Suntory and Toyota International Centres for Economics and Related Disciplines (STICERD) LSE RSS Contact Us YouTube Twitter

Bureaucrats are a core element of state capacity: they are responsible for implementing policy and may therefore have a critical bearing on societal outcomes. Yet, despite their centrality to development and poverty reduction, the incentives civil servants face within bureaucracies are seldom studied. We combine administrative and survey data to study the incentives of elite civil servants in India - officers of the Indian Administrative Service (IAS).

In contrast to firms, bureaucratic organizations are characterized by rigid rules: selection through competitive examinations, absence of discretionary firing, seniority-based progression rules and fixed retirement age. While these rules are a direct response to earlier patronage systems, where appointments and promotions were decided based on personal connections or political favors, heavy reliance on rules may also introduce rigidities. Although time-based promotions are objective and well-suited to limit favoritism and political interference, promoting bureaucrats predominantly based on seniority can weaken the link between effort and return, blunting a critical source of career incentives.

The Indian Administrative Service basically runs India. Officers from the service head up all government departments at the centre and state levels. Itís centrality to policy making during a period of deep economic reforms in a country containing over a billion people make it a particularly interesting focus of study. Literally hundreds of thousands of people in India take civil service exams each year with only the top hundred or so making into the IAS. Upon entry officers are assigned to an Indian state which they remain attached to for the whole of their career though they may also spend periods in central government in Delhi.

As a classic bureaucracy, promotions within the IAS are time-based and occur after 4, 9, 13, 16, 25 and 30 years of service. Though promotions are in part based on an internal performance review, officers do not move to a higher payscale until the minimum number of years of experience has been achieved. With recruits entering aged 21-30 and retiring with 60, this implies that officers entering older will mechanically face barriers in how far they can progress. As Figure 1 shows, officers who enter older than 28 will have almost no chance of getting promoted to the highest payscale of Chief Secretary, which requires at least 30 years of continuous service within the IAS. This rigid progression rule stands in stark contrast to the private sector, where the practice of fast-tracking high performers is often considered to be "good" management practice.

Figure 1. Share of retired officers reaching senior payscales by age at entry. (Click to see larger image.)
figure 1 Notes: Share of retired officers in 2012 reaching senior payscales as a function of age at entry. Number in parenthesis indicates the minimum number of years to qualify for the position.

We are interested in understanding whether these rule based rigidities affect the performance of IAS officers and whether this in turn affects the economic performance of the states over which they have jurisdiction. The key difficulty associated with studying civil servants is the lack of reliable performance measures. Politicians need to win elections, and the performance of CEOs may be reflected in sales or stock prices. What the "output" of civil servants is, is much less clear particularly for generalists like IAS officers who work in a variety of departments across their career. We get around this difficulty by polling a group of stakeholders who operate in the same state as an IAS officer and elicit their perception of the effectiveness of that named civil servant. The key stakeholders we survey include IAS officers, state civil servants, elected politicians, representatives of business associations, local TV and print media, and civil society organizations. For each IAS officer they know, we ask stakeholders to grade them on a 1 (low) to 5 (high) scale for: effectiveness, probity, the ability to withstand illegitimate political pressures, pro-poor orientation and overall rating. We gather this information in the 14 main states of India and cover the majority of centrally recruited IAS officers in each state.

As expected if a lack of promotion prospects is particularly demotivating for officers that enter the service older, we find a negative relationship between officers' perceived effectiveness and their age at entry into the IAS (Figure 2).

Figure 2. Effectiveness score and age at entry. (Click to see larger image.)
figure 1 Notes: Raw correlation between the standardized effectivenesss core and age at entry. Standard errors used are clustered at the respondent-level.

We provide several pieces of evidence consistent with incentives. We find that officers that enter state cadres older and as part of a larger cohort of individuals allocated to the same state cadre in the same year are deemed to be less effective. An older officer that enters in a relatively small cohort will be encouraged by the fact that they will face few competitors and less delays in reaching the higher echelons of the state bureaucracy. The reverse would be true for an older officer entering in a large cohort. In addition, we find that officers which enter the IAS in cohorts with a higher proportion of younger officers are perceived to be less effective. This is consistent with them being disincentivized by having to compete with officers with longer career spans.

We leverage this core finding that officers that enter the IAS older and in larger cohorts are deemed to be less effective to address the question of whether bureaucratic effectiveness affects state-level economic performance in India. We focus on the 1990-2011 liberalization period when extensive reforms were being implemented. We argue that, due to the rule-based manner in which officers are allocated to states at the start of their careers, age at entry and cohort size at entry are not determined by contemporaneous state-level economic performance. This empirical set-up enables us to examine whether having a state cadre which contains a higher fraction of IAS officers which entered older and in larger cohorts adversely affects state-level economic performance.

We find that states containing officers that entered older and in larger cohorts grow less quickly (Figure 3). This is due to effects on the organised industrial and service sectors of the economy which are more dependent on policies controlled by IAS officers. Agriculture, in contrast, which is largely unorganised, is unaffected by the composition of IAS state cadres. Structural change thus appears to proceed more slowly when there is a higher proportion of demotivated, ineffective officers in a state cadre.

Figure 3. State-level GDP per capita and age at entry x cohort size. (Click to see larger image.)
figure 1 Notes: The unit of observation is the state-year. Reporting the partial (residual) correlation between age at entry ◊ cohort size and (real) state-level GDP per capita 1990-2011.

Taken together, we find evidence that the career incentives bureaucrats face influence their effectiveness, and that this has wider impacts on the economic performance of the states over which they have jurisdiction. Our paper thus shines a light on the costs associated with rigid progression rules in public organizations. Given a range of public services from health and education through to the diplomatic services are organized like the IAS, understanding these costs and gaining insights into how bureaucrats might be better motivated represents an important undertaking. Indeed it is central to improving the implementation of public policy, to promoting economic performance and to improving societal outcomes.