Government Crisis Support: What Recent Programs Have Taught Us

Why the experience with crisis support is not over

On March 16, 2020, Germany’s federal and state governments decided on the first lockdown; it took effect on March 22, and on March 27 the Bundesrat, the upper house of parliament, approved emergency aid. A few days later, the first money flowed. Bridging aid, November and December aid, restart aid, and hardship aid followed, later also support against the consequences of high energy prices. For many owners, this was the first experience of the state paying money directly to businesses on a large scale.

The programs have been paid out but are not finished. Final statements were reviewed for years, emergency aid was partly reclaimed, and the question of whether the support reached the right businesses is only now being answered with data. This piece asks: What have government crisis support programs of recent years taught us, and what follows for an owner who wants to prepare for the next crisis? To answer it, we clarify terms, introduce a mental model with three groups of businesses, review evaluation and research, and weigh two views.

Terms: grant, loan, final statement, overcompensation

• Grant: Money that does not have to be repaid as long as the conditions are met. Emergency aid and bridging aid were grants.
• Promotional loan and guarantee: Money that must be repaid, with the state assuming liability toward the bank, for example through the state development bank KfW. They push the problem into the future instead of solving it.
• Short-time work allowance (Kurzarbeitergeld): A benefit from unemployment insurance that replaces part of lost wages when working hours fall temporarily. It secures employment, not the business’s other costs.
• Final statement: The subsequent accounting in which actual sales and costs are compared with the information in the application. Only then is it clear how much support a business is entitled to.
• Overcompensation: A business receives more than it would have needed based on actual losses. A deadweight effect means it takes the support even though it would have survived without it.

Alongside the payments, there was a measure without money: the obligation to file for insolvency in the event of illiquidity or over-indebtedness was temporarily suspended in 2020 and 2021. For suppliers, this was an invisible danger, because customers stayed in the market longer than their figures would have allowed. How defaults then traveled through the supply chain is described in When a customer becomes insolvent.

The mental model: strong, viable, and weak businesses

Pierre-Olivier Gourinchas, Şebnem Kalemli-Özcan, Veronika Penciakova, and Nick Sander (2025) developed a model that assesses crisis support by how well it is targeted. They divide businesses into three groups. Strong businesses survive the crisis even without support. Weak businesses would have exited even without the crisis. Viable businesses would have stayed healthy without the crisis but fail because of it, because they run out of money. Good crisis support saves the third group without giving much money to the first and second.

The model’s most important finding lies in the middle row. Viable businesses resemble strong ones in productivity and growth. They differ only in cash and debt, and there they resemble weak ones. In a sudden crisis, a business therefore fails not primarily because it is bad, but because it has no reserve. An analogy: two equally good runners start a marathon in which the refreshment stations fail. It is not the faster one who reaches the finish, but the one carrying their own bottle.

In a sudden crisis, it is not a business’s quality that decides its survival, but its cash.

What the evaluation of the COVID support shows

The German Federal Ministry for Economic Affairs had the COVID business support evaluated by Prognos (Astor and colleagues 2025). For all programs except hardship aid, the report arrives at around 75.7 billion euros in funding and administration costs of 1.65 billion euros, or 2.18 percent. For emergency aid, which was initially handled decentrally and partly on paper by the federal states, the share was 3.07 percent. From the bridging aid onward, costs fell thanks to a joint digital platform and the involvement of verifying third parties such as tax advisors.

The authors find that the payment behavior and credit rating of businesses eligible for emergency aid improved from June to August 2020, especially among service providers, in retail, and among businesses with fewer than six employees. But they also name the weakness of the first phase: emergency aid initially did not require reporting of actual revenue losses. Spot checks showed overcompensation to a not insignificant extent, which is why a mandatory reporting procedure was agreed with most federal states from 2023. The authors themselves stress one limitation: the effect of the support cannot be cleanly separated from short-time work allowance, loans, and the suspension of the obligation to file for insolvency.

In the final statements for bridging aid, according to the ministry in April 2025, a reviewed volume of around 14.5 billion euros resulted in a balance of almost 308 million euros in favor of the federal government, a little over 2 percent. For most businesses that had applied carefully, the support therefore stood.

What research shows about targeting

Gourinchas and colleagues (2025) tested their model with balance sheet data from small and midsize companies in eleven European countries. For 2017 to 2019, predicted failure rates were on average only 0.69 percentage points off the actual ones. For the COVID crisis, they calculate that without government support the failure rate would have risen by 6.01 percentage points, with 3.1 percent of employment at risk. Support for viable businesses only would have cost 0.77 percent of economic output.

The instruments actually used were broad. In the model, government-guaranteed crisis loans would have mobilized 6.43 percent of economic output, of which 5.45 percentage points went to strong businesses. Of the businesses saved, 56 percent would have been viable and 44 percent weak. John Lorié and Iulian Ciobica (2021), using a simpler model for seven countries, reach a related finding: support prevented insolvencies at the height of the crisis but increased the share of so-called zombie companies that would not be viable without support. In Germany, this share remained low in the second quarter of 2020 and rose markedly in the third.

Two views: fast and broad or precise and slow?

The first view defends fast and broad support. In a sudden crisis, nobody knows which business is viable. Anyone who checks first loses the businesses they want to save. That money also flows to strong businesses is the price of speed, and a later final statement can reduce it.

The second view calls for precision. Broad support is expensive, distorts competition in favor of businesses that would have exited without it, and creates repayment claims that cost trust years later. The evaluation names the high dissatisfaction caused by rules changed after the fact.

I consider the first view right for the first weeks and the second for everything that follows. The evaluation supports this: emergency aid was fast and imprecise, bridging aid slower and more precise. What matters is that the rules are clear from the start. Anyone who promises support today and changes the conditions tomorrow loses the trust the support was meant to create. How much trust in public institutions is worth economically is discussed in Trust as an economic factor.

What owners can prepare for the next crisis

The most important lesson follows from the mental model: the difference between a viable and a strong business is cash. Government support can make up for this difference, but only after some weeks and with conditions. Your own reserve works immediately and without a final statement.

1. Build a reserve

   Decide how many weeks of fixed costs the business should be able to carry without sales, and build this reserve step by step. How quickly growth ties up liquidity is shown in Growth costs cash.

2. Keep documents ready

   Monthly sales and fixed costs for the last three years, contracts, and employee numbers should be available at any time. Almost every support program asks for comparison periods.

3. Keep a cash forecast

   A rolling thirteen-week forecast shows early when a gap will arise and is the most important document in an application and at the bank.

4. Clarify contacts

   Agree with your tax advisor and bank who does what in an emergency. In the COVID programs, verifying third parties filed the applications.

5. Document and set aside

   Keep all the bases of an application and set aside part of the support until the final statement is complete.

Such preparation belongs in the emergency file described in The emergency file and in the business’s risk list. How businesses get through crises overall is discussed in Entrepreneurial resilience.

Conclusion: what crisis support has taught us

To the guiding question: the government crisis support of recent years has shown that the state can help quickly and on a large scale in a sudden crisis and that the support measurably improved the liquidity and credit rating of small businesses. It has also shown that broad support is expensive, flows mostly to businesses that would not have needed it, and keeps some weak businesses alive. The mental model of Gourinchas and colleagues explains why: viable businesses fail because of cash, not quality, and from the outside they are hard to tell apart from others. The limits of this article: the models are calculations based on assumptions, and the evaluation cannot fully isolate the effect of individual programs.

1. This week: Calculate how many weeks your business could carry its fixed costs without sales.
2. This month: Create a folder with sales, fixed costs, and contracts for the last three years.
3. This year: Set a target for your own reserve and set up a rolling thirteen-week cash forecast.

The best crisis support is the one a business has given itself in advance.