The Displacement Effects of Domestic Outsourcing -- by Mayara Felix, Michael B. Wong

TEXT START: Evidence that domestic outsourcing lowers pay comes largely from on-site transfers, in which workers move to a contractor but keep the same jobs.

The Dissection

The paper dismantles the sanitized version of outsourcing in which workers supposedly retain employment while merely changing employers. Its evidence shows a harsher mechanism: firms preserve the function, discard the workers, and refill the role through occupational replacement. Displaced guards eventually regain employment, but often in different occupations and at wages roughly 12% lower. The labor market does not restore their former position; it processes them into cheaper human material.

The important finding is the loss of firm-specific wage premia. Outsourcing does not need to produce permanent unemployment to destroy worker security. It can preserve aggregate employment while stripping individuals of accumulated bargaining power, earnings, and occupational identity.

The Core Fallacy

The paper’s limiting assumption is that reemployment within five years constitutes meaningful recovery. It measures whether workers return to employment, but the relevant structural question is whether they recover productive leverage, income trajectory, and control over the means of production. They do not.

Under Discontinuity Thesis mechanics, this is a lag-defense story, not a survival story. Human labor remains absorbable here because the economy still requires guards. The paper therefore documents transitional displacement under a labor-demand regime that remains fundamentally human-dependent. It does not refute AI-driven collapse; it shows how much damage can occur before outright labor irrelevance arrives.

Hidden Assumptions

• That occupational switching remains a viable escape route when automation attacks entire categories of cognitive and service work simultaneously.
• That a five-year recovery window is socially and financially survivable.
• That aggregate employment can remain a useful proxy for individual viability.
• That firms will continue needing human substitutes rather than replacing the function itself.
• That historical Brazilian outsourcing reforms scale cleanly to AI, despite AI’s broader scope, speed, and replicability.
• That wage scarring is an adjustment cost rather than evidence of permanent class demotion.
• That institutions can manage sequential displacement when P1, P2, and P3 produce synchronized displacement across sectors.

Social Function

Classification: partial truth and transition management.

The paper is empirically valuable because it exposes displacement that headline employment figures conceal. But its frame remains institutional: workers are tracked through reemployment, occupation, and earnings. That makes the damage legible without confronting the terminal implication. The worker is treated as successfully processed once he is economically reabsorbed, even if he returns poorer and less protected.

The Verdict

This is a clean empirical preview of the death sequence: employment survives statistically while worker viability decays materially. Outsourcing first severs the worker from the premium, then from the occupation, and eventually from the bargaining position. AI extends the same mechanism from selected occupations to the cognitive economy itself. The paper proves neither P1 nor P2, but it supplies evidence for the transition pattern: “recovery” can mean continued labor-market circulation after productive status has already been amputated. The post-WWII employment bargain does not need mass unemployment to fail. It only needs labor to become replaceable and reemployment to become a lower-grade holding pattern.