Should firms subsidize worker-owned capital? Experimental evidence from India's platform Economy. Guest post by Shreya Sarkar

This is the 9^th in this year’s series of posts by PhD students on the job market.

Can helping workers buy their own productive assets boost earnings and be profitable for firms? Using an experiment with female gig workers in India, I test whether firms can enable worker investment in productive capital through a new contract design – subsidizing down payments.

Worker productivity often hinges on worker-financed, portable capital – assets employees purchase and use on the job but firms typically don't provide. Think of corrective lenses that let a tailor stitch without errors, cushioned footwear that lets healthcare workers work longer shifts, or vehicles that let gig workers reach more jobs. Such worker-owned capital raises productivity, remains valuable across employers, and benefits workers directly.

Yet adoption often stays puzzlingly low. Standard theory suggests why: firms won't finance portable assets that workers could take to competitors, and workers themselves might face liquidity constraints blocking even high-return investments. But what if firms can capture enough surplus to make subsidizing such capital profitable? I test this in my job market paper with female gig workers in India.

The Setting

I partnered with a large home-services platform firm in India where female workers provide salon services at customers' homes. Thus, efficient last-mile connectivity is a key input for worker productivity, making a scooter (a lightweight motorcycle, like a Vespa) essential worker-owned capital. It costs ~$1,200 – about 20 weeks of earnings – and might impact jobs completed and earnings. (Note: Public transit doesn't provide adequate last-mile connectivity, so workers use personally owned scooters or private taxis.)

Baseline data from 1,531 workers revealed a striking gender gap. There’s a large gender adoption gap – 77% of men vs 31% of women use scooters – and, within women, a large preference–use gap (79% would prefer vs 31% actually do use scooters). Women relying on taxis face unpredictable wait times during peak hours, surge pricing, and high costs. This combination – a clearly productive input, underinvestment concentrated among women, and a portable asset – provides an ideal setting to test whether firms can profitably enable worker investment.

The Experiment: Subsidizing the down payment

I enrolled 350 full-time female workers of this firm who applied to buy a scooter. Control workers faced the standard loan (15% down, 85% financed); treated workers received a $120 subsidy paid to the lender at purchase, cutting the down payment from $180→$60 (two-thirds) and implying a 10% price discount. 

I randomized treatment within strata defined by baseline scooter use: "new adopters" (~70% of sample who were not using a scooter at baseline) and "incumbent users" (~30%, already using one). This is crucial: if the subsidy works by enabling a shift from taxis to scooters, effects should be large for new adopters but zero for incumbents whose commute is unchanged. I use 8 months of daily job-level administrative data to estimate effects.

More Earnings, Better Jobs, Same Hours

The subsidy increased scooter purchases by 32 percentage points (49% in the treatment group vs 17% in control group), a 188% increase. Among new adopters, intent-to-treat estimates show that treated workers completed +0.54 jobs/week (+9% increase) and earned +INR 602/week (+13% increase) (Figure 1 Panel (a)). Hours “available” to work on the platform (labor supply) did not change, thus earnings per hour rose by 10%, with gains appearing immediately and persisting over 8 months

Workers expanded spatially: weekly distance rose 6.67 km (+11%), driven entirely by jobs 10–25 km from home (+20%). Nearby jobs showed no change. They shifted into peak-demand windows: early-morning and late-evening jobs increased 12%, while midday stayed flat (these are the time slots with highest unfilled demand and surge pricing). Service quality improved and travel cost fell. 

For incumbent users already on scooters, all effects are near zero and insignificant – confirming the mechanism operates through improved mobility for new adopters.

Figure 1: Impact of subsidy (ITT) on worker and firm outcomes

Note: Panel (a): Worker outcomes (weekly jobs and weekly earnings ITT). Panel (b): Firm outcome (weekly revenue and weekly profit per worker ITT).

For firms, the subsidy pays for itself in 18 weeks

Among new adopters, weekly revenue increased by INR 1,179 and profit INR 265 (+13% increase) (Figure 1 Panel (b)). These gains are not due to  business-stealing. Control workers' job volumes don't decline around treatment; treated workers' gains concentrate in far-distance and late-evening segments where baseline job non-fill rates are highest, converting previously unmatched demand into completed jobs. Thirteen months after rollout, 90% of the workers remain active, with no treatment induced exit despite three competing platforms and offline salon jobs.

Treatment-on-treated estimates imply weekly profit increased by INR 558 per treated worker; the firm recovers the subsidy in 18 weeks (approx. 4 months) with an implied 12-month ROI of 153%. On the worker side, LATE effects show earnings increase of +INR 2,018/week; with loan payments of ~INR 1,040/week, cash flow is positive from week one; and a 24-month ROI: 73% (57% even without subsidy). 

Yet workers were not investing beforehand. Why? Baseline surveys show workers actually over-estimated returns. Misperception about potential gains is not the barrier.

The Mechanism: It's the upfront cash, not the price

With returns this high, why was adoption so low? The subsidy changes two things at once: it makes the scooter cheaper over its lifetime and it reduces the upfront cash needed. To separate price vs. liquidity, I ran an incentive-compatible discrete-choice experiment with 100 new workers interested in buying scooters. Each evaluated four financing options in random order, indicating whether they would purchase under each.

The four contracts were: (1) standard loan: pay 15% down and finance the remaining 85%. (2) Price discount: get 10% off the price, but still pay 15% down payment on the discounted amount – so the total cost falls, yet the upfront payment remains high. (3) Low down payment: keep the full price but cut the down payment to 5%, rolling the difference into the loan – so upfront is much lower, though total cost is higher. (4) Down-payment subsidy: a grant that covers most of the down payment, sharply reducing upfront cost and price.

Down Payment Relief Wins by a Landslide. Take-up under benchmark: 16%. Price rebate (cheaper overall): 31%. Low down payment (more expensive overall): 66%. Subsidy: 95% (Figure 2 Panel (a)). When split by baseline liquidity, among workers who said they could not mobilize INR 15,000 for a down payment, the price rebate barely moved the needle. But take-up in low down payment jumped to 70%. Among workers who could mobilize cash, all alternatives raised adoption, with price rebate and subsidy performing similarly – both beating low down payment because lifetime cost now matters (Figure 2 Panel (b)).

Figure 2: Discrete Choice Incentive Compatible Experiment: Contract Take-Up Rates

Note: Panel (a): All workers. Panel (b): Sample split by liquidity.

In many contexts, worker-owned capital is pivotal. Portable assets (scooters, tools, devices, connectivity) directly shape worker productivity. Firms can profit from enabling it. Despite portability, labor market frictions may let firms capture enough surplus to make subsidy worthwhile. In this study, the subsidy pays back in 18 weeks with an implied ~153% annual ROI. The binding constraint is upfront liquidity. The contract-choice test shows down-payment relief drives 3–4× more adoption than an equal-value price cut. Policy and practice should target upfront cash, not just cheaper credit or lower prices.

A Private sector path to scale

India is already rolling out large public scooter subsidy programs for women (e.g. Tamil Nadu’s 50% discount up to ~$300 for working women), but these are costly and depend on sustained public budgets. My results point to a complementary, self-financing path: employer-mediated down-payment support. Firms advance the upfront cash, workers use standard loans for the remainder, and higher productivity repays the firm within months, so private incentives can help with scaling. The broader lesson: when workers and firms share gains, innovative contracts can unlock investments that benefit both sides – expanding the production frontier without waiting for credit markets to evolve or governments to subsidize at scale.

Shreya Sarkar is a job market candidate at the University of California, Berkeley.