The Problems With Institutional Legibility

We often hear about the value of things like accountability and transparency in institutions and society, from education and medicine to hiring practices. Only this week, I was listening to a panel on reforming academic peer review and everyone present was very committed to the idea that these are clear goals we should be striving for.

For my part, though, I'm coming round to the view that I'm against accountability and transparency, or at least against making them explicit guiding ideals when regulating and designing institutions and practices. In short, my justification will be familiar to anyone who's read Seeing Like A State or Scott's superb review of it: overwhelmingly when organisations push for more accountability and transparency, this amounts to making individual decisions more institutionally legible, and at best this involves additional paperwork for individuals, and at worst undermines expert decision-making. I would also speculate that it's a major factor in cost disease.

A couple of examples. First, my dad has been a physician for five decades, and has basically been awarded every accolade you get can in family practice. He's operated through most of his career on the basis of intuition and deep knowledge of medicine. Sometimes this means he refers patients to specialists even when they don't meet all the usual criteria for referral, occasionally on the basis of intuition alone. He has a great track record in this regard. On other occasions, he'll prescribe non-conventional treatments, again largely based on intuition. In one case, for example, this involved putting a patient with recurrent nausea on a dose of SSRIs. In short, this patient had had recurrent GI distress but didn't respond to conventional treatments and had already been through a lot of invasive diagnostic procedures with no success. My father had a hunch that the condition was real but psychosomatic, and induced by a particularly unpleasant incident several years earlier when the patient had been hospitalised for food poisoning and had very negative experiences at the hospital. After treatment with SSRIs, the symptoms disappeared. However, my father is finding it increasingly hard to get away with this kind of decision. As the health system in our country moves towards increasing transparency and accountability, doctors are more and more often required to justify their decisions via a series of strict rules and procedures. At best, this means that he has to spend extra hours on paperwork (that could be spent treating patients) to explain why this or that unusual decision was appropriate. At worst, it means he's unable to prescribe or refer patients in virtue of their atypical presentation, and thus cannot exercise the undeniable metis he's acquired from years of experience. He worries a lot about the effect that this increasing bureaucratic workload and the marginalisation of the role of autonomous judgment has had on morale within the medical profession.

Second, education. I've taught in a range of universities, and I've been lucky enough to have a lot of freedom when designing syllabi and marking schemes. I'm an experienced instructor and have a clear idea of what I look for in a good paper. Sometimes, this can mean highly nonconventional papers get a good grade. It can also mean that papers that officially 'tick all the boxes' get a bad grade because the whole is less than the sum of its parts. I am always available to give students more information about why they got the grade they did. On the occasions when I've had to work with an external marking scheme, my procedure is usually pretty straightforward - I decide what grade the paper deserves, and then figure out how to justify it within the context of the marking scheme. This is an annoyance, but a fairly minor one. By contrast, my friends in secondary education are universally despondent about the amount of time they spend justifying and measuring teaching and grading decisions. They complain about the lack of autonomy they have as teachers and the ridiculous amount of paperwork they have to process. In the UK at least, this seems to be a factor in explaining why so many talented teachers choose to work overseas. See this recent Guardian article, for example, and note the quote from one of the teachers interviewed: “In the UK you are constantly having to report to certain people about certain things. Here you are trusted to do what you think is best for the student.” Again, much of this bureaucratic burden is in place in the name of accountability and transparency.

I recognise that the above examples aren't necessarily indictments of transparency and accountability per se. I'm also willing to grant that there may be many cases where these are worthy goals whose pursuit will yield real dividends. However, within in the context of social institutions like those above, my broad view is seeking accountability and transparency will result in major inefficiencies and the destruction of metis. If you want efficient systems that allow employees to exercise their talents, this should be achieved not via institutional-level measures but by focusing on recruiting and retaining high quality staff. Of course, this is easier said than done. But my model for building such systems would be, in short: spare no expense on recruiting high quality staff, and then let them get on with it. (This, incidentally, is one of the lessons I take from The Wire, still my favourite TV show of all time: institutions corrupt and create distorting incentives, thus minimise your organisational footprint and focus on nurturing skilled employees).

Here are a few worries and unanswered questions I have.

(1) How does this apply to different fields? Maybe I'm generalising too much from education and healthcare. What about all the other institutions that maybe have different dynamics at play? How much autonomy to we want to give employees at the DMV, the post office, or social security offices?

(2) How does this apply to high-skill vs low-skill workers? One reason my dad and my friends are maybe good examples of why we shouldn't make employees accountable is that they're smart highly skilled individuals. But there's only so much talent to go around, and there are major cost savings that might be available if we reassign certain tasks from high- to low-expertise workers. But that involves giving autonomy to people who perhaps lack the skills to exercise it appropriately. In these cases, perhaps accountability and transparency as manifest in strict guidelines and frameworks constitute the best model for preventing colossal fuckups and ensuring everyone does their job reasonably well.

(3) Bad actors problem. Again, in the examples above, we can assume the people concerned are highly motivated and conscientious individuals. But what about workers who are lazy and malicious, and perhaps criminally negligent? One way to prevent them from doing harm is to have clear standards in place and externally legible measures of performance. Without such measures, how can we catch the 'bad eggs'?