Work after COVID: A new regime for independent workers

Eduardo Levy Yeyati, Luca Sartorio 16 April 2020



A spectre is haunting the world – the spectre of COVID-19. The ability of nations to deal with the impact of this pandemic is the most important public policy challenge in decades. In addition to the ineludible health emergency, in a context in which productive lockdowns are reproduced on a global scale, governments face the double challenge of sustaining workers' incomes and mitigating demand shortages that can deepen an economic depression.

To avoid a social debacle and stabilise workers' labour income during the lockdown, two broad policy tools have been tested, sometimes simultaneously: job protection through soft credits and direct subsidies to firms conditioned on avoiding layoffs; and, if the crisis is such that not all jobs can be preserved, income protection through the broadening and strengthening of unemployment insurance to workers. 

Is this the right policy choice? In the developed world, yes. As a recent theoretical paper analysing the economic impact of COVID-19 suggests, in the event of a lockdown of contact-intensive sectors, “full insurance payments to affected workers can achieve the first-best allocation” (Guerrieri et al. 2020). Job protection (as opposed to income protection) may generally hinder the efficient allocation of resources (for example, by inhibiting labour turnover to accommodate changes in technology), but in times of exceptional crisis a policy based solely on income protection can lead to an excessive volume of layoffs, generating permanent collateral damages from a transitory shock. Hence, the benefits of combining employment and unemployment subsidies in the current emergency.

Is this the right policy choice in the developing world? Probably, but it is not nearly enough. There is a critical shortcoming to these income-protection policies: by definition, they cover only formal salaried workers, who account for half of the labour force, at best, in the developing world. There is an important additional distinction within this group of excluded outsiders. Whereas informal salaried workers are an anomaly that can, in principle, be contained with better monitoring, lowered labour taxes, and administrative costs, the ‘self-employed’ or independent workers are excluded legally, by design. And there are a lot of them.

To put this in perspective: whereas in OECD countries only 15% of the employed population are self-employed, in Latin America this number more than doubles. Independents amount to 25% of total employment in Argentina, 28% in Chile and Uruguay, 30% in Brazil and Mexico, and more than 50% in Peru and Colombia. Unlike formal salaried workers, the self-employed are fully exposed to drops in their monthly earnings in tranquil times and much more so during crises. 

For starters, in most countries, labour benefits are tied to jobs. This makes historical sense. Those benefits are the product of decades-old struggles between activity-specific unions and business chambers. As a result, they were instrumented for union insiders and, by extension, for non-unionised salaried workers in the same activity. These benefits include working hours, licenses, on-the-job training, severance payments, and social security, which remains fundamentally contributory in most countries and thus restricted to formal, salaried employees who contribute through payroll taxes, deepening the inequality in access to the pension system.2,3  

Moreover, the income instability of the independent worker is highly procyclical. In a context of a drastic activity collapse, not only do they see their working hours  – and associated labour income – reduced; none of the income-stabilising measures listed above protects them, pushing the independent worker to the welfare line.

Protect people, Austrian style

In 2003, Austria implemented a reform that abolished severance payments. It replaced them with occupational pension accounts to which the employer generates monthly contributions (equal to 1.53% of the worker's salary) to a segregated pension account that accrues to the worker only after they’ve quit or been laid off, thereby providing some limited income protection. 

More precisely, upon termination of an employment relationship, the worker owns the accumulated ‘portable’ pension wealth, which is transferred to the new employer if the worker is re-employed. Otherwise, access to the funds is regulated to avoid depleting the pension account in good times: they are available only after three years of tenure, in case of a layoff, when firm and worker agree to terminate the relationship, or after the end of a temporary contract. By contrast, when the worker quits her firm (or is dismissed for misconduct), she keeps the claim but cannot withdraw the money, to avoid misuse or fraud.

This scheme can be easily extended to independent workers in a way that would allow them to save for rainy days much in the way that Austrian workers do while on the job. Indeed, the Corporate Staff and Self-Employment Provision Act of January 2008 broadens the coverage of the Austrian scheme to include freelancers and the self-employed. Since 1 January 2008, Austrian employers are required to pay 1.53% contributions to a self-employment provision fund for freelance employees.

More generally, one could imagine a new regime for independent workers under which registered workers set up a benefits account. For each payment that the worker receives, a proportional sum is transferred directly by the payer. How much would depend on the scope of the coverage. For example, it would be sensible to allocate some money for sick and maternity leave and holidays, in addition to unemployment insurance. It should not be difficult to establish categories by, say, a 12-month moving average of labour earnings, and to estimate declining contribution factors based on each of those brackets to account for the fact that the pension account is meant to secure an income floor. Similarly, while a time tenure may not apply in all cases, withdrawing rules can emulate the Austrian model: money from this account would be withdrawn only if the worker´s registered income in a predetermined period fell below a threshold level.

A ‘grey zone’ regime: The case of faux independents

A frequent problem in countries with a high proportion of independent work is ‘false’ or ‘bogus’ self-employment, which refers to cases in which firms misclassify what should otherwise be an employment relationship as a relationship between a firm and an independent contractor to avoid taxes and regulations. Many workers lie in this ‘grey zone’ between employment and self-employment: ostensibly, they choose when and where to work; in reality, they are economically dependent on a unique ‘client’ in a liaison virtually indistinguishable from formal salaried employment. These situations are often the result of deliberate illegal practices to avoid employment-related charges and hiring bans, both in the private and the public sectors. They are also close, albeit not identical, to the particular case of the drivers working for ride-hailing companies.

To address this loophole, many countries have extended social protections to individuals in this situation (OECD 2019). In 2012, Portugal broadened unemployment protections to “dependent” self-employed workers. The criteria established to determine a “dependent self-employed person” was that at least 80% of the worker's annual income has to come from a single client (in 2018, the criterion was reduced to 50%, further expanding coverage). On the other hand, Spain has a labour category called “economically dependent self-employed worker” (TRADE is its acronym in Spanish) that allows access to social protections such as health and accident insurance, pensions, and unemployment benefits to workers who depend for more than 75% of their income on a single client.

The strengthening of these intermediate figures is the natural way to extend benefits to faux independents and, coupled with the new regime proposed here, should fill in gaps of a labour-benefits safety net for the whole span of independent workers.

A badly needed second best

Remedies to include no salaried workers are not without their disadvantages. For example, the Austrian model can generate financial inefficiencies: if a worker maintains a long-term relationship with an employer, funds accumulate in an occupational pension account that could otherwise be used for consumption or investment decisions. However, if the regime addresses this caveat by allowing employees to withdraw funds after a certain contribution period while still employed, excessive withdrawals may weaken the income coverage upon termination of employment – a practice that is not unusual in the Austrian case (Hofer et al. 2012). Withdrawal conditions should carefully balance these two risks. 

Similarly, a ‘grey zone’ regime can be used perversely by firms to mask salaried relationships. Eligibility criteria must be clearly defined and easily identifiable, since vague definitions of “dependent self-employment” come at the risk of creating two rather than one grey zone: one between ‘employees’ and this third category of workers; and one between this third category and the self-employed (OECD 2019).

That said, nothing can be more damaging to the independent worker´s wellbeing than inaction. Indeed, the COVID-19 pandemic has only highlighted huge inequalities in the safety nets between developed and developing countries – in particular, those arising from the prevalence of independent workers in developing countries. This crisis will not only widen the schism between countries but within countries as well: between insiders and outsiders of the labour force. This duality should be addressed as part of the policy response to the crisis. If it is not, regardless of the size of the income-protection policies that developing countries can afford, disparities in labour markets will deepen the toll of the corona crisis on poverty and inequality.


Guerrieri, V, G Lorenzoni, L Straub, I Werning (2020), “Macroeconomic Implications of COVID-19: Can Negative Supply Shocks Cause Demand Shortages?”, NBER Working Paper No. 26918.

Hofer, H, U Schuh and D Walch (2011), “Effects of the Austrian Severance Pay Reform. Reforming Severance Pay 177”, in R Holzmann and M Vodopivec (eds) Reforming severance pay: an international perspective, World Bank: 177-194.

Kettemann, A, F Kramarz and J Zweimüller (2017), “Job Mobility and Creative Destruction: Flexicurity in the Land of Schumpeter”, University of Zurich, Department of Economics, Working Paper 256.

OECD (2019), Policy Responses to New Forms of Work, OECD Publishing.


1 In those cases, the employee can choose between receiving the severance payment all at once or applying it toward a future pension (Kettemann et al. 2017).

2 More recently, as countries try to mitigate this inequality by introducing a pillar of universal benefits, the divide between contributing salaried workers and non-contributing independent workers deepens the structural imbalances between contributions and benefits within the system, increasing the contingent social security debt.



Topics:  Covid-19 Labour markets

Tags:  COVID-19, coronavirus, workers, freelancers, self-employed

Dean, School of Government, Universidad Torcuato Di Tella

Labour policy advisor, Ministry of Production and Labor of Argentina; Research Affiliate, Center for Evidence-Based Policies


CEPR Policy Research