Author: Mark Hup (Chinese University of Hong Kong)
How and why did modern and centralized fiscal institutions emerge? Prior to centralized tax collection by state actors, tax collection by private parties – tax farming – was widespread. What then drove the use of tax farming as well as its eventual replacement by state-run tax collection? Using newly-collected archival data from colonial Indonesia in the period 1870-1920, this article studies the transition from tax farming to state-run tax collection and thereby contributes to our understanding of fiscal modernization, an integral facet of long-run development. The findings provide evidence for centralized fiscal institutions emerging when the state becomes less dependent on divide-and-rule strategies that route revenue streams through politically weak intermediaries.
State Capacity, Contracting, and the Politics of Revenue Collection
The paper starts from a basic point: the state wants revenue but also faces constraints in raising it. Tax collection suffers from principal-agent problems and requires costly administration and monitoring to measure the tax base and limit stealing and shirking by tax collectors. For low-capacity states, these costs can be prohibitive. But the paper argues that a purely administrative story is not enough. With regards to the tax farming transition, I posit that state capacity played a key role through two channels: the information channel and the divide-and-rule channel.
The information channel revolves around imperfect and costly information collection that induces the state to outsource tax collection. Such outsourcing through tax farming means that taxes are collected via the use of rent contracts rather than wage contracts. Instead of using salaried employees, the state sells tax collection rights to private parties who then hold residual rights to the tax revenue. Tax farming can therefore be a rational institutional response to weak monitoring and limited administrative reach.
The divide-and-rule channel highlights political strategy. The channel posits that tax farming helps the state to curb potential local political competitors by diverting revenue streams away from them. When the state cannot rely on direct collection to achieve this diversion, it can use tax farming to route revenues toward actors who pose little threat such as those from a politically weak minority. In this manner, fiscal design serves political control. Figure 1 shows the predominant role played by the Chinese, a small minority, in Java’s tax farming system.
These two channels yield two hypotheses. First, through mitigating information problems, stronger state capacity reduces reliance on tax farming. Second, through reducing the need for separating revenue streams from powerful actors, the type of state capacity expansion matters for reliance on tax farming. Specifically, I hypothesize that the relationship between state capacity and tax farming depends on the identity of the bureaucrat. Stronger state capacity enables the state to co-opt and incorporate politically powerful groups into its bureaucracy. In turn, such bureaucrats are particularly instrumental in replacing tax farming with state-run tax collection.
This yields a sharp political economy implication: state growth is not just about adding officials; it is about who those officials are, what they stand to gain or lose, and how the state’s revenue strategy reallocates rents.
Identity, Incentives, and Institutional Change in Colonial Revenue Collection
The empirical setting is colonial Java, where tax farming was a major source of revenue. I construct a new database that tracks tax-farm revenues and the rollout of state-run collection of the same revenue streams. Figure 2 displays the tax farm transition.
State capacity is measured using the number of state officials, and crucially, the paper distinguishes officials by legally defined groups. In colonial Indonesia, the state classified inhabitants into three legally distinct groups: Europeans, indigenous Indonesians, and non-indigenous Asians. This makes it possible to operationalize the paper’s political typology.
Under divide-and-rule tax farming, bureaucrats map into three groups: elite, excluded, or wooed. The elite constitute the state bureaucracy’s higher echelons. The excluded are groups the low-capacity state viewed as threats and therefore diverted revenue away from. The wooed are groups the state routed revenue through via tax farming.
These groups face different incentives in the tax transition due to their different stakes in the tax farming system. While the wooed are set to lose from the transition out of tax farming,
the excluded are set to gain. I therefore expect that the excluded (i.e., the indigenous) push hardest for the tax transition and are employed in the greatest numbers to implement it.
The headline findings match this logic. State capacity expansion reduced reliance on tax farming, but effects differ by bureaucratic segment. Indigenous officials – the majority group largely excluded from tax farming – strongly reduced reliance on tax farming. In contrast, non-indigenous Asian officials (mainly Chinese) – the group historically wooed as tax farmers – did not reduce such reliance. On the replacement side, the same pattern appears: different segments of the state bureaucracy differentially impacted the move toward state-run tax collection.
In short, the paper documents a fiscal transition that is not only about stronger administration in the abstract. It is about which officials expand, which networks are displaced, and when the colonial state becomes less dependent on divide-and-rule strategies that route revenue streams through politically weak intermediaries.
Beyond Administration: Bureaucratic Composition and Fiscal Centralization
Recent literature emphasizes fiscal capacity as an informational and administrative problem: monitoring, third-party information, and the technologies that allow states to measure tax bases and enforce compliance. This paper complements that view and pushes the debate in a specific direction. It shows that the relationship between state capacity and taxation depends on the composition of the bureaucracy and the incentives embedded in the revenue regime.
The central takeaway is a political economy claim supported with new data: centralized fiscal institutions can emerge when the state becomes less dependent on intermediated, divide-and-rule revenue strategies, and when capacity expansion incorporates groups previously excluded from the rents and authority associated with tax farming.
The findings imply that we should not treat the state as a unitary actor that automatically centralizes once it has the technical means. Centralization is also a redistribution of revenue access and influence within the state and between the state and social groups. In Java, the transition away from tax farming is therefore a window into how administrative growth, ethnic and legal stratification, and fiscal institutions co-evolve in the making of a modern tax state.