The Tenure Paradox: More Movement, Not Necessarily More Precarity
This series examines how expanding services trade changes labour-market adjustment, with a particular focus on a deceptively simple question: if people spend less time in the jobs they currently hold, does that mean work has become more precarious? The question matters because job duration is easy to count, while the reasons for movement—and what workers gain or lose when they move—are much harder to see. The evidence considered here supports a more careful judgment: services trade can be associated with shorter current job duration without a broad rise in the measured forms of contractual or employment precarity, but that pattern is not the same as proof that every move is voluntary or beneficial.
The subject is broader than exporting firms. Services can be sold across borders, bought from foreign providers, or imported as inputs into other sectors. Their expansion can alter the composition of occupations, the opportunities available in a region, and the local demand for jobs that are not themselves traded. A related European study of offshoring adds another lens: when tasks and intermediate inputs move across borders, some conditions of the jobs that remain can improve while others deteriorate. These are connected questions, not one combined estimate. Together they make it possible to ask not whether trade is simply “good” or “bad” for work, but which outcome changed, for whom, through what route, and with what degree of evidentiary confidence.
Why shorter tenure is not the same as insecurity
Tenure is a snapshot of time. It records how long a person has been in the job they hold when a survey takes place. If average current tenure falls across a group, the most defensible first interpretation is that people in that group are changing jobs more frequently, or that the composition of the group is changing in a way that lowers its average. It is not yet a statement about whether those changes are welcome, forced, or damaging.
Consider two workers who both have short tenure. One may have left a familiar employer for a better match in the same broad occupation. Another may be moving between short contracts because a previous job ended and no stable alternative is available. A third may have entered a growing field recently, while older, longer-tenured workers remain represented elsewhere. The same number—fewer years in the current job—can appear in all three situations. Tenure identifies movement in the background; it does not identify the destination or the experience of the move.
“Insecurity” is therefore a different claim. It may refer to a higher risk of losing employment, repeated spells outside work, involuntary movement into weaker jobs, or a contractual arrangement that offers little certainty. “Job quality” is broader still. It includes the physical environment, work intensity, working-time arrangements, social relations, autonomy, skills, and prospects. A labour market can become more dynamic while improving one of these dimensions and worsening another. Stability can be valuable, but stability alone does not tell us whether a job is safe, developmental, fairly organised, or likely to lead somewhere.
The evidence behind this series begins with the reported association between greater services-trade exposure and shorter average current job duration. That result is important precisely because it refuses to answer the larger question too soon. It tells us that the pace of employment relationships appears to change. It does not tell us whether workers are leaving employment, crossing into different sectors, cycling through temporary contracts, finding better matches, or enduring hidden disruption between observations.
The diagram captures the discipline required by the headline. A tenure change is an observation at the centre of several tests, not a verdict at the end of the analysis. The next step is to examine those tests separately.
The questions hidden inside one number
The first question is whether shorter stays coincide with more people entering or leaving employment. If the answer were yes, a tenure decline could reflect more unemployment or departures from the labour force, which would make a simple opportunity story less plausible. The services-trade evidence does not identify a significant systematic change in these entry and exit shares. That is relevant reassurance, but it is not a guarantee that no individual experienced unemployment. The underlying data generally compare two points about a year apart, so a short spell between them can disappear from the record.
The second question is whether workers are being pushed or pulled across sectors. A sector change can be a route toward a better opportunity, but it can also mean losing job-specific knowledge, accepting a mismatch, or responding to contraction where one previously worked. The evidence does not find a systematic relationship between services-trade exposure and switching into or out of broad sectors. The word “broad” matters: large categories can conceal meaningful changes inside administration, transport, information, or other groupings. A null sector result narrows the interpretation; it does not describe every career path.
The third question concerns temporary contracts. If shorter job duration came with more temporary employment, the pattern would look more like contractual precarity. The report finds no increase in temporary-contract use associated with services exports or imports, and finds a decrease where imported services are used as inputs. This is one of the clearest reasons not to equate shorter tenure with more precarious work. Yet temporary status is still only one observable form of security. A permanent contract can exist alongside heavy work intensity, poor social relations, weak prospects, or a high perceived risk of future restructuring.
The fourth question is what happens to the work itself. The offshoring evidence is useful here because it measures several non-wage conditions separately rather than compressing them into “good job” and “bad job.” Physical safety, time pressure, autonomy, social environment, and career prospects can move in different directions. A reduction in temporary contracts would not cancel out a deterioration in workplace relations; a longer tenure would not necessarily compensate for a hazardous physical environment. The practical implication is that any serious account of trade and work needs more than one stability indicator.
Finally, the question is who and where are represented in the average. Services-trade exposure is not evenly distributed across workers or places. It can be connected to younger or older workers, women or men, urban or rural residents, and different education groups through different trade flows. A changing average may therefore reflect both movement by existing workers and changes in who is entering, remaining in, or leaving a sector. Regions add another layer: workers may move toward new opportunities, while people previously outside the labour force may be drawn into expanding employment. Those distributional changes are part of the subject, not statistical noise to be discarded.
How to read evidence without overclaiming
The sources do not all answer the same question with the same design. The broad services-trade analyses use large labour-force-survey datasets and extensive fixed effects to compare changes across countries, sectors, occupations, regions, and years. For job duration, employment transitions, sector changes, and temporary contracts, the results are reported as associations. The combined pattern makes voluntary movement toward better matches plausible, because the study sees shorter current duration without a corresponding rise in the measured routes most clearly associated with disruption. But the study does not observe each worker’s reason for moving, the wage attached to the next job, or every intervening spell.
The related offshoring study uses instrumental-variable methods for total offshoring, with stronger diagnostics for the full sample and old EU member states. That gives its aggregate working-condition results a different causal status, conditional on the instrument’s assumptions. Its evidence is also more specific: it concerns offshoring of foreign tasks and intermediate inputs in European industries, not the entire set of services-trade flows. Its new-member-state instrument is weak, and its destination-region discussion relies mainly on reduced-form relationships that the authors treat as correlations. Stronger evidence for one result cannot be borrowed to make another result causal.
This distinction is not a technical footnote. It determines what a reader may responsibly say. “Services trade is associated with shorter current job duration” is supported by the evidence. “Services trade causes every worker to change jobs more often” is stronger and not supported. “Those moves are voluntary” goes beyond what is observed, although it is a plausible interpretation of the pattern. “Shorter tenure proves growing precarity” ignores the null findings on employment exits, sector switching, and temporary contracts. “Trade improves job quality” is also too broad, because offshoring can improve physical conditions or prospects while worsening the social environment, and destination-specific effects can point in opposite directions.
The useful starting model is therefore modest but powerful: first identify the exposure, then identify the outcome, then trace the route between them. Ask whether the result concerns a current job, an employment transition, a contract, a working condition, a region, or a local spillover. Ask whose experience is being averaged. Finally, ask whether the study reports an association or a causal estimate under explicit assumptions. That sequence does not eliminate uncertainty. It makes uncertainty visible enough for a better judgment.
The next chapter begins with the measurement problem itself: what a shorter job stay can reveal, what it cannot, and which additional observations are needed before calling labour-market dynamism either an opportunity or a form of precarity.