Portugal wanted to reduce insecure employment and encourage large firms to offer permanent jobs. In 2009, it restricted the use of fixed-term contracts in new establishments. Fixed-term hiring fell. But researchers found no evidence that employers replaced those contracts with permanent ones.
Large companies hired fewer workers and opened fewer establishments. Meanwhile, smaller firms operating in the same labour markets gained opportunities to hire. As a result, the reform’s overall effect depends on more than the response of firms directly covered by the law.
Pierre Cahuc, Pauline Carry, Franck Malherbet and Pedro Martins examine this issue in “Spillover Effects of Employment Protection”, published in 2023. Their research shows why restricting one type of contract does not necessarily increase another.
What changed in Portugal
Before the reform, companies could use fixed-term contracts in new establishments without additional justification. From February 2009, that exemption no longer applied to firms with at least 750 employees.
The change was not a complete ban. Large employers could still use fixed-term contracts for temporary needs and in some other circumstances. Exceptions included certain groups of jobseekers, such as the long-term unemployed.
The reform aimed to reduce insecure work and promote permanent contracts. But its direct restrictions applied to a particular group of firms, not the entire labour market.
Fixed-term hiring fell
The researchers compared firms on either side of the 750-worker threshold. This helped them estimate changes in hiring at new establishments of large employers relative to firms outside the restriction.
In their preferred specification, fixed-term hires in new establishments of large firms were about 77% lower than the estimated level without the reform. Other specifications produced declines of between 73% and 86%.
This is an estimated effect for a specific group of new establishments. It does not mean that all fixed-term employment in Portugal fell by 77%.
The authors found no positive effect on permanent hiring in those establishments. In other words, permanent contracts did not automatically replace the lost fixed-term hires.
Image generated using AI
Employers have another choice
The reform followed an understandable logic. If temporary hiring becomes harder, a firm might offer a permanent contract instead.
But businesses have another option: delay hiring or decide not to open an establishment. The researchers estimate that total hiring in new establishments of large firms declined. Large employers also created fewer establishments.
This does not mean worker protection is unnecessary. Protection can reduce the risks associated with insecure employment. However, Portugal’s experience shows why the effects of a specific restriction should not be assessed through fixed-term contracts alone.
Permanent hiring and vacancy creation matter too. Otherwise, a decline in one type of contract can look like progress while a fall in total hiring goes unnoticed.
Small firms responded differently
Employers compete for workers in the same labour markets. When a large firm hires fewer people, other businesses may find it easier to recruit.
The researchers found positive spillovers for small firms operating in the same local and industry labour markets as affected large employers. Smaller companies increased hiring, and some labour demand shifted across firms.
That does not mean every person who missed a job at a large firm found one at a small business. The study estimates effects on groups of employers. It does not track the outcome of every individual vacancy.
In the authors’ model, firms directly affected by the reform accounted for about 15% of employment. The rest of the market had a much larger share. As a result, relatively small responses across many other firms could materially change the aggregate outcome.
Image generated using AI
Why the aggregate estimate changes
The 2023 paper compares two calculations. The first assumes that other firms do not respond to the reform. Under that assumption, estimated total employment falls by about 1.3%.
The second accounts for interactions between employers and the response of firms not directly covered by the restriction. In that model, the negative effect is about 0.1%. The estimates differ by a factor of roughly 13.
Both figures are estimates relative to a scenario without the reform. They are not counts of jobs actually lost. They depend on the methods and assumptions used.
The authors continued their research. In a later 2026 paper, their general-equilibrium estimate of the aggregate effect differs: it shows a small increase rather than a small decline. This highlights how sensitive the size and even the direction of the overall result can be to the model version.
Protection and access to work
Workers face two important questions. How secure is an existing job? And how easy is it to find another one without employment?
Permanent contracts may offer stronger protection. Yet if a restriction on temporary hiring reduces vacancies at some employers, jobseekers and existing employees may be affected differently.
In their 2023 model, the authors also assess effects on the welfare of workers and unemployed people. Those findings depend on model assumptions. They do not imply that every person’s circumstances changed in the same way.
The researchers also considered a hypothetical extension of the restrictions to all firms. In that scenario, small businesses no longer retain the same advantage over large employers. The estimated result differs from a reform that covers only one group of firms.
What can be learned
Portugal’s experience does not provide a ready-made answer for Kazakhstan. Labour laws, economic structures and hiring conditions differ. Local decisions require separate evidence and an assessment of likely effects.
Still, the research raises a broader question: what happens after rules change? An assessment of labour reform needs to consider total hiring, establishment creation and the response of employers not directly covered by the law.
Reducing fixed-term contracts is not the same as creating an equal number of permanent jobs. And the impact on one group of firms does not always match the outcome for the whole labour market.
Elvira Baturova, EconomyKZ Research expert, exclusively for EconomyKZ.org
