RESEARCH DOSSIER
Insurance and Entrepreneurial Risk Taking
Abstract. Entrepreneurship entails a sequence of contingent commitments: entering a market, acquiring productive assets, financing investment, retaining workers, preserving operating capacity, and surviving adverse states. Insurance changes the economics of these decisions through the allocation of downside loss, the cost and timing of external capital, household consumption risk, and incentives for precaution. This research dossier develops an integrated examination of commercial insurance, agricultural and index protection, health coverage, unemployment protection, informal risk sharing, and firm-level financial risk management. Its theoretical framework draws on established expected-utility, credit-rationing, incomplete-contract, and investment-financing results; its empirical assessment distinguishes randomized assignment, quasi-experimental reforms, observational comparisons, and reviews. Experiments in Ghana, China, and Malawi show that contracts with apparently similar risk-reducing purposes can generate different investment and borrowing responses. Evidence from France identifies a business-entry response to unemployment protection, while United States health-insurance research illustrates the importance of coverage portability for occupational choice. The synthesis finds that credible and appropriately timed downside protection can enable socially valuable experimentation, especially where failure threatens consumption or future financing. It also finds that premiums, basis risk, contract restrictions, incomplete claims performance, endogenous prevention, incumbent displacement, and financing substitution can weaken or reverse an apparent benefit. A positive entry or investment response is therefore an intermediate outcome, not a sufficient measure of welfare. The analysis offers claim-specific empirical interpretation and conditional institutional implications without proposing a new scientific theory, collecting original microdata, or asserting a new causal estimate.
Keywords: Insurance; entrepreneurship; entrepreneurial risk taking; business entry; firm creation; investment; innovation; downside protection; expected utility; financing constraints; risk rationing; insurance-credit complementarity; collateral; index insurance; agricultural insurance; health insurance; unemployment insurance; business continuity; risk management; moral hazard; adverse selection; basis risk; prevention; causal identification; productive risk; small and medium-sized enterprises; welfare; international comparison.
Introduction: Risk Transfer and the Entrepreneurial Decision
An entrepreneurial choice is a commitment to a stream of uncertain receipts and expenditures. The aspiring business owner chooses whether to leave employment, how much capital to commit, which technology to use, and whether to remain in operation after a shock. Those choices occur while future revenues, personal consumption, access to credit, and the value of accumulated business-specific assets remain uncertain. Insurance can alter each margin because a promised transfer in an adverse state changes the cost of failure before failure occurs. Its economic effect is transmitted through a contract, the financing environment, and the entrepreneur's available alternatives.
The central question of this monograph is how the transfer of a particular risk changes productive decisions, for whom, and relative to which feasible arrangement. An increase in insurance purchases is evidence of contract uptake. An increase in business formation is evidence of entry. Neither observation independently establishes that more valuable projects were financed, that households became better protected, or that the net social return increased. Those stronger conclusions require information on claims, premiums, product costs, exits, production, displacement, financing, and the distribution of outcomes.
The economic literature supplies complementary perspectives. Mossin's rational insurance purchasing analysis (1968) explains why a risk-averse person may pay for contingent protection. Ehrlich and Becker (1972) develop the relation between market insurance and self-protection. Froot, Scharfstein, and Stein (1993) analyze financing and corporate risk management when funds from external investors are costly. In entrepreneurship research, these mechanisms converge because the owner often supplies labor, capital, and household wealth to the same risky undertaking.
The dossier examines six connected dimensions: the nature of entrepreneurial risk and financial constraints; risk preferences and insurance demand; investment and entry responses; credit and collateral; prevention, information, and contract quality; and the interpretation of international evidence. Agricultural households are included because they operate production enterprises under observable weather shocks and have been studied through credible experimental designs. Their evidence informs mechanisms but cannot, without additional assumptions, be treated as a universal estimate for technology startups, manufacturers, sole proprietorships, or large corporations. The analysis evaluates commercial property, liability, credit-linked, index, health, and unemployment protection according to the loss and decision each instrument actually addresses.
Sources are evaluated at the level of research design and identifiable claims. Journal publisher records and open research manuscripts establish bibliographic identity; published abstracts and available article text support the findings specified here. The mathematical results below are transparent applications of established economics. Interpretive conclusions are original analytical synthesis, not newly instituted theoretical systems. The discussion is international and educational and does not supply individual insurance, financing, or business advice.
Entrepreneurial Risks and Capital Constraints
The Economic Object: Entry, Scale, Survival, and Experimentation
Entrepreneurial risk taking encompasses at least four distinct choices. The extensive margin is whether an individual establishes or acquires an enterprise. The intensive margin is the amount invested or the scale of production. A technological margin concerns whether to adopt a process or product with an uncertain return. A continuation margin concerns whether to preserve or liquidate an operating business after an adverse state. These margins can respond differently to one insurance intervention. A contract that protects equipment may stabilize continuation without changing the incentive to leave salaried work; unemployment protection may affect entry without covering physical assets; rainfall insurance may change crop selection while offering no protection against an output-price collapse.
It is analytically useful to distinguish risks associated with business operations, the owner's personal economic security, and financing arrangements. Property damage, interruption, product liability, injuries, weather, customer nonpayment, and cyber incidents threaten enterprise cash flow. Illness, loss of dependent health coverage, and household consumption floors influence the individual's willingness to bear business risk. Collateral loss, covenant breaches, and adverse credit terms transmit the same shock into financing costs. A single event can cross these categories. For an owner-operated business, flood damage can destroy inventory, reduce household income, and impair the collateral used to secure a loan.
The entrepreneur may bear risk even where expected operating profit is positive. A project can have a favorable expected financial value and still be rejected when a failure state forces the sale of a home, the withdrawal of children from education, or long-lasting credit exclusion. This possibility matters most when wealth is concentrated in one enterprise and social or family obligations limit the ability to shift consumption across dates. In their empirical work on weather risk and agricultural investments, Rosenzweig and Binswanger (1993) connect wealth, risk exposure, and portfolio composition, illustrating how poorer producers can select safer, less profitable asset portfolios.
A useful accounting distinction separates expected project output from the economic cost of bearing its risk. Expected output is determined by technologies, demand, prices, and production probabilities. The risk-bearing cost includes the shadow value of scarce liquidity in bad states, costly external borrowing, lost continuation options, and the household's aversion to large consumption declines. Insurance can reduce these additional costs without changing the project's productive technology. It can also alter incentives, input choice, or financing, in which case observed production itself changes.
Consumption Risk, Wealth, and the Entrepreneurial Portfolio
Under expected utility, the relevant object is the distribution of consumption rather than an expected cash receipt in isolation. Let entrepreneurial activity be indexed by a choice x, and let state s occur with probability p_s. Write final consumption as c_s(x) = w + y_s(x) - k(x) - d_s(x), where w represents initial resources, y_s business income, k current investment cost, and d_s debt, damages, or other obligations. A risk-averse owner chooses x to maximize the sum of p_s u(c_s(x)), with increasing concave utility u. A large downside loss can have a high marginal welfare cost because utility's marginal value rises when consumption falls. Risk insurance shifts resources toward some of those states, changing the optimal x.
A household's entrepreneurial portfolio includes financial and human capital together. Starting a business often concentrates earnings, savings, and labor-market exposure in one activity. Diversification through public equity, savings, alternative employment, social insurance, or family transfers may reduce the value of an additional policy. Conversely, weak insurance and limited borrowing may make the business owner choose a technology with lower average profits but less exposure to bad weather, illness, or liability. Dercon and Christiaensen (2011) investigate consumption risk and technology adoption using Ethiopian panel data and historical rainfall distributions; their analysis links vulnerability to consumption shortfalls with reluctance to use fertilizer. Its identification strategy is informative about this setting and does not turn a simulation of absent insurance into an observed insurance treatment effect.
Risk preferences are heterogeneous and context-dependent. Binswanger's (1980) experimental risk measurement in rural India shows that elicitation methods matter, while the economic stakes faced by respondents can alter the interpretation of hypothetical answers. Wealth, ability to borrow, trust, prior shocks, and available fallback employment also influence entrepreneurial choices. A preference estimate derived from one lottery should not be imposed mechanically on households facing complex business continuity risks.
Financial Frictions, Cash Flow, and Continuation Value
Financing constraints make a loss relevant to projects that would otherwise be feasible. A firm may generate high expected returns but lack the liquid funds required to purchase supplies, retain workers, restore inventory, or finance a new contract. If equity issuance or additional credit becomes expensive in adverse states, transferring cash to those states can increase project value. The risk management and investment framework of Froot, Scharfstein, and Stein (1993) establishes this logic in a corporate setting. Its direct instruments include hedges; insurance applies the same state-contingent-financing principle where an insured loss and indemnity are contractually specified.
Consider a firm with internal cash A_s after a shock and a worthwhile continuation investment K. If external funding costs rise as the financing gap max(0, K - A_s) increases, the marginal value of extra cash is greatest in states where A_s is low. A valid indemnity I_s can reduce the gap, thereby preserving continuation even when the claim payment merely replaces lost property and creates no accounting profit. The result depends on payment timing, deductibles, exclusions, coverage limits, claim verification, and the lender's rights over proceeds. A payment received after the critical production date may provide compensation yet fail to prevent closure.
The financial mechanism also explains why full coverage is not necessarily optimal. Premium spending can remove liquidity before investment begins; insurance can leave correlated or uninsurable losses; and contractual incentives may raise effort or monitoring costs. The enterprise compares the expected benefit of downside funds with premiums, retained risk, transaction costs, and effects on financial flexibility. The problem has an empirical dimension: observed demand for insurance may be low because it is too expensive, because the actual protection is unreliable, or because borrowers already receive limited-liability protection from their lenders.
Insurance and Risk Preferences
Expected Utility and the Coverage Decision
A simple two-state model makes the conditional nature of insurance's investment effect explicit. Let project failure occur with probability q, generating a monetary loss L relative to the favorable state. Without insurance, expected utility equals (1-q)u(w) + q u(w-L). A fully honored indemnity I=L purchased for a premium P yields u(w-P) if the policy completely equalizes wealth across the two states. Under actuarially fair pricing, P=qL, Jensen's inequality implies u(w-qL) is at least as large as the uninsured expected utility for concave u. The comparison establishes the familiar value of risk pooling, given an exogenous loss and fixed behavior.
Entrepreneurial choices require an extension: the probability and magnitude of loss can depend on the project selected and the entrepreneur's conduct. If the owner chooses an investment x, covered loss L_s(x), premium P(x), and state payment I_s(x), the optimization becomes max_x sum_s p_s(x) u[w + y_s(x) - k(x) - P(x) + I_s(x)]. A contract can affect risk taking through the smaller downside penalty, the premium's upfront liquidity cost, and the dependence of premium or benefits on x. The sign of the net change in x follows from the shape of the objective and the contract. It is not established by concavity alone.
Mossin's (1968) insurance purchasing benchmark and the later insurance-equilibrium literature clarify why observed premium loading matters. A price exceeding the expected indemnity may be worth paying when it protects severe consumption states, but a sufficiently high loading suppresses demand or forces less complete coverage. Contract menus with deductibles and limits can be privately efficient where losses are expensive to verify or where incentives matter. For an entrepreneur, the economically relevant deductible includes both a nominal monetary retention and the additional operating loss caused by delayed or insufficient cash.
It is also necessary to distinguish ex ante protection from ex post return. An insured business may receive no claim in a favorable year yet gain from a lower probability of financial collapse. An insured business may receive a large claim after a loss while still experiencing severe net damage. Neither claims frequency nor the size of indemnity by itself measures the policy's contribution to investment, welfare, or business survival.
Formal and Informal Risk Sharing
Risk sharing often precedes formal insurance. Family transfers, village reciprocity, state-contingent loan repayments, occupational benefits, and informal credit can partially insure business income. Townsend's (1994) village insurance evidence examines whether individual consumption tracks idiosyncratic income shocks after accounting for village-level resources. The findings provide a benchmark for substantial informal smoothing, with important limitations in full risk pooling. Udry's (1994) analysis of rural credit and insurance demonstrates that loan obligations in northern Nigeria can depend on the circumstances of both borrower and lender. A formal policy introduced into such a setting may complement, displace, or duplicate existing risk sharing.
This matters for credit-linked insurance. Default, rescheduling, forgiveness, and enforcement already determine who ultimately bears the bad-state loss. Borrower-facing insurance has little marginal protection value when repayment is already fully contingent on crop failure; it may have considerable value where default threatens collateral, reputation, future lending, or personal assets. Lender-facing coverage may expand credit supply even when the borrowing household does not perceive an immediate change in its own downside exposure. The distribution of indemnity rights must therefore be identified before concluding that a credit-insurance product should encourage adoption.
State-funded unemployment and health insurance address risks that commercial enterprise policies generally do not: income interruption during experimentation and the loss of access to medical coverage when leaving an employer. These instruments can change entry independently of the insurance purchase decision. Their financing through contributions or taxes creates incidence questions that differ from an individually rated property premium.
Risk Perception, Demand, and Contract Comprehension
The value of a contract perceived by an entrepreneur depends on trust in payment, comprehension of excluded causes, current liquidity, and attention to low-frequency loss. The randomized marketing experiments in India analyzed by Cole and colleagues (2013) show price and nonprice insurance barriers, including trust, liquidity and limited salience. Demand rises or falls with the policy's institutional delivery, even where the insured peril is salient to production. This evidence cautions against using low take-up as a sufficient measure of low actuarial or social value.
Insurance literacy is more specific than general financial literacy. An entrepreneur needs to know which event triggers coverage; how deductibles, waiting periods, exclusions and aggregates work; how to document a claim; and whether the insurer pays before liabilities or restoration bills fall due. Index contracts add the distinction between a measured regional indicator and an individual's actual loss. The cost of learning and verifying these terms becomes part of the economic price of protection.
Observed willingness to pay is also conditioned by market eligibility. Policies may be unavailable in particular sectors, regions, firm sizes, or risk classes. The absence of coverage can reflect supply limitations, minimum-premium thresholds, licensing, or market capacity rather than a deliberate choice to self-insure. When estimating demand, the analyst should distinguish being offered a contract, being eligible to buy, purchasing it, renewing it, and receiving a timely valid claim payment.
Investment Incentives under Coverage
Project Selection and Productive Experimentation
Insurance influences project choice by changing the payoff distribution of failure relative to success. Consider two feasible investments. A conservative activity produces comparatively predictable earnings; a new technology delivers a greater expected return with a greater probability of a severe cash shortfall. If losses from the risky project threaten subsistence or destroy the ability to continue operating, a risk-averse or liquidity-constrained entrepreneur may rationally select the conservative activity. Reliable coverage against a relevant bad state can improve the relative valuation of the experimental activity. This response is economically meaningful even if the insured event never occurs during the observation period.
The mechanism requires specificity. A project has technological risk, commercial demand risk, management risk, and financing risk. A rainfall policy may insure a weather-induced component of output loss while leaving the value of crop sales or managerial errors exposed. A property policy may reimburse damaged capital but leave the revenue foregone during restoration only partially covered. General liability coverage can limit financial ruin from an eligible judgment while leaving compliance expenditures, litigation time, and reputational damage with the firm. The investment response depends on the protected component of the loss distribution, not on the name or nominal face value of the policy.
Insurance also can affect incentives to invent and test products. When entrepreneurial experimentation creates information for future entrepreneurs, the private entrant may bear downside costs while some knowledge benefits spread to others. Downside protection can support experimentation, but evidence of a higher number of attempts does not reveal the value of those spillovers. A socially valuable experiment may fail commercially; a subsidized entrant may compete mainly by reallocating customers from an established firm. The welfare assessment must consequently combine the number and composition of entries with their productivity, survival, employment, and effects on incumbent businesses.
Agricultural Evidence from Ghana and China
Karlan, Osei, Osei-Akoto, and Udry (2014) conducted experiments in northern Ghana allocating cash grants, rainfall-index insurance, or combinations of the two. Their experimental analysis of credit and risk constraints finds that insurance encouraged greater agricultural investment and riskier production choices. Within the study design, the result supports the proposition that uninsured risk can itself restrict productive investment, rather than insufficient cash always being the sole binding obstacle. Insurance purchase and subsequent investment respond to trust and payout experience, so treatment assignment, actual uptake, and realized claims remain separate variables.
The Ghana evidence is important because it identifies a decision made before harvest and before the eventual payout. If protection changes input expenditure at planting, the treatment has modified ex ante investment incentives. This mechanism is more specific than a general association between insurance-market development and economic growth. It also shows why grants and protection should be tested separately: a cash grant improves the initial balance sheet, whereas insurance reallocates resources conditionally across future weather states. Their marginal effects can be substitutes or complements depending on the household's shadow value of cash and bad-state consumption.
Cai (2016) evaluates the provision of agricultural insurance through a Chinese program using quasi-experimental difference-in-differences and triple-difference methods. The published study of production and financial decisions reports that insurance provision increased insured-crop production by 16 percent and borrowing by 29 percent in the study setting, while total household savings did not change in the same way as the composition of more flexible savings. These are estimates of an identified local policy exposure, not universal elasticities for commercial insurers or for every form of entrepreneurship. Their relevance lies in the simultaneous movement of production and credit use, consistent with insurance affecting both productive choice and financial capacity.
The two studies support a conditional inference: when an insured exposure is economically important, coverage is usable and the entrepreneur retains an incentive to invest, risk transfer can release resources or willingness to undertake productive activity. This does not establish that a fixed premium subsidy will always produce such an effect. The evidence has to be interpreted together with the underlying coverage, market institutions, household alternatives, and the method by which treatment reached beneficiaries.
Counterevidence from Malawi and Interpretation of Heterogeneity
A particularly instructive counterexample is the Malawi field experiment of Giné and Yang (2009). Farmers were offered credit for improved seeds; the treatment arm received a similar loan bundled with mandatory rainfall insurance that could forgive a portion of the loan following poor rainfall. In the published insurance-and-credit field experiment, take-up was lower by approximately 13 percentage points for the insured loan, while 33 percent accepted the uninsured loan. The authors discuss the possibility that ordinary limited liability already offered borrowers implicit protection against bad harvests, making the explicit premium costly relative to incremental protection.
The result does not show that insurance is intrinsically harmful to investment. It establishes that the marginal contract mattered: protection was added to a loan that already embedded state-dependent default rights, and the farmer faced an additional cost. Mandatory bundling altered the choice set and could change how product risk was perceived. It is therefore a valuable test of the proposition that increasing the amount of named insurance automatically relaxes an economic constraint. The entrepreneur responds to the overall loan-and-insurance contract, including implicit default insurance, collateral exposure, pricing, and information.
Taken together, Ghana, China, and Malawi rule out a context-free prediction about the sign of investment effects. In Ghana, the experimental intervention addressed meaningful uninsured rainfall exposure and increased investment. In China, insurance provision was associated with higher insured production and borrowing under a quasi-experimental design. In Malawi, bundled insurance reduced credit uptake relative to a loan with existing implicit risk sharing. The comparison reveals plausible mechanisms for differences; it is not a pooled causal effect because the treatments, populations, outcomes, and counterfactuals are distinct.
Downside Protection and New Business Formation
Agricultural investment is one form of enterprise risk taking. Business creation supplies an additional extensive-margin test. Hombert, Schoar, Sraer, and Thesmar (2020) study a French unemployment-benefit reform that provided downside income protection to unemployed people starting businesses. Their evidence on entrepreneurial entry in France reports increased firm creation, with new entrants initially smaller but subsequent survival, employment growth, and productivity measures comparable to those of earlier entrants. A central qualification is that jobs created by new firms displaced employment at incumbent firms almost one-for-one in their analysis, although the new businesses were more productive. The evaluation therefore distinguishes entrepreneurship's private entry response from its net labor-market effect.
This study addresses a specific institutional mechanism: insured income during entry and early operation lowers the cost of trying self-employment. It is not evidence that a commercial property policy has the same effect. The payout entitlement is tied to a social-insurance reform and a particular labor-market transition; consequences depend on eligibility, benefit duration, financing, and the structure of the labor market. Nonetheless, the causal question is highly relevant to the general claim that personal downside protection can change the willingness to establish a business.
The French findings also illuminate the welfare problem. An additional entrant may create productive knowledge, improve allocation by replacing a less productive incumbent, or earn a privately attractive return after shifting some failure costs to a public system. More firm registrations alone cannot sort these outcomes. The evaluation needs business-quality measures, fiscal costs, incumbent adjustments, and the treatment's effect on the income security of both successful and unsuccessful entrants.
Credit and Insurable Collateral
Contracts, Loan Repayment, and Risk Rationing
Credit and insurance interact through the assignment of bad-state losses. A bank may deny a loan because collateral is insufficient, because the borrower may default after a shock, or because the interest rate itself alters incentives and the quality of the applicant pool. Stiglitz and Weiss (1981) examine credit rationing with imperfect information, a foundation for understanding why the price of credit need not clear a risky lending market. Insurance can improve recovery prospects or reduce exposure to identified physical hazards, but it does not resolve every informational, enforcement, or aggregate risk that limits lending.
A simplified credit contract shows the distinction. Let the lender advance K and receive repayment R when the project succeeds. In failure the lender recovers collateral V and may receive eligible assigned insurance proceeds J. Expected gross recovery is (1-q)R + q(V+J), subject to legal, practical, and solvency limits. Increasing J can make lending feasible at a given price if it raises expected recovery and meets liquidity or capital requirements. If the borrower instead receives I and can freely spend it, the lender's effective recovery might rise only indirectly through better continuation prospects. The legal identity of the beneficiary and the priority of proceeds matter to the economic mechanism.
The borrower's expected utility is also separate from the lender's recovery. When limited liability caps personal repayment after failure, an additional loan-insurance premium may reduce the borrower's welfare without meaningfully changing bad-state consumption. This possibility explains why the Malawi experiment needs analysis from both sides of the contract. A lender can value insurance because it stabilizes its balance sheet even when the borrower has limited incentives to purchase the same coverage. Conversely, personal guarantees and pledged household assets can make coverage valuable to the borrower even where the enterprise is legally incorporated.
The relevant financing constraint can be interpreted as quantity rationing, price rationing, or risk rationing. Quantity rationing occurs when external capital is unavailable at desired scale. Price rationing occurs when external financing is offered at an uneconomic cost. Risk rationing arises when the borrower declines available debt because default would impose an unacceptable downside burden. Insurance can affect each margin differently. The contract may lower lender loss and raise loan supply, lower the owner's default risk and raise demand, or simply add an upfront premium and reduce near-term liquidity.
Collateral, Cash Flows, and the Timing of Indemnity
Physical collateral is economically useful only if its value and enforceability survive the relevant shock. Coverage on pledged equipment or buildings can preserve the lender's expected recovery and the firm's ability to resume production. This effect is strongest where the insured event is a common source of collateral destruction. It weakens when payouts exclude a major peril, replacement costs substantially exceed policy limits, or the proceeds are encumbered in a manner that prevents rapid restoration of operating capacity.
A basic financing calculation clarifies the timing issue. Suppose a firm must finance continuation expenditure K at date t, using cash A_t and external funds F_t = max(0, K - A_t). A covered loss reduces A_t; insurance increases it only when the indemnity actually arrives. If payout I is received after a financing deadline, it may reduce eventual losses while leaving the investment decision at t unchanged. An identical nominal limit can thus produce different productive results under different claims procedures. The corporate-finance argument in Froot, Scharfstein, and Stein (1993) supports the importance of matching contingent financing to investment opportunities.
For SMEs, the practical contract terms include reinstatement and replacement conditions, loss-adjustment proof, business-interruption waiting periods, coverage for contingent supply-chain damage, liability defense costs, and exclusions for correlated events. These terms determine the risk actually transferred. An indemnity against property loss may stabilize the balance sheet, while separate business-interruption protection is needed to finance some earnings losses. A parameter such as insured value is therefore an insufficient summary of operating continuity insurance.
Risk pooling also confronts correlated loss. A storm can harm many borrowers and damage their common collateral simultaneously; a regional price collapse can affect borrowers without triggering a local property claim. Correlated claims can strain insurers, reinsurers, and public budgets at the same time that firms seek external finance. Effective entrepreneurial protection may require solvent counterparties, diversified capacity, catastrophe reinsurance, ex ante reserves, or access to emergency liquidity. Those arrangements change the probability that a contractual promise will perform in the states when it is most valuable.
Entrepreneurial Health Coverage and Social Insurance
Personal insurance arrangements can determine whether a prospective founder leaves salaried employment. In the United States, employer-linked medical coverage historically made business creation costly for workers with expected health expenses or dependents needing coverage. Fairlie, Kapur, and Gates (2011) study this phenomenon using matched Current Population Survey observations and comparisons around Medicare eligibility. Their analysis of health insurance and entrepreneurship finds patterns consistent with health-insurance demand discouraging entry among workers who lack alternative spousal protection. These designs are informative but remain sensitive to labor-market selection, retirement incentives around age 65, and the identification assumptions specific to each comparison.
This mechanism differs from commercial insurance for the new enterprise. The founder's difficulty is potential loss of personal medical coverage during a change of employment status. A portable health-insurance arrangement can remove that particular exit cost even if it offers no protection against the startup's product or financial risks. Its entrepreneurship effect may therefore operate through occupational mobility and household welfare rather than through the firm's production technology.
Unemployment insurance illustrates a parallel channel. The French reform analyzed by Hombert and colleagues (2020) protected income during a transition to self-employment. Such programs change the downside value of experimentation, while also raising fiscal, eligibility, and behavioral questions. A public scheme can be funded broadly and can promote some entry even where a commercial insurer would face costly selection and monitoring. Evaluating it requires the full fiscal incidence and the performance of firms that would otherwise never have been formed.
The international comparison must retain institutional differences. The meaning of self-employment, employment-based health benefits, contributory unemployment eligibility, insolvency discharge, and personal guarantees varies by country. A reform can be transferable in economic mechanism while not being transferable in estimated magnitude. The causal claim should therefore specify eligibility rules, comparison group, observation window, and outcome rather than exporting a percentage point result across jurisdictions.
Business Continuity and Catastrophe Recovery
Insurance can prevent the destruction of organizational value that is difficult to recreate: customer relationships, trained teams, supplier networks, permits, and local market knowledge. When a shock interrupts business, a timely, payable claim may finance replacement assets and avert a permanent exit. Such a benefit differs from the expected value of physical loss reimbursements because it preserves future operating options. The magnitude depends on restoration time, labor and input availability, and whether competitors or suppliers survive the same event.
Kousky's (2019) review of disaster insurance and recovery finds evidence that insurance can improve recovery while emphasizing a limited rigorous evidence base on mitigation and land-use responses. Its subject includes households and properties rather than providing a stand-alone causal estimate of SME continuity; the general lesson concerns the importance of recovery outcomes and prevention incentives. A claim paid months after a disaster may support rebuilding while doing little to preserve working capital during the first critical weeks.
Among pastoral enterprises in northern Kenya, Jensen, Barrett, and Mude (2017) compare a cash-transfer program with index-based livestock insurance using panel data and exogenous variation in participation. Their comparative evaluation of transfers and insurance reports that insurance was associated with greater productivity-enhancing investment, fewer distress livestock sales, and higher income per adult equivalent; cash transfers supported other important household outcomes. The designs and objectives differ, and the comparative costs involve program-specific assumptions. This evidence is pertinent to livelihood enterprise resilience, while direct extrapolation to formal-sector manufacturing or services requires separate support.
Entrepreneurial continuity encompasses a chain of outcomes: retained productive assets, uninterrupted liquidity, resumed trading, survival, employment, and eventual profits. These outcomes should be measured at relevant horizons. A recovery grant or credit facility may substitute for insurance in certain states, while insurance can complement financing by restoring collateral and reducing the risk of nonpayment. Distinguishing these mechanisms is essential when evaluating private insurance, public disaster finance, and social protection within one recovery program.
Moral Hazard and Prevention Tradeoffs
Prevention Incentives and Risk-Based Terms
Insurance changes the division of an avoidable loss between the insured and the insurer. If a business bears the entire cost of an accident, the owner has a direct financial incentive to maintain equipment, enforce safety rules, and protect data. Coverage can reduce the private share of an otherwise identical loss. Ehrlich and Becker's (1972) comparison of insurance and self-protection supplies the classic economic foundation for distinguishing the reduction of loss probability from the transfer of the financial consequences of a realized loss. Risk transfer and prevention can be complements or substitutes according to prices, technology, and contractual incentives.
A transparent expected-cost benchmark illustrates the mechanism. Let e be prevention expenditure, q(e) the probability of loss with q'(e)<0, L the monetary loss, and I the insurer's promised indemnity if a loss occurs. If the business is locally risk-neutral, its own expected burden, ignoring a fixed premium, is C(e)=c(e)+q(e)(L-I). With interior optima, the first-order condition is c'(e)=-q'(e)(L-I). Higher indemnity reduces the marginal privately borne benefit of prevention when the premium and coverage terms do not respond to e. The formula is a textbook incentive comparison; risk aversion, endogenous premiums, contract conditions, limited liability, detection, reputation, and legal duties modify the optimum.
An insurer can condition prices or eligibility on credible safety systems, deductibles, inspection rights, security controls, or professional standards. Such contract features may protect prevention incentives and reduce expected losses. Holmström's (1979) analysis of moral hazard and observability explains the value of informative performance signals when effort cannot be directly observed. In practical insurance, an observed loss is an imperfect signal of conduct because random bad luck also causes claims. Verification and monitoring therefore consume resources and can shift risk back to the insured.
A stronger entrepreneurial conclusion follows from examining productive and preventive risk taking jointly. A well-designed policy may encourage investment in a more innovative technology while preserving incentives to maintain machinery or comply with cyber controls. The relevant question is which dimension of risk is being increased. Productive experimentation can create output and knowledge; preventable losses can destroy resources without generating corresponding value. Insurance contract design is evaluated by its effect on the complete pattern of decisions, rather than by a single summary label of increased risk taking.
Selection, Basis Risk, Insurer Reliability, and Contract Complexity
Insurance markets face informational asymmetry when entrepreneurs know more about their risk types or actions than the insurer. Rothschild and Stiglitz (1976) analyze competitive insurance under imperfect information, in which contract menus and coverage restrictions can arise because applicants have different underlying risk. This literature provides a reason why a uniform premium and unlimited coverage may be infeasible or economically unstable. An entrepreneur entering a new industry may additionally lack a claims history, creating uncertainty for both parties.
Basis risk is especially important in parametric and index products. The contract pays according to an observed index rather than the enterprise's verified loss; the two may diverge. Morsink, Clarke, and Mapfumo (2016) set out measures of reliable index protection, including the risk that severe losses occur without corresponding payouts. For investment incentives, this is more than a measurement problem. If a farmer expects a catastrophic private loss to remain uncovered in some states, the policy may fail to remove the very downside exposure that suppresses adoption. Lower administrative expenses do not automatically compensate for protection that performs poorly when the owner most needs it.
The reassessment of agricultural index insurance by Carter, de Janvry, Sadoulet, and Sarris (2017) emphasizes persistent low uptake without substantial subsidies and the potential value of more reliable indices, better contract design, and complementary risk-resistant technology, savings, and credit. This review supports a portfolio conception of entrepreneurial resilience: insurance works alongside adaptation and financial buffers, and the instruments must be compared by the states they cover and the costs they impose. It does not offer a universal policy effect size.
Counterparty performance is part of actual protection. Policyholders consider the probability of insurer distress, dispute, or delayed settlement. A promised indemnity is economically less valuable if the firm may close before it receives it, or if contractual exclusions make the most relevant causes of loss ineligible. A clear policy, accessible claims process, and sufficiently capitalized insurer can therefore strengthen the entry and investment incentives generated by an otherwise identical face amount. Claim reliability, not nominal coverage alone, is the economically relevant variable.
Contract complexity can itself influence adoption. A bundled product may combine a loan, an index trigger, an underwriting covenant, and a premium payment that is difficult to evaluate. The Malawi experiment in Giné and Yang (2009) demonstrates why consumer choice should be analyzed over complete bundles. The Indian experimental evidence of Cole and colleagues (2013) shows that price and nonprice constraints jointly matter. These results support separate measurement of affordability, understanding, trust, and the incremental loss reduction generated by the particular contract.
Private Returns, Crowding Out, and Social Welfare
The entrepreneur's private return can differ from the social value of the activity. An insurer may transfer losses into the premium pool; a government may finance benefits from taxation; and an entrepreneur may shift some losses onto employees, creditors, neighboring firms, or the environment. Likewise, a new entrant may create knowledge or productive spillovers not captured by its accounts. The welfare effect of increased risk taking depends on these external consequences. A private willingness to invest more is a response to altered incentives, not a complete welfare test.
The French reform provides an empirical illustration. Hombert and colleagues (2020) document increased entrepreneurship alongside substantial displacement of employment from incumbent firms. The finding that entrants were more productive matters for allocation, while the near one-for-one crowding out of jobs limits interpretation of gross job creation as net economy-wide employment growth. A valid evaluation counts displaced activity as well as the new firms' outputs, fiscal transfers, and changes in risk borne by entrants.
A subsidy may be justified where uninsured risk suppresses high-return projects with positive spillovers or where market failures prevent effective contracting. Its net benefit depends on how the subsidy changes adverse selection, incentives, administrative cost, and the distribution of public expenditure. For a subsidized insurance program, total program cost includes subsidy transfers and the social cost of raising public funds; claims paid to entrepreneurs are transfers from the financing side, while administrative spending and real resource losses are economically distinct. The valuation of distributional gains is an explicit social choice and should not be hidden inside a production percentage.
The welfare benchmark therefore requires a specified counterfactual. One should compare a given insurance arrangement with realistic alternatives: precautionary savings, a contingent credit line, a guaranteed loan, public income support, an improved safety technology, an index policy with different triggers, or direct financial assistance after a disaster. The comparison should include the reliability and timing of transfers. Otherwise, attributing all investment made by insured firms to their insurance contract overstates the identified mechanism.
Evidence and Economic Implications
Identification, Estimands, and Measurement
Causal evaluation begins by naming the outcome and the intervention. An insurance offer, price subsidy, compulsory bundle, expansion of eligibility, delivered policy, paid claim, and public benefit reform are different treatments. Entry, investment volume, risk composition, borrowing, technology adoption, business survival, employment, profit, and household consumption are different outcomes. A treatment effect on one of these variables does not mechanically identify effects on all others.
Random assignment of insurance offers can identify an intention-to-treat effect for eligible study participants under successful implementation. It does not directly measure the effect of actually holding insurance when some assigned participants refuse coverage and some controls obtain alternatives. Under appropriate exclusion, monotonicity, and other instrumental-variable assumptions, assignment may help identify a complier-specific effect of coverage. Those assumptions should be stated and tested where possible. In observational comparisons of insured and uninsured firms, differential wealth, productivity, management skill, risk preference, and access to brokers threaten causal interpretation.
Policy reforms provide other sources of variation. The French unemployment-benefit study uses institutional reform exposure to investigate entry, while the Chinese agricultural-insurance study uses difference-in-differences and triple differences. Both require consideration of counterfactual trends, affected populations, policy anticipation, and changes that coincide with the reform. The United States medical-insurance study uses labor-market comparisons and the eligibility threshold for Medicare; discontinuities around an age threshold can also affect retirement and labor-force choices. Estimates should be tied to their own identifying variation rather than put on a common numerical scale without adjustment.
The major empirical sources can be summarized by their directly supported contribution:
| Research setting | Identification or design | Observed outcome relevant to entrepreneurship | Principal interpretive limit |
|---|---|---|---|
| Ghana, Karlan et al. (2014) | Randomized cash and rainfall-risk interventions | Greater farm investment and riskier production with insurance | Agricultural climate exposure, local uptake and basis risk |
| China, Cai (2016) | Natural-experiment differences and triple differences | Reported 16% insured-crop production and 29% borrowing increases | Reform-specific exposure, parallel-trend requirements |
| Malawi, Giné and Yang (2009) | Randomized loan versus insured-loan offers | Insured credit take-up lower by about 13 percentage points | Mandatory bundling and preexisting limited liability |
| France, Hombert et al. (2020) | Unemployment-benefit reform | More new firms, similar later quality, substantial incumbent employment displacement | Social-insurance eligibility and national labor institutions |
| United States, Fairlie et al. (2011) | Panel comparisons and Medicare threshold | Evidence compatible with employment-linked health-coverage barriers to entry | Health selection, age-related changes and observational assumptions |
| Northern Kenya, Jensen et al. (2017) | Program participation variation and panel comparisons | Investment and reduced distress livestock sales under index insurance | Pastoral context and distinct program objectives |
Outcome measurement requires attention to horizons and attrition. A newly insured firm may invest immediately, fail during a later shock, or survive with reduced long-run profitability. Short panels can detect first-period expenditure but miss exits, debt repayment, or changes in the business's competitive environment. Survivorship bias arises when unsuccessful businesses disappear from administrative data; program evaluation must track failures as outcomes rather than simply omit them. Reported business profits can mix returns to the owner's labor and capital, especially among sole proprietors. The measured unit may be an individual, household, establishment, firm, or lender, and each yields a different answer.
Selection into the purchase of insurance creates further challenges. An entrepreneur who anticipates high losses may buy more coverage; an entrepreneur with low risk but high risk aversion may also buy more. The sign of the observed correlation between coverage and claims may therefore be difficult to interpret without preference, risk, and information measures. A firm's decision to purchase coverage may also follow lender demands, new contract opportunities, or improved compliance systems. These channels can confound comparisons if the econometric model treats the observed policy as an exogenous standalone intervention.
Cross-Country Interpretation and Institutional Design
The theoretical and empirical evidence identifies several design conditions that increase the probability of productive responses. The policy should insure a consequential downside loss; the entitlement and claim trigger should be understandable; the payment should arrive at the decision-relevant time; the insured should retain appropriate incentives for prevention; and the premium should not consume the liquidity needed to use the productive opportunity. A market that supplies coverage to a credit-constrained founder at a prohibitive upfront cost may offer valuable actuarial protection without enabling that founder's investment.
The conditions vary by sector. In farming, crop-loss covariance, weather indices, informal credit, seasonal financing, and basis risk dominate. In service businesses, professional liability and income interruption may be more relevant. In manufacturing, loss of physical assets, supplier dependence, and environmental liability can be decisive. In technology firms, cyber damage, intellectual property disputes, and high uncertainty over commercial success require separate contract analysis; ordinary business insurance often covers selected operational hazards rather than failure to find product-market fit. A social-insurance reform operates through household security and labor-market status, while a commercial policy operates through specified insured events.
A comparison of risk-bearing institutions should account for legal enforceability and the location of financial risk. A public unemployment-benefit system financed by broad taxation shares risk across households and time. A mutual insurer allocates claims among members and reserves. A stock insurer accepts defined claims in exchange for premiums backed by capital and possible reinsurance. A credit guarantee may protect lenders while borrowers still face substantial personal liabilities. Such structures can finance similar entrepreneurial activities while distributing bad-state consequences very differently.
The 2026 OECD Financing SMEs and Entrepreneurs Scoreboard documents the international policy concern with SME borrowing costs, lending conditions, and access to finance. The report supplies contemporary institutional context; it is not treated as causal evidence that a specific insurance instrument increases entrepreneurial activity. Correspondingly, cross-country levels of insurance premiums or policy counts cannot identify the contribution of insurance to entrepreneurship without accounting for income, sector composition, regulation, institutional quality, and financing opportunities.
The analytical implications for policymakers and insurers are operational. Product evaluation should document a contract's target risk, the implicit coverage already supplied by loans and social programs, and how indemnity rights interact with the entrepreneur's balance sheet. Claims data should be linked to investment and continuity outcomes where privacy, consent, and suitable governance allow. Index-product assessment should record catastrophic failures of the payout rule. Insurance-credit bundles should be tested against the relevant unbundled and limited-liability alternatives, rather than solely against an imagined credit market with no risk sharing. Public programs should report fiscal incidence, displacement, and productivity as well as new-firm counts.
Original Analytical Synthesis and Remaining Research Questions
The studies collectively support a set of analytical conclusions. First, insurance has a productive option value: credible protection in an adverse state can preserve the ability to make an investment whose profitability depends on remaining solvent. This is an interpretation of standard contingent-finance and expected-utility reasoning, not a newly named financial option or theory. Its strength rises with financing frictions and the severity of avoidable continuation loss; it declines when payments are delayed or irrelevant to the binding exposure.
Second, insurance may change entry through the entrepreneur's personal outside option even when the enterprise itself acquires no business policy. The French social-insurance reform and United States health-insurance evidence demonstrate distinct institutional paths from household downside protection to self-employment. For comparative work, firm coverage and founder coverage should therefore be measured separately. Studies of one channel should not be used as causal proof of the other.
Third, a premium is paid at a particular time, while a claim is contingent on a future state. The product can simultaneously reduce expected downside exposure and tighten initial liquidity. For an entrepreneur facing a binding cash constraint, the upfront premium can make a profitable project infeasible even as its protection would increase the willingness to invest. Premium subsidies, deferred premiums, and financing arrangements have different incidence and incentive effects. The opposing Ghana and Malawi results become comprehensible when the incremental contract, liquidity, and prior protection are made explicit.
Fourth, the quality of insurance lies in the joint distribution of realized loss and timely net indemnity. A high insured sum does not indicate that a policy pays in the entrepreneur's relevant adverse state. Basis risk, exclusions, counterparty solvency, and settlement delay determine whether the policy actually supports productive experimentation. This inference applies across index, property, and liability products, although the precise trigger and measurement requirements differ.
Fifth, additional risk taking has to be decomposed into productive innovation, insured financial exposure, and preventable hazard creation. A contract can increase the first while preserving controls on the third. Contractual deductibles, monitoring, prevention conditions, and risk-sensitive terms can move these incentives, with tradeoffs in cost and accessibility. The socially preferred design depends on the value of experimentation, the observability of effort, the distribution of losses, and the institutions that bear residual risk.
Finally, causal evidence is strongest where a policy or experiment identifies a specified treatment on a specified decision margin. Agricultural investment findings, business-entry reforms, and insurance-demand studies together form a coherent but heterogeneous body of evidence. The appropriate synthesis is a map of mechanisms and boundary conditions, not a pooled claim that insurance always increases entrepreneurship or economic growth. An estimate for a rainfall policy offered to smallholders is informative about production under weather risk; it does not substitute for a randomized evaluation of commercial business-interruption insurance for urban SMEs.
Further research should measure the interaction of policyholder liquidity and premium timing, verify the relationship between contractual coverage and time to actual payment, compare business survival under insurance versus contingent credit, distinguish founder-level from firm-level protection, and examine how insurer capital and reinsurance affect performance during correlated disasters. Longer panel studies should track entrepreneurial entry, investment quality, labor displacement, persistence, exits, owner household welfare, and fiscal costs together. Where a research design cannot identify all these outcomes, reporting the estimand and its limit is a substantive scientific result.
Conclusion
Insurance can alter entrepreneurial risk taking by moving economic resources toward adverse states, reducing the expected personal cost of failure, supporting collateral values, preserving liquidity, and making experimentation financially survivable. The mechanism is strongest when the insured event is central to the entrepreneur's decision and when the entitlement produces credible, timely net payments. It can operate through commercial enterprise insurance, agricultural and index products, protection embedded in credit, or personal social insurance that changes the decision to leave employment. These channels involve different counterparties, institutions, and welfare consequences.
The empirical literature confirms meaningful but heterogeneous responses. Insurance interventions in Ghana increased agricultural investment and riskier production; insurance provision in China increased insured-crop output and borrowing in the studied reform setting; mandatory insured lending in Malawi attracted fewer borrowers than a comparable ordinary loan. A French unemployment-benefit reform increased the number of startups while showing substantial employment displacement from incumbents, and United States health-insurance research identifies a plausible barrier to entry through employer-linked benefits. Each result is interpreted within its own treatment, population, identification design, and observation period.
The decisive economic finding is that the value of insurance for entrepreneurship depends on the complete contract and the preexisting allocation of risk. Premium timing, default rights, basis risk, loss prevention, lender claims, insurer reliability, and effective post-loss liquidity shape the productive response. Evaluating coverage solely by take-up, nominal limits, new-firm registrations, or aggregate insurance penetration misses the mechanisms that explain divergent outcomes. The appropriate scientific standard combines clear contingent-claim analysis with causal design, measured business performance, and explicit welfare incidence. The conclusions constitute an original integration and interpretation of established scholarship; no new theory, proprietary empirical estimate, or externally reviewed certification is claimed.
References
Binswanger, Hans P. (1980). “Attitudes Toward Risk: Experimental Measurement in Rural India.” American Journal of Agricultural Economics 62(3): 395–407. https://doi.org/10.2307/1240194.
Cai, Jing. (2016). “The Impact of Insurance Provision on Household Production and Financial Decisions.” American Economic Journal: Economic Policy 8(2): 44–88. https://doi.org/10.1257/pol.20130371.
Carter, Michael R., Alain de Janvry, Elisabeth Sadoulet, and Alexandros Sarris. (2017). “Index Insurance for Developing Country Agriculture: A Reassessment.” Annual Review of Resource Economics 9: 421–438. https://doi.org/10.1146/annurev-resource-100516-053352.
Cole, Shawn, Xavier Giné, Jeremy Tobacman, Petia Topalova, Robert Townsend, and James Vickery. (2013). “Barriers to Household Risk Management: Evidence from India.” American Economic Journal: Applied Economics 5(1): 104–135. https://doi.org/10.1257/app.5.1.104.
Dercon, Stefan, and Luc Christiaensen. (2011). “Consumption Risk, Technology Adoption and Poverty Traps: Evidence from Ethiopia.” Journal of Development Economics 96(2): 159–173. https://doi.org/10.1016/j.jdeveco.2010.08.003.
Ehrlich, Isaac, and Gary S. Becker. (1972). “Market Insurance, Self-Insurance, and Self-Protection.” Journal of Political Economy 80(4): 623–648. https://doi.org/10.1086/259916.
Fairlie, Robert W., Kanika Kapur, and Susan Gates. (2011). “Is Employer-Based Health Insurance a Barrier to Entrepreneurship?” Journal of Health Economics 30(1): 146–162. https://doi.org/10.1016/j.jhealeco.2010.09.003.
Froot, Kenneth A., David S. Scharfstein, and Jeremy C. Stein. (1993). “Risk Management: Coordinating Corporate Investment and Financing Policies.” Journal of Finance 48(5): 1629–1658. https://doi.org/10.1111/j.1540-6261.1993.tb05123.x.
Giné, Xavier, and Dean Yang. (2009). “Insurance, Credit, and Technology Adoption: Field Experimental Evidence from Malawi.” Journal of Development Economics 89(1): 1–11. https://doi.org/10.1016/j.jdeveco.2008.09.007.
Holmström, Bengt. (1979). “Moral Hazard and Observability.” Bell Journal of Economics 10(1): 74–91. https://doi.org/10.2307/3003320.
Hombert, Johan, Antoinette Schoar, David Sraer, and David Thesmar. (2020). “Can Unemployment Insurance Spur Entrepreneurial Activity? Evidence from France.” Journal of Finance 75(3): 1247–1285. https://doi.org/10.1111/jofi.12880.
Jensen, Nathaniel D., Christopher B. Barrett, and Andrew G. Mude. (2017). “Cash Transfers and Index Insurance: A Comparative Impact Analysis from Northern Kenya.” Journal of Development Economics 129: 14–28. https://doi.org/10.1016/j.jdeveco.2017.08.002.
Karlan, Dean, Robert Osei, Isaac Osei-Akoto, and Christopher Udry. (2014). “Agricultural Decisions after Relaxing Credit and Risk Constraints.” Quarterly Journal of Economics 129(2): 597–652. https://doi.org/10.1093/qje/qju002.
Kousky, Carolyn. (2019). “The Role of Natural Disaster Insurance in Recovery and Risk Reduction.” Annual Review of Resource Economics 11: 399–418. https://doi.org/10.1146/annurev-resource-100518-094028.
Morsink, Karlijn, Daniel J. Clarke, and Shadreck Mapfumo. (2016). “How to Measure Whether Index Insurance Provides Reliable Protection.” World Bank Policy Research Working Paper No. 7744. Washington, DC: World Bank. https://www.gov.uk/research-for-development-outputs/how-to-measure-if-index-insurance-provides-reliable-protection.
Mossin, Jan. (1968). “Aspects of Rational Insurance Purchasing.” Journal of Political Economy 76(4, Part 1): 553–568. https://doi.org/10.1086/259427.
OECD. (2026). Financing SMEs and Entrepreneurs 2026: An OECD Scoreboard. Paris: OECD Publishing. https://doi.org/10.1787/075d8058-en.
Rosenzweig, Mark R., and Hans P. Binswanger. (1993). “Wealth, Weather Risk and the Composition and Profitability of Agricultural Investments.” Economic Journal 103(416): 56–78. https://doi.org/10.2307/2234337.
Rothschild, Michael, and Joseph Stiglitz. (1976). “Equilibrium in Competitive Insurance Markets: An Essay on the Economics of Imperfect Information.” Quarterly Journal of Economics 90(4): 629–649. https://doi.org/10.2307/1885326.
Stiglitz, Joseph E., and Andrew Weiss. (1981). “Credit Rationing in Markets with Imperfect Information.” American Economic Review 71(3): 393–410. https://www.jstor.org/stable/1802787.
Townsend, Robert M. (1994). “Risk and Insurance in Village India.” Econometrica 62(3): 539–591. https://doi.org/10.2307/2951659.
Udry, Christopher. (1994). “Risk and Insurance in a Rural Credit Market: An Empirical Investigation in Northern Nigeria.” Review of Economic Studies 61(3): 495–526. https://doi.org/10.2307/2297901.