An Overview on Civil Service and Government Expansion

About the Study

The structure of democratic administrations underwent significant alterations in the late 19^th and early 20^th century. First, several rounds of civil service reforms depoliticized government employment processes. Initial initiatives usually established a merit system, requiring that government personnel be hired based on their official qualifications and exam results rather than their partisan affiliations. Later, tenure systems were established to safeguard civil officials from politically driven dismissals. The 1883 Pendleton Act established a merit civil service in the United States, which was steadily expanded to incorporate most federal employees during the following decades. President McKinley issued an executive order in 1897 that established minimal rights against dismissal, which were codified in the Lloyd-LaFollette Act of 1912. As a result, voters in all civil service systems have legitimate, but distinct, grievances about government services.

I propose a dynamic two-party election model in which the ruling party transfer tax revenue to the bureaucracy, which can either spend it on public goods or divert it to political expenditures that sway voters' re-election decisions. I presume that bureaucrats are chosen on the basis of merit, i.e., they have the expertise to generate public goods at no cost and have no inherent bias toward one political party over another. Indeed, I expect that officials chosen through an apolitical procedure will have neither the desire the skills, or the need to manipulate spending to favour the government's re-election.

Indeed, because bureaucrats hired through a political process should not have the motivation, skills, or obligation to distort expenditure in order to promote the government's re-election, I suppose that they can only do so if they make costly partisan investments. As a result, in equilibrium, the bureaucracy's partisan expenditure must be backed up by endogenous preferences for the incumbent's re-election that are strong enough to overcome these partisan investment costs.

If a ruling party returns to power, it can engender these preferences by offering large compensation to the bureaucracy. As described in Section 7 following the presentation of my primary results, several components of the model can readily be made more general. For example, I show that if bureaucrats have only limited choice over government expenditure, all of my results are preserved. What matters is that bureaucrats keep their ability to shirk by squandering government funds. Other factors, on the other hand, are vital to my success. For example, I show that a civil service structure that lacks both merit and tenure cannot even enable the supply of partial public goods. However, public goods can be delivered efficiently if the relationship between politicians and bureaucrats is sufficiently clear that expenditure decisions made by bureaucrats do not influence voters' election choices.

Both advocates and opponents of civil service reforms were concerned about the negative consequences of tenure, such as the impact on officials' motivation to work hard. Some reformers even claimed that a well-functioning merit system would eliminate the need for tenure: if politicians can't distribute dismissed bureaucrats' positions to political cronies, why dismiss them in the first place? 2 In Section 4, I examine untenured bureaucracy and find evidence to support the idea that merit systems can limit political use of public monies on their own. In Section 6, my last main findings address the question of when voters want a tenured bureaucracy. I show that this occurs when the advantages of public goods that can only be provided by politically insulated bureaucrats outweigh the labour costs of producing them. As I mentioned in my first paragraph, civil service reform and the scope of government activities are inextricably linked, with the electorate's desire for public goods serving as the underlying predictor. This is consistent with historical studies of civil service reforms, which show how politically subordinate bureaucracies in the 19^th and early 20^th centuries were unable to fully execute the new tasks that citizens were urging their governments to take on.