, if only they knew it. Either the disruption of climate change consequences, or of social insurrection due to inequality – or, as it turns out, the effect of a global pandemic, though the book predates this – will destroy capitalism. Unless businesses across the board change their ways.

I got the book as it was one of the FT Business Book award shortlist titles I hadn’t yet read (I don’t fancy the ones on Netflix or Instagram, but have read & enjoyed all of Deaths of Despair, If/Then and The World Without Work). Henderson clearly has vast experience of engaging with businesses of different kinds, and much of the book is about the success stories – those that have re-engineered themselves to become oriented toward purpose rather than profit. The examples include Unilever and Aetna, described in some detail, as well as old chestnuts like the worker-owned John Lewis and Mondragon, and a range of smaller companies, and industry initiatives like the move to purchase sustainable palm oil.

The book has a good term for what’s needed to make the kinds of changes described in these examples: architectural disruption. As Henderson acknowledges, many more businesses are still oriented toward short term profit and share price rather than long-term social purpose – even though the purpose-driven businesses ultimately do far better in conventional terms. She identifies some key barriers, among them the necessary big internal re-organisation and culture change. Becoming a purpose-driven and high productivity business requires a high level of trust within the firm, and many managers are unwilling and able to embark on this programme.

There are external barriers too: the short-termism of some investors, the difficulties of getting co-operation among businesses, and the political and regulatory context. So reforms to corporate governance and finance (including proper risk-measurement and accounting), and to the political climate of ideas will be needed in addition. List all that’s needed and it can seem daunting. But we’re all on a burning platform. My guess is that several forces will converge to bring about change – millenial employees demanding better, political upheaval given the state of the world, and un-ignorable consequences of the damage to nature. Whether the change will happen fast enough is another matter.




Law and technology, power and truth

I’d been looking forward to reading Between Truth and Power: The Legal Constructions of Informational Capitalism by Julie Cohen. I start out sympathetic to the key argument that the legal system is the product its society – the material economic and technological and also the political conditions. Therefore current legal contests reflect underlying changes in these same forces, and the final legal system for the digital era is still in the process of being shaped. Law and the implementation of technologies in society influence each other – the ‘rules of the game’ are not exogenous.

There are indeed some interesting insights in the book. I most enjoyed some of the earlier parts, which are more descriptive of the extension of the idea of intangibles, ideas, as “intellectual property”. This is not new – James Boyle for one has written superbly about it. But the detail here is interesting, plenty of nuggets about the US legal system and how truly, gobsmackingly awful it can be.

I also appreciated the chapter on regulation, and its basic point that many regulators are now having to “move into the software auditing business”, and indeed may have to evaluate software controls designed to evade regulation. Implications here for analysing regulation in the digital economy in terms of hyper-asymmetric information and algorithmic complexity.

On the whole though, I was disappointed. The book is almost entirely US focused – and is upfront about it – but the US system is distinctive even compared to other common law jurisdictions. Nowhere else, for instance, has its 1st Amendment fetishism. It would have been terrific to have some reflections on the extra-territoriality of US lawmaking and court judgements in the digital domain.

The book has some tantalising reflections about the limitations of law, based on concepts of individual rights, in the face of collective effects: as I’ve been arguing for a while eg here, digital power spells the end of individualism, including what we all now call the neoliberalism that gave birth to it. More on this would have been great.

There’s also just too much sub-Zuboffian rhetoric (rather than argument) about the ‘surveillance-industrial complex’. I’m all too willing to believe this exists, so all the more disappointed when the analysis is so vague. There’s also a lot of allusion to Foucault – “biopolitics”, “governmentality” –  in the book without it ever – as far as I spotted – actually citing and deploying The Birth of Biopolitics.

All in all, there’s a lot of detail in this book that didn’t, at least for me, cohere. Too many trees, not enough overview of the wood. Perhaps it’s time for me to try the highly-praised The Code of Capital by Katarina Pistor.



Tech self-governance

The question of how to govern and regulate new technologies has long interested me, including in the context of a Bennett Institute and Open Data Institute report on the (social welfare) value of data, which we’ll be publishing in a few days’ time. One of the pressing issues in order to crystallise the positive spillovers from data (and so much of the attention in public debate only focuses on the negative spillovers) is the development of trustworthy institutions to handle access rights. We’re going to be doing more work on the governance of technologies, taking a historical perspective – more on that another time.

Anyway, this interest made me delighted to learn – chatting to him at the annual TSE digital conference – that Stephen Maurer had recently published Self-governance in Science: Community-Based Strategies for Managing Dangerous Knowledge. It’s terrifically interesting & I recommend it to anyone interested in this area.

The book looks at two areas, commerce and academic research, in two ways: historical case study examples; and economic theory. There are examples of success and of failure in both commercial and academic worlds, and the economic models summarise the characteristics that explain whether or not self-governance can be sustained.

So for instance in the commercial world, food safety and sustainable fisheries standards have been adopted and largely maintained largely through private governance initiatives and mechanisms, synthetic biology much less so, having an alphabet soup of competing standards. Competitive markets are not well able to sustain private standards, Maurer suggests: “Competitive markets can only address problems where society has previously addressed some price tag to the issue.” Externalities do not carry these price tags. Hence supply chains with anchor firms are better able to bear the costs of compliance with standards – the big purchasing firm can require its suppliers to adhere.

Similarly, in the case of academic science the issue is whether there are viable mechanisms to force dissenting minorities to adhere to standards such as moratoria on certain kinds of research. The case studies suggest it is actually harder to bring about self-governance in scientific research as there are weaker sanctions than the financial ones at play in the commercial world. Success hinges on the community having a high level of mutual trust, and sometimes on the threat of formal government regulation. The book offers some useful strategies for scientific self-governance such as building coalitions of the willing over time (small p politics), and co-opting the editors of significant journals – as the race to publish first is so often the reason for the failure of research moratoria to last.

The one element I thought was largely missing from the analytical sections was the extent to which the character of the technologies or goods themselves affect the likelihood of successful self-governance. This is one aspect that has come up in our preparatory work – the cost and accessibility of different production technologies. The analysis here focuses on the costs of implementing standards, and on monitoring and enforcement.

This is a fascinating book, including the case studies, which range from atomic physics to fair trade coffee. It isn’t intended to be a practical guide (& the title is hardly the airport bookstore variety) but anybody interested in raising standards in supply chains or finding ways to manage the deployment of new technologies will find a lot of useful insights here.



Calculating the economy

One of the books I’ve read on this trip to the AEA/ASSA meetings in San Diego is The People’s Republic of Walmart by Leigh Phillips and Michael Rosworski. This is a very entertaining projection of the socialist calculation debate onto modern capitalism.

41JGcj2r26L._SX329_BO1,204,203,200_The starting point is the Simon/Coase realisation that big firms are internally planned economies – if it works for Walmart, why wouldn’t it work at larger scale? The authors’ hypothesis is that economic planning might work better now that we have so much more powerful computers and better data.

I’d recommend the book as an introduction to the socialist calculation debate for those unfamiliar with it (ideal for students). It cites some of my favourite books including Francis Spufford’s Red Plenty and Eden Medina’s Cybernetic Revolutionaries. Some chilling lines – about Stalin’s purges, for instance: “Anyone with any expertise was placed under suspicion.” It’s a great read.

Am I persuaded? Not entirely. Technology clearly will change organisational configurations, but it has just as much been decentralisation of firms and extended supply chains as it has been giant Walmart-type firms. I’m also sceptical that the data available is actually the information needed to plan an economy, or that it’s easy to access and join up. Still, it’s the right question, and a reminder that the boundary between market, state and other forms of organistaion is not set in stone but needs constant negotiation – in fact, I know a great book about this about to be published: Markets, State and People.

0FBA95F6-E1DB-4017-A90A-89A2D3C43EB7As seen at ASSA2020 in San Diego