It’s common to note that history matters for understanding contemporary politics. But does it? Here’s an interesting problem in the context of regional autonomy in Indonesia that raises some questions about whether history matters and how we would know if it did.

Regional autonomy is a big issue in Indonesia: in a phenomenally diverse country with a volatile and violent political history, it’s reasonable to worry that too much autonomy to particular regions might prompt secessionism (or at least intensify center-periphery tensions). Indeed, today Indonesia is a unitary state, with constant worries about the potentially fissiparous consequences of excessive regionalism. But Indonesia has also undertaken a massive decentralization exercise in the past 15 years. How does that work?

The answer is that it has decentralized most policymaking authority “two levels down,” to the district level rather than to the provincial level. Smart, no? Districts are too small to mount secessionist claims, but if they have most of the devolved powers under decentralization, then local politicians’ incentives will dominate those of provincial politicians, keeping any potential regional movements divided across multiple jurisdictions. At least, that’s the idea.

Now when go back into independent Indonesia’s political history, you can find some discussion about whether provincial autonomy was a good idea for Indonesia. In fact, Indonesia’s post-1949 constitution did contain provisions for provincial autonomy, and these were progressively rolled back by Sukarno and then Soeharto. And one alternative account of the reasons why this was done is that there was no historical precedent in the pre-colonial polities of what is today Indonesia for anything like provincial autonomy. Here’s a quote by Gabriele Ferrazzi in the colorfully-titled essay “Using the “F” Word: Federalism in Indonesia’s Decentralization Discourse.”

One argument in the public arena justified the abolition of provincial autonomy by reference to history, holding that provinces were created by the Dutch, with no real roots in Indonesian tradition; the Mataram and Majapahit kingdoms only recognized district (kabupaten) and village (desa) autonomy

Mataram and Majapahit are two great Java-centered empires, and the latter covered much of what is today Indonesia. Majapahit serves at times as the pre-colonial “justification” for why Indonesia is a coherent nation. The reasoning goes that if they didn’t have provincial autonomy, then there’s no history of provincial autonomy upon which contemporary Indonesia can draw. It’s not Indonesian to have provincial autonomy.

Now this argument, as Ferrazzi notes, was clearly for public consumption, but it does get us to the heart of the matter of how history matters and the difference between historical legacies and comparable incentives. Why didn’t Majapahit recognize provincial autonomy? Because it—like Indonesia—was worried about empowering its regions. Even though neither Mataram nor Majapahit were states in the Westphalian or Weberian sense, Majapahit especially faced the same general problem of ruing over a sprawling archipelago that Indonesia faces today.

So it could be that it’s true that Indonesia avoids provincial autonomy because there is no historical legacy of provincial autonomy. Or it could be that Indonesia avoids provincial autonomy because provincial autonomy is threatening, and that’s the same reason that Majapahit avoided provincial autonomy. One is an argument for historical state forms determining contemporary policies, and the other is an argument that historical and contemporary state forms face the same structural problems, and any historical connection is just epiphenomenal on those common incentives.