DC Bulletin

Can basic liberties be valued in monetary terms?

Many people object to the very idea of placing a “price” on liberty, even among the Chicago-school economists, who are usually eager to estimate the price of everything from human life to rain-forest. But sometimes, policy-makers are forced to weigh up civil liberties and competing social concerns. Can we assess such hard choices in monetary terms?

In one sense, yes. A method exists …

Between 2020 and 2021 all US states restricted basic liberties to varying degrees, to curb the spread of COVID-19, leading to high spatial and temporal variation in the enjoyment of civil liberties. This variation provides a unique opportunity to apply Revealed Preference valuation techniques to civil liberties, which are particularly hard to evaluate under normal circumstances. I will spare the reasons here. It is a long list.

I explore whether the opportunity cost of not curbing liberties provides an indicator of policymakers’ `willingness-to-pay’ for – or reluctance to forego – basic liberties.

I use US state and county-level data from 2020 to mid 2021 to explore the output costs of restrictions on liberties of movement and assembly, and the benefits they brought in terms of reduced COVID-19 cases, morbidity, and mortality. An optimal (net cost-minimising) control level is identified for the two liberties, which is compared with actual responses to infer reluctance and estimate its opportunity cost. I test the robustness of the solution to analytic choices, including data encoding, indexing, econometric specification, and monetary valuation method for morbidity and mortality (VSL).

When GDP bounce-back is considered, which drastically reduces the net output costs of restrictions, actual restrictions on movement and assembly were far less stringent than the optimum, but several plausible specifications return negative net benefits to restrictions on liberty of assembly. On average, the model suggests that the two liberties are jointly valued at 4 – 6 dollars per person day.

But there is no one “price” of liberty

When we look at variation between states, we find that the stringency of controls, and hence the valuation of liberty, is strongly associated with voting patterns in the 2020 Presidential elections. With some notable exceptions, controls tended to be tighter in more Democrat-leaning states. So the price of liberty depends on the political setting.

Different modelling choices also change the “price”, so a politically motivated analyst (is anyone immune?) might make modelling choices to find the “right” price. For example, just ignoring bounce-back effects would be enough to dismiss any idea that controls on liberty were worth it.

And it is not just a problem for liberty

Some of the leading proponents of monetary evaluation (such as Matthew Adler and Eric Posner) make the argument that just about everything can be valued and compared in a common (i.e., monetary) unit. I call this the “strong commensurability” position. I think they are right, but the existence of a method does not ensure the existence of a set of prices at which we can happily weigh up human life, liberty, justice, and so on.

For social or political goods that are not traded, the “price” at which one group of people in one set circumstances would be prepared to trade x and y could be wildly unacceptable for another group. It is one thing to say that we can put a price on everything, but quite another to conclude that this price provides an objective basis for evaluating tough trade-offs.

My (“weak commensurability”) position is that whereas methods may exist for monetary valuation of everything, the “price” reflects the mean or median political opinion in a given time and location. It cannot be extrapolated to other political constituencies and circumstances without making huge paternalistic assumptions. It is unlike the price of a traded good. We constantly extrapolate to make price predictions of decide on purchases.