Aspirational service delivery: A theory of bureaucratic governance under digitization in India

Background and the shared puzzle

India’s digital governance reforms are more centralized and ambitious than most other reforms. A single biometric identifier underwrites welfare delivery for hundreds of millions of citizens. CCTNS standardizes police record-keeping across thousands of stations. State-level dashboards visualize district performance in near real time. The stated rationale is familiar: Reduce frontline discretion, curb corruption, raise efficiency, and produce a legible state.

The empirical record from local field sites tells a different story. Bureaucrats describe the new tools as imposing rigid compliance demands that are difficult to meet given intermittent connectivity, uneven training, and the messiness of citizen lives. In NREGA, block-level computer operators report deleting beneficiaries to keep dashboards clean. In police districts, officers report adopting predictive policing software not because it solves a local crime problem, but because neighboring districts have it. Across both domains, bureaucrats report that the discretion the technology was meant to remove has migrated upward into platform designers and middle managers who control the data infrastructure.

The puzzle our collaboration addressed is therefore not whether digital tools work, but why bureaucrats continue to comply with, and even champion, technologies that they themselves describe as exclusionary or inefficient. We theorize that compliance signals state aspiration: Bureaucrats are performing the modern, legible, data-driven state that central reformers have asked them to embody, and they are performing it for an audience that is largely lateral and upward rather than downward.

Individual contribution I: Technological signaling in justice (Pabbaraju)

Pabbaraju’s paper examines how digital policing infrastructure — CCTNS, e-FIR portals, CCTV procurement, women’s help desks, and command-and-control centers — has reshaped the citizen-state interaction in Telangana and Andhra Pradesh. Where the welfare case (described below) isolates performative compliance, the policing case foregrounds a complementary strategy: technological signaling, in which departments deploy visible digital infrastructure to project competence to peers and superiors, often independent of whether the technology meaningfully alters the underlying service.

The empirical work draws on three streams of evidence. A state-year panel from 2009 to 2025, assembled from the Bureau for Police Research and Development, NCRB statistics, and the e-PRAGATI dashboard for CCTNS rollout, captures hardware procurement, software adoption, institutional capacity (women police strength, modernization budgets, training budgets), and two outcome families: case registration and case disposal. Two-way fixed effects models with clustered standard errors estimate direct effects and the moderating role of capacity mechanisms. These quantitative findings are supplemented by surveys of street-level bureaucrats and over fifty in-depth interviews with senior officers and civil society actors, with a complementary set of fifty-four interviews with women in Telangana that grounds the analysis from the citizen side.

Three findings emerge. First, digital hardware significantly increased police intake but not adjudication. CCTV procurement is significantly and positively associated with reported violence-against-women cases (p < 0.05); computer procurement shows a similar pattern at intake. Neither hardware family produced statistically significant effects on case disposal rates. Second, capacity mechanisms operate at intake, not at disposal. Training budgets exert a strong positive effect on registered cases (p = 0.003); women police strength operates in the opposite direction (p = 0.028), consistent with diversion through informal mediation. None of these levers register at the disposal stage. Third, CCTV evidence has a distinct catalytic effect at intake: CCTV procurement remains significant once mechanisms are introduced while computer procurement attenuates, suggesting that hardware binds bureaucratic discretion most effectively when it alters the evidentiary economy of the interaction rather than merely the procedural one. The substantive interpretation is that digital policing reforms reliably restructure the moment of citizen–state contact — lowering the cost to women of approaching the state and raising the cost to officers of refusing to file a complaint — while leaving the substantive decisions that produce outcomes (investigation, adjudication, disposal) substantially undisturbed. Technological signaling captures the bureaucratic logic that drives this asymmetry: Visible infrastructure is rewarded; invisible adjudicative work is not.

Individual contribution II: Performative compliance in welfare (Dhorajiwala)

My paper, Byte-Sized Rights: Welfare in Digital India, examines how the 2023 mandate to make all NREGA wage payments through the Aadhaar-Based Payment System (ABPS) affected the world’s largest public works program. The mandate was issued in January 2023 with an initial deadline of one month. At that point, only about 46 percent of payments in my sample were being made through ABPS. Local bureaucrats were thus required to shift the remaining payments to a system that depended on each worker having an Aadhaar number correctly linked to a verified bank account — conditions that were not met for a substantial share of the existing workforce.

The paper argues that local bureaucrats facing this mandate in low-capacity settings respond through performative compliance: Rather than carrying out the slow, error-prone work of seeding and verifying every worker’s Aadhaar–bank linkage, they delete the records of workers for whom this work cannot be completed in time. The dashboard then registers full compliance with the mandate, while genuine workers are quietly removed from the program.
The argument is tested using a factorial difference-in-differences design proposed by Xu, Zhao, and Ding (2025) on a panel of 250 blocks across 16 states from financial year 2021 to 2025. Because adoption rates are endogenous to outcomes, I use a pre-treatment infrastructural capacity index, constructed from Census 2011 measures of financial institutions, electricity, ICT availability, and road connectivity, as a proxy for differential exposure. I employ entropy balancing on covariates and two-way fixed effects with standard errors clustered at the block level.

The findings are twofold. First, the 2023 mandate is associated with a sharp increase in job card deletions in low-capacity blocks: on average, low-capacity blocks deleted roughly 1,360 more household job cards post-mandate than high-capacity blocks (p < 0.01). Second, these deletions had real welfare consequences. The reduced-form effect on per capita wage expenditure is a decline of about Rs. 150 in low-capacity blocks relative to high-capacity blocks (p < 0.01), with a marginally significant decline in employment per household. More strikingly, the relationship between deletions and employment reverses post-mandate: each additional deleted job card was associated with about 15 additional person-days of employment in the pre-policy period (consistent with the removal of fake or duplicate records), but with a net reduction of about 12 person-days post-mandate. This reversal is the empirical signature of performative compliance: The deletions began excluding active, eligible workers rather than cleaning out ghost beneficiaries. The contribution is to introduce performative compliance as a mechanism distinct from the standard explanations in the literature — capacity constraints, institutional inertia, and bureaucratic backlash. Performative compliance is a survival tactic in low-capacity, high-pressure environments that reconciles the apparent paradox of efficiency gains coexisting with stark exclusionary outcomes.

Conclusion

Digital governance in India is often described as a story about efficiency. This collaboration argues that it is better understood as a story about audience and asymmetry: about whom the state is addressing when it digitizes, and about where in the service delivery chain the technology actually binds behavior. The empirical work across welfare and justice shows that centralized digital reforms reliably restructure the moment of citizen-state contact while leaving the substantive decisions that produce outcomes substantially undisturbed. The theory developed here, organized through two complementary 2×2 typologies, names that pattern and locates it within a more general account of bureaucratic governance under digitization. As nations across the global South navigate the fraught terrain of digital welfare and digital policing, the lesson from our work is that the pursuit of legibility through digital means risks replicating, rather than resolving, the inequalities it was meant to address.