The platform digital economy has radically transformed labour markets, with food delivery couriers emerging as one of its most visible symbols. This column examines the controversial practice of sidestepping traditional employment regulations and social protections by subcontracting food delivery couriers. The authors suggest that well-intended policies requiring firms to reclassify these workers as dependent employees have the potential to harm the very workers they aim to protect. Such policies need to be carefully paired with measures that boost demand for regular employment.
The platform digital economy is expanding rapidly worldwide. For example, in the EU, the number of platform workers (28 million) has reached nearly the number of manufacturing workers (30 million). This phenomenon has brought about a radical transformation in labour market relationships, as platform companies often sidestep traditional employment regulations and social protections by subcontracting their workers as independent contractors rather hiring them as dependent workers (Silberman et al. 2019, Katz and Krueger 2019, Abraham et al. 2021).
Online food delivery services stand out as one of the most prominent examples of these controversial work arrangements. Pedalling through the streets of big cities with their distinctive oversized backpacks, couriers — commonly known as ‘riders’— have emerged as a visible symbol of an intense debate over whether they should remain self-employed or be reclassified as dependent employees (Frenken et al. 2018). Proponents of the latter policy argue that employee status would grant riders essential social rights, including fixed-work schedules and collective bargaining. Opponents counter that self-employment offers riders greater flexibility in working hours and lower barriers to entry in the labour market, as all it takes to start working is signing up to a digital platform. Like characters described in the classic song by the legendary 1970s California rock group the Doors, these riders seem to be weathering a storm full of uncertainties about their labour conditions.
Recent regulatory efforts reflect these concerns. The EU Council has proposed policy action to “improve the working conditions and social protection of people working through digital labor platforms” (EU Directive 2024). Spanish legislation has gone further, approving the so-called Riders’ Law in September 2021, which establishes the legal presumption of riders as employees.
Previous research on the regulation of flexible work arrangements (Scarfe 2019, Dolado et al. 2022) and the informal economy (Zenou 2008, Albrecht et al. 2009) analysed similar policy issues. However, evidence on the effects of regulating platform work remains scarce, mainly due to lack of good data. To fill this gap, our recent work (Dolado et al. 2025) aims to quantify the effects of the Spanish Riders’ Law by combining administrative and original survey data with a structural model capturing the above-mentioned trades-offs.
Why some riders choose casual over regular employment
A key feature of this reform is that some platforms refused to comply with it and, as a result, accumulated heavy administrative sanctions. 1 This lack of compliance enables us to compare the working conditions between riders operating as dependent employees (i.e. those with ‘regular jobs’) with those of independent contractors (‘casual jobs’) before and after approval of the Riders’ Law .
We find that regular-job riders earned an average wage premium of 18 log points over casual-job riders with similar observable characteristics and hours worked, consistent with the hypothesis that trade unions are able to extract rents for employees through collective bargaining. Despite this wage differential, a sizeable share of riders opt for casual jobs. Figure 1 provides two explanations for this puzzling fact. First, some casual-job riders work longer hours (exceeding six hours per day) than regular-job riders to take advantage of the flexible schedules. Second, other casual-job riders work similar hours to regular-job riders, suggesting that search frictions limit access to regular employment. That is, in contrast to casual jobs, platforms offering regular jobs hire only the number of workers needed to fulfil their demand for food deliveries, and they screen those workers to ensure long run profitable matches.
Figure 1 Distribution of hours worked
To understand these empirical patterns and evaluate policy interventions, we develop a model wherein workers with heterogeneous preferences for leisure — representing differences in home, education, or caregiving commitments — choose between casual and regular jobs, which differ in terms of employability, work schedules, wages, and who pays payroll taxes.
The model captures key policy trade-offs: workers can instantly access casual jobs, which provide control over their work schedules, but this comes at a cost. They have to pay their own payroll taxes and are compensated on the basis of deliveries, rather than by actual hours worked. The latter feature captures the fact that casual-job riders spend a fraction of their work schedule just waiting for orders in front of restaurants, an interval that increases in periods of low demand. Conversely, regular jobs are subject to search frictions, offer fixed work schedules, hourly wages determined through bargaining, and employers who pay payroll taxes.
The calibrated model rationalises the coexistence of both types of jobs. Riders who prefer long hours choose casual jobs due to their flexibility, while those preferring fixed schedules either look for regular jobs or accept casual job offers initially to escape unemployment and then conduct an on-the-job search for their preferred regular jobs.
The labour market effects of the Riders’ Law
Our model accounts for the possibility that a substantial proportion of riders could benefit from policies forcing platforms to transform casual into regular jobs, since they offer substantially higher wages. This rationale is likely behind politicians’ desires to promote regular employment. Thus, we use our framework to quantify the labour market effects of the Riders’ Law and propose a few policy recommendations.
By imposing heavy administrative sanctions on non-compliant platforms, the Riders’ Law raises marginal costs of delivering food orders through these self-employed riders. This reduces the labour demand for casual-job riders, which increases their unpaid waiting times and effectively lowers their hourly pay. Facing lower wages, some casual-job riders transition to regular jobs. Overall, the reform reduces the share of casual employment by 13 percentage points and average wages for casual-job riders by 7%. As more riders search for regular jobs, platforms in this sector create more vacancies. This increases the employment share of the regular sector by 6 percentage points, thus only partially absorbing the casual job losses. Moreover, wages in the regular sector fall slightly because the deterioration of riders’ outside options in casual jobs weakens their bargaining position. We thus conclude that the reform decreases riders’ overall welfare due to employment and wage losses. These figures, however, have to be interpreted as deviations from a growing trend in the demand for online food deliveries since the pandemic, which we take as given and which has led to a cumulated employment growth rate of 25% in this industry since 2020.
Policy recommendations: Tax bonuses for regular employment
The negative effects of the reform stem primarily from the difficulty platforms face in substituting regular jobs for casual jobs. To ameliorate this problem, we propose complementing the original reform with payroll tax bonuses for platforms hiring regular-job riders.
Our simulations show that a moderate payroll tax reduction of 9 percentage points, relative to the benchmark tax rate of 29%, is sufficient to boost the labour demand for regular workers to preserve pre-reform welfare levels. A larger tax cut of 21 percentage points is required to maintain pre-reform employment levels. While this second policy would have larger fiscal revenue implications, it would lead to significant welfare gains for riders.
Conclusions and broader implications
Based on our findings, we conclude that policies aimed at increasing R- employment by raising the marginal costs of the casual sector will be partially passed through to workers in the form of lower employment and wages. These effects extend beyond the casual sector, affecting regular-job workers due to the important role that causal jobs play in workers’ outside option in wage bargaining.
We argue that effective regulation requires a two-pronged approach: addressing misclassification concerns while simultaneously expanding demand for regular employment through targeted fiscal incentives. This balanced strategy can preserve welfare and employment with mild budgetary consequences.
In sum, as digital platforms continue to transform labour markets worldwide, policymakers should carefully consider the complex trade-offs involved in regulation, and seek to design interventions that protect workers without undermining the flexibility and employment opportunities provided by platform work.
Source : VOXeu
