The Talisman Test

India's governance splits in two: it succeeds at episodic, low-discretion, mission-mode tasks and fails at routine, high-discretion, transaction-heavy ones. The failures in the routine row are built from millions of small adjustments: the citizen who makes one more trip, the clerk who defers one more file, the broker who smooths one more case. Each adjustment solves a problem for one person on one day. Taken together, they absorb every shock that might force reform and lock the system into a stable bad equilibrium that spreads across actors, sectors, and space.

The paper first asks a simple question: when does each person meet an institution, in what kind of task, and what do they do when it fails? It answers with existing data (the IHDS panel, administrative logs and household surveys) and builds four linked maps: cognitive, scholarship, experience and data. It closes with testable propositions on how success and failure propagate, using complex-systems analytics. It follows the first paper in the Tailwinds series, which forecasts where change moves; this paper explains why change so rarely starts.

The title adopts Gandhi's talisman as its standard. The talisman is a short note Gandhi wrote in his last years, offered as a test for any moment of doubt; it is printed at the front of Indian school textbooks, and generations of Indians know it by heart. It asks the reader to "recall the face of the poorest and the weakest" person they have seen, and to ask whether the step they are about to take will be of any use to that person: whether it will give them something, and whether it will restore to them control over their own life. Applied to governance, the test asks how an arrangement treats the person least able to adjust; that person is where this paper ends.

The split runs along two axes: how often a task recurs, and how much judgment the official must exercise. India performs in three cells and fails in one, and that one cell holds most of what citizens meet day to day.

The typology follows Pritchett and Woolcock's distinction between tasks that are discretionary and transaction-intensive and those that are not, and Kapur's account of a state that both fails and succeeds. The paper's claim is that the bottom-right cell is held in place by its users as much as by its officials.

Every encounter pairs a citizen with an institutional actor inside one task cell. The actor may be the state, a market, a community institution or a hybrid of them; a loan recovery call or a caste council summons is as much an encounter with governance as a visit to the tehsil office. Citizens meet institutions at moments of the life course they cannot skip: birth, schooling, a ration card, inheritance of land, illness, a crime, marriage, old age, death. What separates them is what they can do when the encounter fails.

Each cell of the matrix should be read through these positions; a tailender who is an older widow meets a different state than a young man with the same network.

The same failure shows up twice, once on each side; and it moves between institution types as tasks migrate from state to market to hybrid rails. The clerk's fear of the vigilance file and the citizen's come tomorrow are the same delay seen from either side of the counter.

Adjustment is familiar to anyone who has flown. A flight is delayed and the airline desk has no answer; passengers rebook themselves, wait it out, pay for a hotel, call a friend who works for the airline, or choose another carrier next time. Each move solves one traveler's problem and leaves the airline's schedule as it was, and the complaint that might have changed the airline is never filed. In India, the same moves are made at the tehsil counter, the police station and the clinic, every day, by hundreds of millions of people. Adjustment, in this paper, is the private repair of a public failure.

Adjustment does two things at once. It absorbs each failure where it lands, so the citizen gets the certificate and the clerk keeps the file moving. It also removes the signal that would force a fix, so the failure repeats with the next case and the next.

This is the resilience of a bad equilibrium: the system bends at the edges and never breaks at the center. Persson, Rothstein and Teorell describe corruption as a collective-action trap; this paper adds the citizen's own tolerance as the material the trap is built from.

Governance should work hardest for people who cannot adjust: persons with disabilities, autistic and other neurodivergent children, older people, and the chronically and seriously ill. The hypothesis is that high-income states give these groups a floor (an entitlement in law, a plan or navigator for each person, and a duty on the institution to come to them), while India legislates the entitlement and leaves the adjustment to the family.

The typology predicts this. Their needs are continuous, routine and high-discretion, so they sit in the failure cell, where campaigns and digital rails help least. Adjustment also demands what they lack: mobility, time, a network, a working fingerprint, the stamina to queue. The burden passes to carers, most of them women, who become the tailenders of the tailenders.

The hypothesis needs care in its framing. Rich countries burden these groups too: Herd and Moynihan document heavy burdens in US disability and health programs, and families in England wait for, and appeal against, decisions on EHC plans. The claim is narrower and testable: a floor exists, and the institution rather than the family carries most of the adjustment. India's own record holds counter-examples worth learning from, Kerala's neighborhood palliative care networks among them.

Carer corollary
Where the user cannot adjust, failure passes to the carer, and shows up as lost work, lost schooling and ill health in the household. Test: employment and time use of carers against the presence of a person with disability or long-term illness (IHDS, LASI, NSS time-use survey).

The paper is built as four linked maps, each answering one question and handing its answer to the next. The data map closes the loop by testing the rules that the first three maps generate.

The cognitive map speaks many languages. The same feelings carry their own idioms across India; the list below draws on six Indian languages and is a starting set, open to correction. Readers are warmly invited to send corrections and idioms in their own languages.

A coding scheme for grievances should record the language and phrase as spoken, then map each to a feeling, with the speaker's gender and age recorded alongside.

The four maps say where failure lives; a fifth, simpler map says how hard each encounter is to complete. It scores thirteen common services on three plain measures: how much of the service can be done online, how many times a person must typically appear in person, and how large the front line is that delivers it. The figure places each service by the first two measures, sizes it by the third, and colors it by the task cell of Figure 1.

Online access and visits are the author's indicative coding from the services' public portals and procedures; they vary by state and will be measured in the first empirical pass. Front-line counts are cited where a public source was checked; others are named but not yet counted.

Read together, the map shows the bifurcation at a glance. Services on rails cluster at the bottom right: online, few visits, low discretion. The failure cell piles up at the top left: offline or partly online, many visits, delivered through the largest front lines in the country. Size is the trap; the services with the most counters are the ones citizens must visit most often, and every visit is a place to adjust.

Three families of existing data can measure the four things the theory needs: frustration, adjustment, brokerage and exit. The IHDS panel carries the encounter matrix, because it records both whom households know and whom they trust. Administrative logs record delay from inside the state; household surveys record exit.

One baseline from the first IHDS wave already shows the shape: households reported confidence in the courts at 55%, the police at 23% and politicians at 11%, and about a third had an acquaintance in government (Desai et al. 2010).

Success spreads along platforms; failure spreads along hierarchies and networks. Each rule is stated as a complex-systems rule, with the test existing data allows.

1. Downhill. A failure is absorbed by whoever has the least capacity to exit, so delay flows down to tailenders. Test: grievance closure times and claim success by household network access (IHDS linked to district grievance and right-to-services logs).
2. Exit cascade. When the cost of exit falls below a threshold, the most vocal users leave; quality falls; more leave. Test: private school share against government school learning, by district over time (ASER, NSS).
3. Broker. Brokers earn from failure, so a fix that bypasses them meets resistance, and discretion moves to the next chokepoint. Test: rejected MGNREGA payments and PM-KISAN instalments held back after Aadhaar seeding and eKYC.
4. Fear symmetry. Vigilance pressure on officials raises deferral, which citizens feel as delay. Test: disposal and closure times before and after anti-corruption drives or high-profile suspensions.
5. Islands. Success holds where a task is converted out of the routine, high-discretion cell: run as a campaign, or put on digital rails. Test: the same department's campaign-mode and routine outputs, such as vaccination drives against routine immunization.
6. Neighbor. Reform crosses state borders by imitation where states carry the program, and stops at them where they do not. Test: the border design of Tailwinds 2026-01, applied to right-to-services acts and grievance portals.
7. Threshold. Reform starts only when a shock exceeds what adjustment can absorb, usually a mass-casualty event. Test: event studies of regulatory change after the cough-syrup deaths and comparable crises.

Rules 1 to 4 explain lock-in; rules 5 to 7 explain the rare escapes; the carer corollary (section 4) says where the cost lands when no one can adjust.

The design runs in three passes: classify, link, test. A codebook will assign about forty common services, from certificates and land mutation to the FIR, the ration card, the pension and the drug license, to the four cells and to an actor pair, and will measure each on the complexity map. Households in the IHDS will be linked to district extracts from court, grievance and employment records, and the vocabulary of feelings will be coded into Tailenders' grievance dataset by language, gender and age. Each propagation rule then runs as its own test, so the work can report which rules hold and which fail. The first pass will test the downhill and broker rules, which fit the available data best; the next paper in the series takes the rules into complex-systems models.

This paper stands on the work of others. Warm thanks to the teams at NCAER and the University of Maryland who built and sustain the India Human Development Survey; to the public agencies that publish court, grievance, employment and health records; to the ASER Centre and the International Institute for Population Sciences for ASER, NFHS and LASI; and to the scholars whose work on brokerage, burden, bureaucracy and the commons is cited below. Errors are the author's own.

1. Auerbach, A. M. 2020. Demanding Development: The Politics of Public Goods Provision in India's Urban Slums. Cambridge University Press.
2. Auyero, J. 2012. Patients of the State: The Politics of Waiting in Argentina. Duke University Press.
3. Banfield, E. C. 1958. The Moral Basis of a Backward Society. Free Press.
4. Bussell, J. 2019. Clients and Constituents: Political Responsiveness in Patronage Democracies. Oxford University Press.
5. Chandra, K. 2004. Why Ethnic Parties Succeed: Patronage and Ethnic Head Counts in India. Cambridge University Press.
6. Chaudhury, N., Hammer, J., Kremer, M., Muralidharan, K. and Rogers, F. H. 2006. Missing in action: teacher and health worker absence in developing countries. Journal of Economic Perspectives 20(1): 91–116.
7. Chhatre, A. and Agrawal, A. 2008. Forest commons and local enforcement. PNAS 105(36): 13286–13291.
8. Cleaver, F. 2002. Reinventing institutions: bricolage and the social embeddedness of natural resource management. European Journal of Development Research 14(2): 11–30.
9. Das, J. and Hammer, J. 2007. Money for nothing: the dire straits of medical practice in Delhi, India. Journal of Development Economics 83(1): 1–36.
10. Dasgupta, A. and Kapur, D. 2020. The political economy of bureaucratic overload: evidence from rural development officials in India. American Political Science Review 114(4): 1316–1334.
11. Desai, S. et al. 2010. Human Development in India: Challenges for a Society in Transition. Oxford University Press. pdf
12. Eban, K. 2019. Bottle of Lies: The Inside Story of the Generic Drug Boom. Ecco.
13. Gupta, A. 2012. Red Tape: Bureaucracy, Structural Violence, and Poverty in India. Duke University Press.
14. Herd, P. and Moynihan, D. P. 2018. Administrative Burden: Policymaking by Other Means. Russell Sage Foundation.
15. Hirschman, A. O. 1970. Exit, Voice, and Loyalty. Harvard University Press.
16. Kapur, D. 2020. Why does the Indian state both fail and succeed? Journal of Economic Perspectives 34(1): 31–54. link
17. Kruks-Wisner, G. 2018. Claiming the State: Active Citizenship and Social Welfare in Rural India. Cambridge University Press.
18. Lipsky, M. 1980. Street-Level Bureaucracy: Dilemmas of the Individual in Public Services. Russell Sage Foundation.
19. Mathur, N. 2016. Paper Tiger: Law, Bureaucracy and the Developmental State in Himalayan India. Cambridge University Press.
20. Muralidharan, K., Das, J., Holla, A. and Mohpal, A. 2017. The fiscal cost of weak governance: evidence from teacher absence in India. Journal of Public Economics 145: 116–135.
21. Ostrom, E. 1990. Governing the Commons: The Evolution of Institutions for Collective Action. Cambridge University Press.
22. Persson, A., Rothstein, B. and Teorell, J. 2013. Why anticorruption reforms fail: systemic corruption as a collective action problem. Governance 26(3): 449–471.
23. Pritchett, L. 2009. Is India a flailing state? Detours on the four lane highway to modernization. Harvard Kennedy School Working Paper RWP09-013.
24. Pritchett, L. and Woolcock, M. 2004. Solutions when the solution is the problem: arraying the disarray in development. World Development 32(2): 191–212.
25. Pyarelal. 1958. Mahatma Gandhi: The Last Phase, Vol. II. Navajivan. (Gandhi's talisman.)
26. Stokes, S. C. 2005. Perverse accountability: a formal model of machine politics with evidence from Argentina. American Political Science Review 99(3): 315–325.
27. Stokes, S. C., Dunning, T., Nazareno, M. and Brusco, V. 2013. Brokers, Voters, and Clientelism: The Puzzle of Distributive Politics. Cambridge University Press.
28. Taylor, K.-Y. 2019. Race for Profit: How Banks and the Real Estate Industry Undermined Black Homeownership. University of North Carolina Press.
29. Taylor, M. 2011. 'Freedom from poverty is not for free': rural development and the microfinance crisis in Andhra Pradesh, India. Journal of Agrarian Change 11(4).
30. Tailenders Analytics. 2026. Tailwinds: How governance and grievance change moves across India, and how to forecast it. Tailwinds Working Paper 2026-01. doi:10.5281/zenodo.23005367

Data and laws. India Human Development Survey (NCAER and University of Maryland); National Judicial Data Grid; CPGRAMS and state right-to-services portals; MGNREGA MIS; NSS, including the 76th round on disability (2018); NFHS-5; ASER; LASI wave 1; UDISE+; Census of India 2011; front-line counts from PIB releases (fair price shops, 2023; Common Service Centres, 2025; police stations, BPR&D, 2020), Health Dynamics of India 2022–23 (primary health centers) and an IANS report citing the government (post office passport centers, 2025). Rights of Persons with Disabilities Act 2016; National Trust Act 1999; Maintenance and Welfare of Parents and Senior Citizens Act 2007; Americans with Disabilities Act 1990; UK Equality Act 2010; US Individuals with Disabilities Education Act; Children and Families Act 2014 (England).