← Papers

Cheaper AI can either hollow out formal jobs or bolster them: if imported AI capital substitutes for formal workers it weakens formal labor demand and swells the informal sector, but if it complements formal labor it increases formal employment, wages, investment and aggregate output.

Cheaper AI, More Informality? A Dual Labor Market Model for Developing Economies

Gabriel Montes-Rojas, Fernando Toledo, Juan Manuel Rodríguez Repeti · July 16, 2026

arxiv theoretical low evidence 7/10 relevance Full text usable extracted full text Source PDF

Structured author observations

Linked only from stored provider relations; the raw author line above is never matched by name.

Semantic Scholar

Latest observation: July 23, 2026

Gabriel Montes‐Rojas provider ID
Fernando Toledo provider ID
Juan Manuel Rodríguez Repeti provider ID

A calibrated open-economy DSGE model shows that falling AI prices can either shrink formal employment and expand informal-sector work when AI substitutes for formal labor, or expand formal employment, output, wages, investment, and capital accumulation when AI complements formal workers.

Citation observations

Cumulative provider counts captured on specific dates; providers are never combined.

0 cumulative citations

Semantic Scholar · Observed July 22, 2026

View corpus context

This paper studies what happens when AI gets cheaper, with emphasis on the labor market outcomes, whether it creates formal jobs or whether it pushes workers into informality. We argue that the answer depends on the elasticity of substitution between imported AI capital and formal labor. We build a small open economy DSGE model with a dual labor market, imported AI capital, and country risk, calibrated to an economy where informality is pervasive. The same decline in AI prices produces sharply different labor-market outcomes depending on whether AI substitutes or complements formal workers. Under substitution, cheaper AI weakens formal labor demand and increases the role of the informal sector as an employment buffer. Under complementarity, it expands formal employment and amplifies output, wages, investment, and capital accumulation. The model therefore shows that AI can become either a source of displacement pressure or a driver of formal-sector expansion, depending on how it interacts with human labor.

Summary

Main Finding

The labor-market effect of cheaper AI in developing economies depends critically on the elasticity of substitution between imported AI capital and formal-sector labor (σKL). If AI is a substitute for formal labor, falling AI prices weaken formal labor demand and push workers into the informal sector (informality rises as an employment buffer). If AI is a complement to formal labor, cheaper AI expands formal employment and amplifies output, wages, investment and capital accumulation. General-equilibrium channels (relative prices, demand spillovers, capital accumulation and the external-debt channel) shape the magnitude of these outcomes.

Key Points

Pivotal parameter: elasticity of substitution between imported AI capital and formal labor (σKL). Complementarity vs substitution produces opposite labor-market outcomes.
Model environment:
- Small open economy DSGE with a dual labor market (formal and informal sectors).
- AI modeled as an imported intermediate capital input used only in the formal sector.
- Informal sector produces non-tradable goods with a labor-only technology and decreasing returns.
- Households have GHH preferences; consumption is a CES composite of tradable (formal) and non-tradable (informal) goods (intra-temporal elasticity η).
- Free labor mobility equalizes wages across sectors at the margin.
- Country risk premium / debt-elastic interest rate (Schmitt-Grohé & Uribe / Neumeyer & Perri style) for external vulnerability.
Calibration and simulations:
- Calibrated to a representative Latin American economy with high informality (~50%).
- Simulate a persistent (AR(1)) international AI price shock (1% drop) and compare substitution vs complementarity regimes.
Mechanisms highlighted:
- Direct factor substitution/augmentation in the formal sector through the σKL parameter.
- Informal sector acts as a buffer absorbing displaced workers when AI substitutes formal labor.
- Demand-side and external-finance channels can amplify or damp responses (e.g., changes in real exchange rate and non-tradable prices).
Empirical and practical context:
- Uses stylized facts and existing estimates of AI exposure in Latin America (ILO/World Bank, Azuara et al., Ciaschi et al.).
- Notes that formal workers are more exposed to AI, while informal workers are less exposed (so informality shields from displacement but also from productivity gains).
Deliverables and transparency:
- Paper provides analytical steady-state conditions linking informal-sector parameters to observables and supplies replicable Dynare code.
Limitations acknowledged:
- Representative-agent setup — cannot speak to distributional impacts across worker groups.
- Simplifying assumptions: no heterogeneous firms/workers, abstract from endogenous AI adoption decisions and some financial frictions.

Data & Methods

Model type: Small open economy DSGE with dual labor markets and nested CES technologies.
- Inner nest: formal-sector production combines physical capital, formal labor and imported AI capital with a CES specification where σKL is the substitution elasticity between AI and formal labor.
- Consumption: CES between tradable (formal) and non-tradable (informal) goods (elasticity η).
- Preferences: GHH (no wealth effect on labor).
Shocks and dynamics:
- International AI price shock: modeled as a persistent AR(1) process (1% drop in simulations).
- Productivity and other shocks follow persistent processes per emerging-market literature.
- Country risk premium is debt-elastic to ensure stationarity of external debt.
Calibration:
- Parameter values drawn from macro and automation literatures and regional statistics (ILO/World Bank, region-level informality data).
- Calibrated to capture Latin America-like informality and exposure patterns.
Solution and experiments:
- Solve and simulate the model (Dynare code provided).
- Compare model outcomes under two regimes: σKL < 1 (complementarity) vs σKL > 1 (substitution) — qualitatively different results.
Empirical inputs referenced:
- AI exposure and occupation/task metrics from ILO/World Bank (2024), Azuara et al. (2024), Ciaschi et al. (2025).
- Informality benchmarks from Gasparini & Tornarolli (2009); firm/informality micro-foundations from Ulyssea (2018) and related literature.

Implications for AI Economics

Policy hinge: estimate σKL. The right policy depends on whether imported AI complements or substitutes formal labor.
- If complementarity: policies that lower barriers to AI adoption in formal firms, invest in complementary human capital, and encourage formalization can magnify growth and formal employment gains.
- If substitution: anticipate displacement pressures in the formal sector; prioritize social safety nets, active labor-market policies, retraining, and measures to support transitions (and tax/base adjustments because informality growth erodes the tax base).
For developing economies:
- Informality acts as an employment buffer but can trap workers outside productivity-enhancing adoption (they’re shielded from AI exposure but also from augmentation).
- Policy should focus on enabling informal firms/workers to access complementary technologies and finance so formalization can capture productivity gains if complementarities exist.
- External-finance vulnerability matters: shifts in capital accumulation and external debt amplify or mitigate AI shocks — macroprudential and external-debt management remain important.
Empirical research priorities:
- Micro and macro estimation of the elasticity of substitution between AI capital and formal labor (σKL) in different industries and occupations.
- Measure heterogeneous exposure and adoption probabilities across formal vs informal firms and workers.
- Track informality flows over time in response to AI adoption events (panel data, matched employer–employee datasets).
- Evaluate distributional consequences by extending models to heterogeneous agents and firms.
Modeling and research extensions suggested:
- Introduce heterogeneous firms and workers, endogenous AI adoption, richer financial frictions, and explicit redistribution to analyze welfare and distributional outcomes.
- Explore policy experiments (taxation, subsidies for AI adoption, training programs, formalization incentives) conditioned on σKL regimes.
Takeaway for researchers and policymakers: cheaper AI is neither unambiguously good nor bad for formal employment in developing economies — the sign and welfare implications hinge on the nature of AI–labor interactions (σKL) and on general-equilibrium feedbacks mediated by prices, demand and external financing.

Assessment

Paper Typetheoretical Evidence Strengthlow — Results are derived from a theoretical calibrated model rather than empirical causal estimation; conclusions depend critically on model structure and parameter choices (notably the elasticity of substitution) and lack micro- or macro-level validation with observed data. Methods Rigormedium — The paper uses a standard, rigorous DSGE framework augmented with a dual labor market, imported AI capital, and country risk and performs calibrated simulations; however, the approach rests on strong functional-form assumptions, selected parameter values, and limited robustness/validation against empirical microdata. SampleA calibrated small open economy with a large informal sector (stylized middle-income economy); uses macro aggregates and literature-based parameter values to simulate dynamics following exogenous declines in imported AI capital prices — no primary microdata or causal identification from observational data. Themeslabor_markets productivity IdentificationComparative-static analysis of a calibrated small-open-economy DSGE model with a dual labor market: causal claims follow from structural model assumptions (production functions, elasticities of substitution between imported AI capital and formal labor, country risk) and simulated responses to exogenous declines in AI prices. GeneralizabilityRelies on assumed elasticity of substitution between AI capital and formal labor; outcomes hinge on this parameter, Calibrated to a single stylized middle-income economy with pervasive informality — may not generalize to high-income or very low-income contexts, Assumes AI enters as imported capital; ignores domestic AI development, firm-level adoption heterogeneity, and adoption costs, Abstracts from important micro frictions (skills mismatch, job search, enforcement of labor regulations, heterogeneous worker/firm responses), No empirical validation with microdata or cross-country evidence; sensitive to functional-form and parameter choices

Claims (10)

Claim	Direction	Outcome	Confidence & Evidence	Details
We build a small open economy DSGE model with a dual labor market, imported AI capital, and country risk, calibrated to an economy where informality is pervasive. Other	null_result	model specification / calibration	Reading fidelity high Study strength high	not reported 0.2
The labor-market outcome of a decline in AI prices depends on the elasticity of substitution between imported AI capital and formal labor. Other	mixed	interaction effect between AI capital and formal labor (elasticity of substitution)	Reading fidelity high Study strength medium	not reported 0.12
When AI is a substitute for formal workers, cheaper AI weakens formal labor demand. Employment	negative	formal labor demand / formal employment	Reading fidelity high Study strength medium	not reported 0.12
Under substitution, cheaper AI increases the role of the informal sector as an employment buffer. Employment	positive	informal sector employment / share of informal employment	Reading fidelity high Study strength medium	not reported 0.12
When AI complements formal workers, cheaper AI expands formal employment. Employment	positive	formal employment	Reading fidelity high Study strength medium	not reported 0.12
Under complementarity, cheaper AI amplifies aggregate output. Fiscal And Macroeconomic	positive	aggregate output	Reading fidelity high Study strength medium	not reported 0.12
Under complementarity, cheaper AI raises wages. Wages	positive	wages	Reading fidelity high Study strength medium	not reported 0.12
Under complementarity, cheaper AI increases investment and capital accumulation. Firm Productivity	positive	investment / capital accumulation	Reading fidelity high Study strength medium	not reported 0.12
The same decline in AI prices produces sharply different labor-market outcomes depending on whether AI substitutes or complements formal workers. Employment	mixed	labor-market outcomes (formal vs informal employment, wages, output)	Reading fidelity high Study strength medium	not reported 0.12
AI can become either a source of displacement pressure or a driver of formal-sector expansion, depending on how it interacts with human labor. Job Displacement	mixed	job displacement vs. formal-sector expansion	Reading fidelity high Study strength medium	not reported 0.12