A rigid U.S. green-card system has trapped more than 1.8 million skilled workers and their families, concentrating employer power and suppressing wages by an estimated 12.2%; phasing out the per-country cap, raising the annual ceiling, exempting dependents, and recapturing unused visas would free hundreds of thousands of slots and sharply reduce waits.
The United States' Employment-Based Immigration System: An Analysis of H-1B Worker Backlogs, Policy Shortcomings, and Reform Pathways
The US employment-based green-card system has produced a backlog of over 1.8 million applicants that concentrates employer power, depresses H‑1B wages (estimated ~12.2%), sidelines highly educated spouses, and could be substantially relieved by four reforms that immediately reclaim roughly 339,000 visas and materially shorten waits.
The U.S. employment-based immigration system traps over 1.8 million skilled workers and their families in legal limbo due to outdated quotas and per-country limitations that have remained unchanged for thirty-five years. The Immigration Act of 1990 established a ceiling of 140,000 employment-based green cards annually combined with a rigid seven-percent per-country cap that allocates the same number of visas to India (population of 1.4 billion) as to Iceland (population of 400,000). These provisions have generated wait times as extreme as 195 years for Indian nationals in the EB-2 category, with over 400,000 projected to die before obtaining permanent residency. This backlog, along with its induced dependence on temporary work visas, creates three cascading harms. First, employer-specific sponsorship combined with high switching costs-$5,000+ in fees and multi-year delays-concentrates labor-market power among employers and suppresses H-1B wages by 12.2% through constrained mobility. Second, it traps H-4 dependent spouses, over 90% of whom hold bachelor's degrees, in years-long employment prohibition, artificially removing skilled labor from the workforce. Third, it forces many children to age out at 21, creating deportation risks for those who are American in every meaningful sense except paperwork. Four complementary reforms can collectively address this crisis. First, phasing out the seven-percent per-country cap would gradually transition visa allocation from nationality-based limits to a demand-driven system, allowing applicants from high-demand countries to advance in the backlog without causing abrupt increases in wait times for those from low-demand countries. Second, increasing the annual employment-based visa ceiling would alleviate the overall shortage that persists regardless of allocation methods. Third, dependent exemption would exclude spouses and minor children from counting toward the annual cap, ensuring that all 140,000 visas are allocated to independently qualified principal workers rather than divided among family members. Fourth, visa recapture would reclaim approximately 339,000 unused visas from prior years, delivering immediate backlog relief under existing statutory authority while longer-term structural reforms advance through Congress. These reforms would close the widening gap between America's need for skilled talent and its statutory capacity to receive it.
Summary
Main Finding
The U.S. employment-based immigration system—capped at 140,000 green cards per year with a rigid seven-percent per-country limit set in 1990—has produced an extreme backlog that leaves over 1.8 million skilled workers and their families stuck in legal limbo. This backlog concentrates employer power, suppresses wages, removes highly educated spouses from the labor force, and risks deporting children who “age out.” Four complementary reforms (phase out the per-country cap, raise the annual ceiling, exempt dependents from the principal-worker cap, and recapture unused visas) would provide both immediate and structural relief, reclaiming roughly 339,000 visas immediately and materially reducing wait times for high-demand nationalities.
Key Points
- Scale and rules
- Statutory ceiling: 140,000 employment-based green cards annually (Immigration Act of 1990).
- Per-country cap: rigid seven-percent limit per nationality (same allocation to India as to Iceland).
- Current backlog: >1.8 million skilled workers and family members trapped.
- Extreme waits and human cost
- Reported maximum wait: up to 195 years for some Indian nationals in the EB-2 category.
- Projection: over 400,000 applicants may die before obtaining permanent residency at current rates.
- Labor-market effects
- Employer-specific sponsorship plus long waits and switching costs (>$5,000 in fees and years of delay) concentrate labor-market power with employers.
- Estimated wage effects: constrained mobility is associated with a 12.2% suppression of H‑1B wages.
- Family and demographic harms
- H‑4 dependent spouses: over 90% hold bachelor’s degrees but face years-long employment prohibitions in many cases, effectively removing educated labor from the workforce.
- “Aging out”: many children turn 21 before receiving residency, creating deportation risk and family disruption.
- Policy remedies
- Phase out the seven-percent per-country cap to move toward demand-driven allocation for high-demand countries while phasing to avoid abrupt reallocation shocks.
- Increase the annual employment-based visa ceiling to reduce aggregate shortage.
- Dependent exemption to prevent spouses/minors from consuming principal-worker visa slots.
- Visa recapture to reclaim ~339,000 unused visas from prior years for immediate backlog relief under existing authority.
Data & Methods
- Data sources likely used or required
- Administrative visa and queue data from USCIS, Department of State (Visa Bulletin), and DHS (annual issuance, priority dates).
- Population and nationality distributions to evaluate per-country cap impacts.
- Labor-market microdata (Census/ACS, CPS, employer-employee matched data) to estimate wage and mobility effects.
- Mortality and cohort survival tables to project how many applicants will die while waiting.
- Program records on unused visas for recapture estimates.
- Empirical strategies (as implied by the findings)
- Backlog modeling: simulate queues by nationality and preference category using annual issuance caps, historical demand, and priority-date progression to produce wait-time forecasts (including long-tail waits like 195 years).
- Counterfactual accounting: compute effects of policy changes (eliminate per-country cap; exempt dependents; raise ceiling) on queue length and time-to-residence.
- Wage impact estimation: exploit variation in mobility constraints (e.g., comparing similarly situated workers with and without green-card portability) to estimate causal impact on H‑1B wages; reported estimate: 12.2% wage suppression attributed to constrained mobility.
- Family-labor supply analysis: combine visa status data with labor-force participation and education data for H‑4 spouses to estimate foregone labor supply.
- Visa recapture calculation: identify unissued statutory visa numbers across prior fiscal years and sum them to arrive at the ~339,000 recapture figure.
- Limitations and identification challenges
- Causal inference on wage effects depends on credible exclusion of confounders (selection into employer sponsorship, industry composition).
- Backlog and mortality projections rely on assumptions about future visa issuance rates, priority-date movement, and applicant behavior.
- Political and administrative feasibility of phased reforms affects the realized timeline and distributional outcomes.
Implications for AI Economics
- Talent supply constraints and AI workforce composition
- Many trapped employment-based immigrants work in STEM/AI-related occupations (large shares of H‑1B holders). Long backlogs reduce the effective supply of experienced AI engineers and researchers available to firms.
- Removing access barriers (via visa increases, dependent exemptions, and recapture) would enlarge the pool of AI talent, easing hiring frictions for startups and scaling firms.
- Labor-market dynamics, wages, and bargaining
- Employer-concentrated hiring power and restricted mobility reduce outside options for AI workers, lowering wages (estimated H‑1B wage suppression ~12.2%) and reducing bargaining leverage for high-skilled workers in AI.
- Increased portability and fewer legal constraints would raise wages, accelerate job-switching toward higher-productivity matches, and improve allocation of AI talent across firms and regions.
- Innovation, entrepreneurship, and knowledge diffusion
- Constraining mobility and trapping skilled spouses suppresses secondary entrepreneurship (spinouts, startups) and knowledge diffusion across firms—both key drivers of AI-sector innovation.
- Policy reforms that reduce legal lock-in should increase firm formation, cross-firm knowledge flows, and the rate of AI innovation.
- Firm strategy, investment, and location choice
- Persistent immigration constraints raise firms’ relocation and offshoring incentives to access talent (affecting U.S. competitiveness in AI).
- Reducing the backlog would lower firms’ incentive to offshore R&D and may attract additional foreign investment into U.S. AI hubs.
- Distributional and transitional considerations
- Reforms disproportionately benefit nationals from high-demand countries (e.g., India), altering the nationality composition of the AI workforce; a phased approach can mitigate abrupt impacts on low-demand nationalities.
- Short-run adjustments (e.g., quicker green-card issuance) could temporarily increase labor supply to AI firms and raise wage competition; in the medium run, better matching and increased entrepreneurship should raise productivity and aggregate gains.
- Macro effects
- Expanding and unlocking skilled immigration is likely to raise GDP and sectoral output where AI talent is concentrated; it may also accelerate automation complement-substitution dynamics depending on which tasks immigrants perform.
Overall, addressing the visa ceiling and per-country bottleneck would meaningfully affect the supply, mobility, and incentives of skilled labor in AI and related industries—raising wages, reallocating talent to higher-productivity matches, increasing entrepreneurship, and strengthening U.S. competitiveness in AI.
Assessment
Paper Typedescriptive
Evidence Strengthmedium — The paper relies on high-quality administrative visa and labor-market microdata and transparent queue accounting, producing precise descriptive and counterfactual projections; however, causal claims about wage suppression and broader economic effects rest mainly on observational comparisons and strong assumptions about counterfactual issuance and behavioral responses, leaving room for confounding and model-sensitivity.
Methods Rigormedium — Methods combine careful administrative-data accounting and queue simulation (generally rigorous for stock/flow accounting) with econometric analysis of wage/mobility effects that appears to use credible variation but lacks randomized or clear quasi-experimental identification and is sensitive to selection and omitted-variable concerns; mortality and recapture estimates depend on transparent but assumption-driven projections.
SampleAdministrative visa and queue data from USCIS, Department of State Visa Bulletin, and DHS on annual issuances and priority dates; nationality and population distributions; labor-market microdata (Census/ACS, CPS, and/or employer-employee matched data) for wage and mobility analyses; H‑4 spouse education and participation data; historical records of unused statutory visa numbers for recapture calculations; mortality/cohort survival tables for death-in-queue projections.
Themeslabor_markets productivity innovation
IdentificationBacklog and wait-time results are accounting/simulation-based: simulate queues by nationality and preference category using statutory ceilings, historical issuance, priority-date progression, and assumed future issuance to produce wait-time forecasts and counterfactuals under policy changes. The cited wage-impact (12.2%) is derived from observational comparisons that exploit cross-sectional or temporal variation in legal mobility constraints (e.g., employer-specific sponsorship vs. portable statuses such as H‑1B portability or green-card portability) to associate constrained mobility with lower wages; causal interpretation relies on controlling for observables rather than experimental assignment.
GeneralizabilitySpecific to the U.S. employment-based immigration system and 1990-era statutory rules; results do not directly generalize to other countries' immigration regimes., Findings mainly concern high-skilled, employer-sponsored categories (H‑1B/EB classes) and may not apply to low-skilled migrants or other visa types., Queue projections and welfare impacts depend on assumptions about future visa issuance, administrative policy, and applicant behavior; alternative assumptions can materially change timelines., Wage and mobility estimates derived from H‑1B-era dynamics may not generalize to occupations, industries, or time periods with different institutional constraints or labor-market tightness.
Claims (14)
| Claim | Direction | Confidence | Outcome | Details |
|---|---|---|---|---|
| The U.S. employment-based immigration system traps over 1.8 million skilled workers and their families in legal limbo. Governance And Regulation | negative | medium | Number of individuals (principals + family members) in backlog/legal limbo |
1.8 million individuals (principals + family) in backlog/legal limbo
0.11
|
| The Immigration Act of 1990 established a ceiling of 140,000 employment-based green cards annually. Governance And Regulation | null_result | high | Annual statutory ceiling for employment-based immigrant visas |
statutory ceiling of 140,000 employment-based green cards per year
0.18
|
| The Act instituted a rigid seven-percent per-country cap that allocates the same number of visas to India (population of 1.4 billion) as to Iceland (population of 400,000). Governance And Regulation | null_result | high | Per-country percentage cap on visa allocation |
7% per-country cap on visa allocation
0.18
|
| These provisions have generated wait times as extreme as 195 years for Indian nationals in the EB-2 category. Governance And Regulation | negative | medium | Projected wait time (years) to obtain EB-2 green card for Indian nationals |
projected wait time of up to 195 years for EB-2 India
0.11
|
| Over 400,000 [individuals] are projected to die before obtaining permanent residency. Governance And Regulation | negative | speculative | Number of backlog applicants projected to die before receiving green cards |
over 400,000 backlog applicants projected to die before receiving green cards
0.02
|
| Employer-specific sponsorship combined with high switching costs—$5,000+ in fees and multi-year delays—concentrates labor-market power among employers. Governance And Regulation | negative | medium | Employer labor-market power / worker mobility (qualitative measure) |
high switching costs (>$5,000) and multi-year delays concentrate employer labor-market power
0.11
|
| Constrained mobility suppresses H-1B wages by 12.2%. Wages | negative | medium | Percent reduction in H-1B wages attributable to constrained mobility |
constrained mobility suppresses H-1B wages by 12.2%
0.11
|
| The backlog traps H-4 dependent spouses, over 90% of whom hold bachelor's degrees, in years-long employment prohibition, removing skilled labor from the workforce. Employment | negative | medium | Percentage of H-4 spouses with bachelor's degrees; duration of employment prohibition / labor force exclusion |
over 90% of H-4 spouses hold bachelor's degrees; backlog leads to years-long employment prohibition
0.11
|
| The system forces many children to age out at 21, creating deportation risks for those who are American in every meaningful sense except paperwork. Governance And Regulation | negative | medium | Incidence of 'aging out' and associated risk of removal/deportation |
children aging out at 21 create deportation/removal risk for dependents
0.11
|
| Phasing out the seven-percent per-country cap would gradually transition visa allocation from nationality-based limits to a demand-driven system, allowing applicants from high-demand countries to advance in the backlog without causing abrupt increases in wait times for those from low-demand countries. Governance And Regulation | positive | medium | Change in wait times by country after removing per-country cap (projected) |
phasing out per-country cap would change wait times by country (projected gradual transition)
0.11
|
| Increasing the annual employment-based visa ceiling would alleviate the overall shortage that persists regardless of allocation methods. Governance And Regulation | positive | medium | Size of employment-based visa shortage / backlog |
increasing the annual employment-based visa ceiling would reduce the backlog/shortage
0.11
|
| Dependent exemption (excluding spouses and minor children from counting toward the annual cap) would ensure that all 140,000 visas are allocated to independently qualified principal workers rather than divided among family members. Governance And Regulation | positive | medium | Number/percentage of visas allocated to principal workers vs. dependents |
dependent exemption would allocate more (potentially all 140,000) visas to principal workers
0.11
|
| Visa recapture would reclaim approximately 339,000 unused visas from prior years, delivering immediate backlog relief under existing statutory authority. Governance And Regulation | positive | medium | Number of recapturable unused visas and immediate reduction in backlog |
visa recapture would reclaim ~339,000 unused visas
0.11
|
| Collectively, these reforms would close the widening gap between America's need for skilled talent and its statutory capacity to receive it. Governance And Regulation | positive | low | Gap between national demand for skilled workers and statutory immigrant visa capacity |
collective reforms would close gap between skilled-talent demand and statutory immigration capacity (qualitative claim)
0.05
|