Two Programs, Two Kinds of Firm Incentives
Chile's two main firm-support programs during the COVID-19 shock did not offer the same kind of help, even though both were intended to prevent temporary disruption from becoming permanent damage. The credit-guarantee program changed the price and availability of borrowing; employment protection changed how a firm could keep workers attached while reducing activity. The useful question here is how those rules altered the choices of firms, banks, and the state. Once the difference is clear, later evidence about risky borrowers and bank approvals can be read as an outcome of incentives rather than as a mysterious feature of emergency lending.
Guaranteed credit lowered the price of liquidity without removing the bank
The expanded FOGAPE COVID program operated through the banking system. The state strengthened an existing public guarantee fund, which promised to cover a portion of a bank's loss if an eligible borrower defaulted. That arrangement matters because it is not equivalent to the government handing every firm a grant. The bank still originates the loan, evaluates the firm, and retains some exposure. The public guarantee changes the size of the loss a bank expects to bear; it does not make repayment irrelevant.
The intended use was working capital: money to keep ordinary operations funded during an interruption in sales. The study describes loans that could cover up to three months of sales, with a six-month grace period followed by repayment over 24 to 48 months. Their interest-rate ceiling was 3.5 percent, close to a zero real rate in the period under study. For a firm facing a temporary shortage of cash, those terms made borrowing much cheaper than ordinary credit. The grace period also moved the first repayment pressure away from the most acute phase of the shutdown. The combination affected both the firm's willingness to apply and the amount it could reasonably carry into the recovery.
The rules nevertheless left several decisions in place. Only firms that were current, or no more than 30 days past due, when they applied were eligible. A bank had to conduct a risk analysis and could approve or reject the application. The guarantee was partial and varied with firm size, and a deductible meant that the bank did not pass every loss to the public fund. These features gave a lender a financial reason to care about the borrower's repayment prospects. A bank that approves a poor loan may recover part of the principal through the guarantee, but it can still bear a meaningful first loss, the administrative cost of dealing with default, and the consequences of a weak credit decision.
That creates an important asymmetry between the firm's incentive and the bank's incentive. A riskier firm typically has more to gain from a subsidized loan because ordinary borrowing would be expensive or unavailable. The guarantee and low interest ceiling reduce that private price. The firm may therefore demand the loan precisely when its repayment prospects are less comfortable. The bank sees the same low price from the other side: the public guarantee makes it more willing to lend than it would be under a fully private contract, but the portion of risk it retains still makes screening worthwhile.
The bank's problem is not simply whether a borrower is “good” or “bad.” It is whether the expected return on this particular loan covers the cost of funds, the cost of administering it, and the loss that remains if the firm defaults. A public guarantee lowers the last component but does not make it disappear. The answer can therefore depend on loan size, the firm's existing debt, its sales prospects, and the share of the claim that the bank would still have to absorb. Two firms facing the same advertised interest ceiling can present different economics to the lender if one is much larger, more leveraged, or less covered by the guarantee.
The interest ceiling creates another tension. In a fully private market, a bank might try to compensate for higher expected default by charging a riskier borrower more. A low program ceiling limits that adjustment. That can protect borrowers from an unaffordable price, but it also means the bank cannot use a higher rate to make every risky loan worthwhile. It must instead reject more applications, reduce the amount approved, or rely on the guarantee and other information to keep the loan viable. The ceiling can thus broaden access while making non-price screening more important. The same rule that makes credit attractive to a risky firm can prevent a bank from simply pricing all of that risk into the contract.
Eligibility works as a first filter before the bank's individual assessment. Excluding firms that were already more than 30 days past due removes some borrowers whose repayment difficulty was visible before the pandemic. That rule narrows the program's reach, including for firms that may have needed help most urgently, but it also limits the chance that a temporary liquidity facility becomes a way to finance debts that were already seriously impaired. The bank's remaining decision is still not trivial: being current before the shock does not guarantee that a firm can repay after months of disrupted revenue.
This is why “guaranteed” should not be translated into “automatically approved.” The guarantee changes the bank's loss exposure, while the eligibility gate and the retained share of risk shape what the bank is willing to do with that exposure. A lender can be encouraged to process a large number of applications quickly and still reject the riskiest applications within that flow. It can also respond differently to firms of different sizes because the amount of the loan and the effective public coverage change the cost of a mistake.
The program's design therefore has two distinct effects. At the demand margin, cheap money makes borrowing attractive to firms that would not seek ordinary credit at its usual price. At the supply margin, partial coverage reduces the bank's downside without eliminating it. The eventual borrower pool depends on both effects. Looking only at the public guarantee tells us why credit became available; it does not tell us which firms actually received it.
Employment protection changed operating capacity—and the information attached to a firm
The Employment Protection Law worked through a different bottleneck. It allowed a firm and its workers to freeze an employment contract temporarily or reduce the work schedule while workers received payments from the unemployment-insurance system. The firm could preserve the employment relationship without paying the full wage bill during a period when employees were not working. Access depended on the employment arrangement rather than on a bank's assessment of creditworthiness, and applications were approved when the relevant agreement was in place.
That does not mean the instrument was costless from the firm's point of view. Its cost was largely an opportunity cost. A firm that put workers under employment protection had to reduce its activity in proportion to the work that was suspended or cut. It gave up the revenue and profit those workers might have generated. The choice made more sense when a lockdown or falling sales had already made continued operation difficult. It made less sense for a firm whose activity remained profitable and whose main problem was simply the need for working capital. Timing matters here. Credit is a bridge from a current cash shortfall to future revenue, so its value depends on the firm being able to keep enough activity alive to repay. Employment protection is a bridge from a current operating interruption to a possible restart, so its value depends on preserving the employment relationship without paying for work that cannot currently be performed. If the interruption is short, both bridges may help; if the underlying business cannot recover, credit can leave a larger debt claim while employment support may only postpone a difficult adjustment. The sources do not measure these welfare trade-offs directly, but the distinction clarifies why take-up cannot be read as a simple preference ranking.
The contrast can be stated simply. Guaranteed credit helps a firm keep operating while it borrows against future revenue. Employment protection helps a firm endure a period in which some operation is deliberately paused or reduced. Credit preserves productive capacity by financing it; employment protection preserves the employment relationship by accepting lower current activity. Neither instrument is automatically superior. Their value depends on whether the binding constraint is liquidity, the ability to operate, or both.
The different costs also change what a firm's choice communicates. A firm that applies for cheap credit reveals a desire for liquidity, but that desire is difficult to interpret by itself: a healthy firm may want to build a precautionary buffer, while a risky firm may need cash urgently. A firm that uses employment protection reveals something about current operations as well. It has accepted the loss of some activity in exchange for wage support and employment continuity. That decision can be consistent with temporary external disruption, financial distress, or a deliberate strategy to preserve workers until conditions improve. It is not a complete credit score, but it is additional information about how the firm is responding to the shock.
For a bank reviewing a guaranteed-credit application, that signal can cut in two directions. Employment support may make a firm less dependent on new borrowing because part of its immediate labor expense is covered. At the same time, using the program may suggest that the firm has reduced its operating scale and may have less cash available to service debt. The World Bank study reports that banks were less likely to approve guaranteed-credit applications from firms using employment protection, after accounting for the study's measures of pre-pandemic risk and sales performance. The most careful interpretation is not that employment-program users were all intrinsically bad borrowers. Rather, their program choice was a real-side fact that banks incorporated into the credit decision.
The diagram makes the division of responsibility visible. The employment-insurance system helps absorb the wage consequences of a reduction in work. The bank remains the gatekeeper for credit, and the public guarantee changes the bank's exposure after a loan is made. A firm can use both paths, but the paths do not cancel one another out. Employment protection may reduce the amount of cash the firm needs, while the fact that it used the program may make a bank more cautious about the credit that remains necessary.
This also explains why joint use is not evidence that the programs were perfect substitutes. A firm can need employment protection because its operations are restricted and still need a loan for fixed costs, suppliers, or the restart. Conversely, a firm can borrow without freezing contracts if it expects to keep operating. The combination can reflect overlapping needs without making the instruments identical: wage support may cover one cost while credit addresses another, and the bank still evaluates the loan on its own terms.
A useful way to keep the mechanism straight is to follow three questions. First, what does the firm lose if it does not use the instrument? For credit, the loss may be an inability to finance operations; for employment protection, it may be the wage cost of retaining workers during a shutdown. Second, what does the firm give up when it uses it? Credit creates repayment obligations; employment protection sacrifices some current activity. Third, who makes the final gatekeeping decision? The firm chooses whether to seek support, but the bank decides whether guaranteed credit is approved. The state defines eligibility and shares some losses, while the unemployment-insurance system carries the employment-support payment.
Those questions matter more than a long list of program provisions because they identify the transmission path from design to outcome. A low credit price can increase demand among borrowers who face the greatest private financing difficulty. A partial guarantee can make banks more willing to meet that demand while preserving a reason to screen. An employment program can satisfy part of a distressed firm's need and simultaneously reveal that its operations have contracted. Together, the programs change the composition of the credit market before any aggregate default number is calculated.
The essential design details are therefore the ones that alter choices or information: the credit ceiling and repayment timing, the eligibility gate, the bank's retained loss, the employment program's operating opportunity cost, and the fact that bank approval applies to credit but not to employment protection. The exact administrative sequence matters only where it changes one of those channels. With that map in place, the next step is to ask how pre-pandemic risk entered application and approval decisions, and whether the bank's screening response was strong enough to offset the riskier demand created by cheap guaranteed borrowing.