Versions of the state

Brad DeLong has a "reaction essay" entitled ""Perhaps. And Sometimes.", to Scott's Seeing Like a State recap discussed in the previous post. What he's reacting to is Scott's contention that the state's interest in so-called "legibility" as a means of ensuring state control comes at a cost to local or vernacular knowledge, systems, and practices, and is therefore often oppressive. Using a classical example, DeLong makes a good point that the control such legibility enables may well be preferable to an alternative in which local lords and warlords become much more oppressive than the central government. But, just as Scott tried to overextend his argument re: standardization, so DeLong overextends his good point -- as he puts it in the terms Scott uses:

A state that makes civil society legible to itself cannot protect us from its own fits of ideological terror, or even clumsy thumb-fingeredness. A state to which civil society is illegible cannot help curb roving bandits or local notables. And neither type of state has proved terribly effective at constraining its own functionaries.

In some ways, the “night watchman” state — the state that enables civil society to develop and function without distortions imposed by roving bandits, local notables, and its own functionaries, but that also is content to simply sit back and watch civil society — is the most powerful and unlikely state of all.

Translating that into more common terminology, he's saying that there are only two likely versions of the state available to us -- either a non-intrusive but weak version (i.e., what's usually called a "failed" state today) or a strong but intrusive, and frequently oppressive, version. But this rather bleak set of binary alternatives is really just an illusion resulting from flattening all versions of the state into a 1-dimensional spectrum, from weak to strong. In reality there are many other dimensions -- states can range along an autocratic to democratic spectrum, for example, and there can be weak or strong versions at either end. They can also vary along a  constitutional to tyrannical dimension, and again display weak or strong, autocratic or democratic versions at either extreme. And once we see that there are more than two possible versions of the state, we can see our way out of the trade-off between freedom and safety that DeLong presents. We can see, for example, that a limited state is not the same as a weak state -- i.e., that there's no inherent reason a state needs to be oppressive itself in order to control and suppress local oppressors.

What DeLong seems really to be talking about is just the perennial tendency toward oppression by rulers everywhere, whatever the scope of their rule -- which is certainly both an old problem and a current one. But that's only to say that the work to achieve, defend, and extend human freedom is as perennial as the efforts to oppress it. In our day, that work takes the form of upholding a version of the state that does indeed, as DeLong indicates, enable civil society without at the same time disabling it through bureaucratic intrusions. It's called making progress.