Short contracts, a two-speed labour market

The share of fixed-term contracts in hiring has risen sharply in France since the 2000s, driven mainly by the growth of very short contracts, sometimes lasting a single day, in certain sectors.

The share of employees on fixed-term contracts at any given moment has risen much less. The gap is explained by turnover: the same post may give rise to many successive hires. A simple example clarifies the mechanism. A post held all year by the same person on a permanent contract counts as one hire; the same post filled through one-week contracts counts as several dozen, for a single job at any given moment.

The theory of the segmented labour market, formulated in the 1970s by Peter Doeringer and Michael Piore, distinguishes a protected segment and an unstable one, with little movement between them.

French studies of trajectories find real transitions to stable employment, more frequent for graduates and in certain sectors, alongside a share of people kept in a succession of short contracts over the long term.

The short contract therefore lends itself to two readings that coexist in the data: a stepping stone to stable employment for some people, a lasting situation for others. The debate is less about whether these two cases exist than about their relative weight.

Several measures have sought to discourage very short contracts, notably varying unemployment insurance contributions according to a firm's separation rate, introduced in France at the end of the 2010s.

The first available evaluations find measurable but limited effects on contract length, with reactions varying across the sectors concerned. The scheme applies only to certain sectors and certain firm sizes, which limits its reach from the outset.

Part of the turnover stems from genuinely discontinuous needs, notably in accommodation, food service, the arts and social work, where activity varies sharply over time.

The law takes this into account: in sectors listed in the regulations, the so-called customary fixed-term contract (CDD d'usage) allows successive short contracts where it is standard, long-established practice not to use permanent contracts for certain jobs.

Separating that share from the share that substitutes for permanent employment is the hard part of evaluations, and published studies have not settled it.

Contract type says nothing about how stable income actually is. A succession of short contracts with the same employer can provide continuous income, and a permanent contract can be ended quickly.

Analyses based on matched income data rather than on contract type give a noticeably different picture of measured precariousness.