Something to complain about: How minority representatives overcome ethnic barriers

We live in tumultuous times. There is a renewed interest and need for policies that help minorities overcome systemic sources of discrimination. A series of important studies in development economics shed light on the consequences of affirmative action policies in public sector jobs and in local and national government (Pande 2003, Besley et al. 2004) . Yet, even while serving in government, minorities face discrimination from their majority group peers. What institutional innovations can the state design to give minorities in government a leg-up?

Formal complaints technologies

Our study, set in Bihar, India, describes one such innovation – a formal complaints technology. Such technologies, common across the developed¹ and developing world, are used by citizens to air their grievances with the local state to members of the higher state. In 2016, Bihar launched its own such system called the Bihar Right to Public Grievance Redressal System (BPGRS). Under the system, once complaints are filed, they are resolved via three-person hearings featuring the complainant, the relevant local government official and a dispute resolution officer from the higher state. 

As it turns out, local government officials would often repurpose this system to air some of their own complaints. In these instances, hearings, therefore, would feature two disputing local government officials and the dispute resolution officer. Why would this be the case?

Two social hierarchies

To understand this better, we need to acquaint ourselves with two key hierarchies amongst local government officials in Bihar:

• The first hierarchy is social and is determined by caste. India’s caste system has long been considered the organising structure of society. Caste is said to determine an individual’s place in the social hierarchy. At the bottom of this hierarchy are the Scheduled Castes (SC), a collection of formerly “untouchable” sub-castes who continue to lag behind others along many socioeconomic dimensions (Banerjee and Somanathan 2007, Deshpande 2011).
• The second hierarchy is administrative. Bihar’s local government comprises 8400 village councils called Gram Panchayats (GPs). These GPs are headed by an elected village head.² GPs are divided into approximately 13 smaller units called “wards”. Wards are relatively tiny – comprising roughly 225 households – and are caste homogenous. In comparison to the powerful village head, ward representatives – called “ward members” – are considerably less important: until 2017, the village head was solely responsible for the village’s development funds.³

The doubly disadvantaged minorities

SC ward members are therefore twice disadvantaged,⁴ occupying the bottom of both hierarchies. Over 80% of such ward members govern under a non-SC village head. Things came to a head when, in 2017, the government of Bihar announced that implementation of key water and sanitation (WAS) public goods would be handed over to ward members. Ward members now had access to some of the village’s financial resources. However, the rules mandated that these funds would reach the ward member through the village head. Thus, if the village head so wished, they could choose to hold up projects by refusing to release funds.⁵

Results

Focusing on SC ward members, our study established three causal results (see Figure 1 for a stylised representation of the findings). We begin by establishing how being quasi-randomly paired with non-SC village heads affects outcomes. To do so, we rely on a regression discontinuity (RD) design strategy.⁶

Public goods provision is affected

Using an administrative dataset of over 98,000 WAS projects, we show that SC ward members deliver an initial 29% fewer WAS public goods.⁷ While things get better with time, the most disadvantaged sub-castes amongst SCs implement fewer projects even a full two years after the scheme had begun. 

Crucially, reversing roles does not have the same effect: non-SC ward representatives have no trouble working with SC village heads. Thus, a combination of the two hierarchies makes SC representatives worse off. 

Figure 1 Three main results

SCs respond by filing complaints

Our second result – identified via the aforementioned RD design – describes the role of formal complaint technologies in this setting. To do so, we match data on nearly 100,000 ward members with the universe of nearly 500,000 complaints filed under the formal complaints system. We find that SC ward members are more than twice as likely to file complaints about WAS public goods when governing under a non-SC village head. Once more, we find no such pattern for non-SC ward members, reinforcing our claim that caste hierarchies lie at the root of the problem. 

These two results, therefore, establish that SC ward members are particularly disadvantaged. Since complaint-filing is not costless (it requires attending multiple hearings in a distant dispute resolution office), it suggests that the breakdowns in collaboration between SCs and non-SCs are significant. 

The formal complaints system improves public goods provision

Does filing complaints improve public goods provision? To understand this, we conducted a large field experiment across 1,629 SC ward representatives.⁸ These come from villages whose total population is nearly 15 million. Our main treatment arm involved contacting these representatives and randomising an offer to file complaints regarding stalled projects in their jurisdictions: 41% of those offered the choice filed complaints.⁹  

An endline survey was conducted over the phone, between 3-4 months after the intervention was run. Treated SC ward members report an additional 6.4 percentage points (24%) increase in WAS project initiation (see Figure 2).¹⁰ There are positive spillovers: neighbours of treated representatives file more complaints (2.5 percentage points) and report more projects being implemented (8 percentage points) than neighbours of control representatives. 

Back-of-the-envelope calculations suggest that the intervention is highly cost-effective, costing 2.5 cents for every dollar's worth of public goods provided.

Figure 2 Impact of intervention on main outcome variables (RCT)

Policy implications

We began by asking how minorities in government can be given a leg-up. Our results indicate that formal complaints technologies could be a possible solution. These technologies give voice, particularly in decentralised set-ups, improving minorities’ strategic bargaining power with upper-tiered members of the local state. This has important, positive implications for public goods provision. 

Our findings, therefore, speak to two linked policy agendas around deepening democracies in developing country settings:

1. By showing how local government officials could effectively repurpose them, the findings suggest a non-citizen-based pathway via which formal complaints technologies make the state more accountable. 
2. They also contribute to the thinking around making decentralisation most effective (Bardhan and Mookherjee 2006), by arguing in favour of an active formal complaints technology to be used by members of the local state.

References

Banerjee, A and R Somanathan (2007), “The political economy of public goods: Some evidence from India”, Journal of Development Economics 82(2): 287-314.

Bardhan, P and D Mookherjee (2006), Decentralization and local governance in developing countries: a comparative perspective, Vol. 1, The MIT Press.

Besley, T, R Pande, L Rahman and V Rao (2004), “The politics of public good provision: Evidence from Indian local governments”, Journal of the European Economic Association 2(2-3): 416-426.

Deshpande, A (2011), The grammar of caste: Economic discrimination in contemporary India, Oxford University Press.

Sharan, M R and C Kumar (2019), “Something to Complain About: How Minority Representatives Overcome Ethnic Barriers”.

Pande, R (2003), “Can mandated political representation increase policy influence for disadvantaged minorities? Theory and evidence from India”, American Economic Review 93(4): 1132-1151.

Sachchidananda (2007), Empowerments of Dalits through Panchayati Raj: The Bihar experience.

Endnotes

1  Over 200 American cities have such systems. 

2 The GP head in Bihar is called a “Mukhiya”. 

3 Subhas Baitha, a 35-year-old ward member, described his state in 2007 thus: “Ward members of the Gram Panchayat have no financial power. They cannot sanction any scheme. [The Village head] is all in all in Panchayat” (sic). (Sachchidananda 2007)

4 Ashok Rajak, a ward member said: “[the village head] helped only his own caste men”. (Sachchidananda 2007)

5 As mentioned, there exists funds for every ward to implement their own projects. Also, allocations are determined by a fixed set of rules. So, all our results abstract away from preferences or rivalries over which wards should get which goods and when.

6 An affirmative action policy to encourage representation of SCs in local government mandates that villages with an SC population above a threshold would only have SC village heads. Thus, villages close to the threshold on either side are similar, but only vary in the caste of the village head.

7 While this gap reduces by the end of year 2, it remains large (27%) and significant for those sub-castes among SCs who are at the bottom of the caste hierarchy.  

8 This study is registered in the AEA RCT Registry and the unique identifying number is: AEARCTR-0004308.

9 An additional treatment measured if information alone was sufficient to increase complaint filing: our results indicate that this increases complaint filing by 7 percentage points. This indicates that the transaction costs of filing complaints are significant. 

10 They are also more likely to say that the main problem that caused projects from being implemented was resolved and that projects are likelier to start in the coming week (Figure 2).