Skill-Badge Decay Trims 9-Week Budget Streaks at Tier 4

In the 2024–25 financial year, several Tier 4 training centres under India's National Skill Development Corporation (NSDC) ecosystem reported something peculiar: learners who had maintained unbroken nine-week budgeting streaks were abandoning them at week six or seven, not because they had failed a test, but because a digital badge they had earned earlier had visibly expired. The question worth asking is whether the badge expiry itself, rather than any change in financial circumstance, is what broke the streak—and if so, what that tells us about how behavioural reinforcement schedules interact with credentialing design in low-income finance training.

The Architecture of a Nine-Week Streak

Most financial literacy programmes targeting Tier 4 learners—small towns and rural blocks outside the district headquarters—are built around short, repeatable modules. Budgeting, savings discipline, debt awareness, digital payment hygiene. The nine-week format is common because it maps roughly onto a quarterly cycle and fits the rhythm of self-help group meetings, which typically convene weekly.

The streak mechanic is deliberate. Learners log a budgeting exercise each week. Complete nine consecutive weeks, and the programme records a "budget discipline" milestone. This is not gamification for its own sake; it is an attempt to convert an abstract behaviour (planning expenditure before income arrives) into something observable and socially legible.

What makes the streak work, behaviourally, is not the final reward. It is the weekly confirmation. Each completed week produces a small, immediate signal: a tick, a badge, a message. In operant terms, this is closer to a fixed-interval schedule with continuous reinforcement—predictable, low-variance, and effective for habit formation in the early stages.

The problem arises when the reinforcement architecture of the badge system operates on a different schedule from the reinforcement architecture of the streak system.

Skill-Badge Decay: A Mechanism Meant for Portability

Digital skill badges—Open Badges 2.0 and similar standards—were designed with expiry dates for a defensible reason. A credential in "GST filing for small enterprises" issued in 2022 may not reflect current portal requirements. Decay signals currency. Employers and training providers wanted badges that degrade gracefully rather than mislead.

In practice, most badge frameworks set validity windows of 12 to 36 months. But in the Tier 4 programmes examined, badge validity was often compressed to match funding cycles: 6 to 12 months, sometimes shorter. A learner who earned a "Basic Budgeting" badge in month one of a programme might see it enter a "decaying" or "expiring soon" state by week seven of a nine-week streak.

This is where the design conflict becomes visible. The streak says: keep going, you are building something. The badge says: what you built earlier is losing value.

The Loss Aversion Asymmetry

Kahneman and Tversky's foundational work on loss aversion established that losses loom roughly twice as large as equivalent gains. A learner who has completed six weeks has accumulated something—six ticks, a partial streak, a sense of momentum. But they also hold a badge that is now visibly fading.

The streak is a prospective gain: nine weeks, if completed. The badge is a retrospective loss: something already earned, now eroding. Under loss aversion, the eroding badge commands disproportionate attention. And because the badge's decay is time-based rather than effort-based, the learner has no action available to arrest it. The one thing they could do—earn a new badge—requires completing the very streak they are now inclined to abandon.

This is not a motivational failure. It is a predictable response to competing reinforcement schedules.

Variable-Ratio Reinforcement and the Tier 4 Context

There is a second behavioural layer. Many Tier 4 programmes, under pressure to demonstrate completion rates to funders, have introduced variable rewards: bonus badges for "exceptional" budgeting entries, surprise recognitions, leaderboard placements. The intent is to mimic the engagement mechanics of variable-ratio reinforcement, which produces high, persistent response rates.

But variable-ratio schedules work best when the base rate of reinforcement is stable. If the baseline—the weekly tick—is being undermined by badge decay, adding unpredictability on top does not increase persistence. It increases attrition. The learner cannot tell whether the system is rewarding them or quietly withdrawing recognition.

Field data from a 2023 pilot in eastern Uttar Pradesh, covering 1,240 learners across 18 centres, showed that streak completion dropped from 71% to 48% when badge validity was shortened from 18 months to 9 months, with no other programme change. The study, conducted by a regional skilling nonprofit, attributed the drop to "credential anxiety"—learners reporting that they felt their earlier work was being "cancelled."

That phrase—cancelled—is worth sitting with. It suggests that the badge was not experienced as a portable credential but as a live record of competence, and that its decay was read as a verdict on the learner, not on the currency of the skill.

What a Better Design Might Look Like

The obvious fix—remove expiry dates—is not available. Currency matters, and funders need to know that a badge issued in 2023 reflects 2023 standards.

A more workable approach is to decouple the streak's reinforcement from the badge's validity. The streak tick should be permanent and non-decaying for the duration of the programme. The badge can carry its expiry, but the streak record should not inherit it. In system terms: the badge is a credential; the streak is a behaviour log. They should not share a decay function.

A second adjustment is to make badge renewal effort-based rather than time-based. Instead of a badge silently expiring at month nine, the learner receives a short renewal task at month eight: a single budgeting exercise that refreshes the credential. This converts a passive loss into an active, low-cost maintenance behaviour. It also aligns the badge's reinforcement schedule with the streak's—both now require weekly or periodic action, not mere waiting.

Third, and perhaps most importantly for Tier 4 contexts: the decay notification itself should be reframed. "Your badge is expiring" is a loss signal. "Your badge is due for renewal" is a maintenance signal. The behavioural difference between the two is not cosmetic. It is the difference between a learner who feels their record is being erased and one who feels their record is being maintained.

The Forward Question

As India's skilling architecture scales—NSDC's own projections point to over 15 million certified learners by 2027—the interaction between credential design and behavioural reinforcement will stop being a niche concern. Every badge that decays is a potential streak-breaker. Every streak that breaks is a learner who does not complete the module, does not earn the next credential, and does not enter the formal financial system with the habits the programme was built to instil.

The design question for the next programme cycle is not whether badges should expire. It is whether the streak—the behavioural engine—should be asked to carry the badge's decay. It should not. The two systems need separate clocks.