AI as Coordination-Compressing Capital: Task Reallocation, Organizational Redesign, and the Regime Fork

Task-based models of AI and labor hold organizational structure fixed. We introduce agent capital: AI that reduces coordination costs, expanding spans of control and enabling endogenous task creation. Five propositions characterize how coordination compression affects output, hierarchy, manager demand, wage dispersion, and the task frontier. The model generates a regime fork: the same technology produces broad-based gains or superstar concentration depending on who benefits from coordination compression. Simulations with heterogeneous workers confirm sharp regime divergence. Economy-wide inequality falls in all regimes through employment expansion, but the manager-worker wage gap widens universally. The distributional impact hinges on who controls organizational elasticity.

Summary

Main Finding

Introducing "agent capital" — AI that lowers coordination costs inside firms — changes organizational structure endogenously. Lower coordination costs expand managers’ spans of control and enable new tasks. The model yields five propositions about output, hierarchy, manager demand, wage dispersion, and the task frontier, and produces a regime fork: the same coordination-compressing technology can deliver either broad-based gains or superstar concentration depending on who captures the coordination benefits (who controls organizational elasticity). Simulations with heterogeneous workers confirm sharp divergence across regimes. Across regimes employment expands and overall inequality falls, but the manager–worker wage gap widens in all cases; distributional outcomes hinge on who controls the gains from coordination compression.

Key Points

Agent capital: AI reduces coordination costs (coordination compression), enabling managers to supervise more subordinates and allowing firms to create new tasks endogenously.
Five core consequences (propositions):
Output increases due to reduced coordination friction and task-creation possibilities.
Hierarchies compress (larger spans of control) — fewer layers may be needed for the same organization.
Aggregate demand for managers changes non-trivially (can increase) because managerial roles are amplified by coordination gains.
Wage dispersion is affected: manager–worker wage gaps widen universally even if overall inequality falls.
The task frontier expands (new tasks become profitable as coordination costs fall).
Regime fork: identical technology can produce either broad-based wage and output gains or superstar concentration depending on allocation of coordination-rent (e.g., whether benefits accrue to many workers/firms or to a few top agents).
Simulations with heterogeneous workers reproduce the analytical predictions and show sharp divergence between regimes.
Net effect on overall inequality: employment expansion reduces economy-wide inequality in both regimes, but distributional details differ; control over organizational elasticity determines concentration of gains.

Data & Methods

Theory: Formal task-based model augmented with endogenous organizational structure via "agent capital" that compresses coordination costs. The model holds traditional task productivity but allows coordination costs to vary and affects span of control and task creation.
Analytical results: Five propositions derived characterizing comparative statics of coordination compression on output, hierarchy, manager demand, wage dispersion, and the task frontier.
Numerical methods: Simulations/calibrations with heterogeneous agents (workers and managers) to trace distributional outcomes under alternative allocations of coordination benefits (the two regimes).
Identification / empirical implications: The model suggests observable firm-level and economy-wide moments (changes in spans of control, manager share of payroll, new task creation rates, employment growth, shifts in wage distribution) that can be used to test the theory.

Implications for AI Economics

Organizational channels matter: Beyond task substitution/complementarity, AI that lowers coordination costs fundamentally reshapes firm structure and task creation — policy and empirical work should measure organizational change (spans of control, layers, task portfolios).
Distributional policy hinge: Who controls organizational elasticity (ownership, bargaining power, institutions) crucially determines whether coordination-compressing AI yields broad-based gains or superstar concentration. Policies that influence control over coordination rents (e.g., corporate governance, labor bargaining rights, ownership structures) will shape inequality outcomes.
Managerial labor: Expect rising demand for managers (or higher value per manager) and widening manager–worker wage gaps even when overall inequality declines; labor-market and training policies should anticipate changing returns to managerial skills.
Measurement priorities: Empirically track indicators such as span of control, layers per firm, manager wage premia, new task incidence, employment growth across occupations, and firm-level AI adoption/organizational investments.
Policy levers: To steer outcomes toward broad-based gains, consider interventions that distribute coordination gains (profit-sharing, worker ownership, stronger worker bargaining, antitrust to limit superstar firm dominance), and investments in worker retraining to capture task frontier gains.
Research directions: Empirical tests linking firm-level AI adoption to changes in organizational structure and wage distribution; analysis of contract/institutional arrangements that determine who captures coordination rents; dynamic extensions with entry, firm competition, and capital ownership.

Assessment

Paper Typetheoretical Evidence Strengthn/a — Theoretical derivations produce clear comparative statics and the calibrated simulations demonstrate mechanism-driven heterogeneity across regimes, but there is no empirical testing or causal estimation using real-world data to validate the model's quantitative predictions. Methods Rigorhigh — The work develops a formal model with clear assumptions, derives multiple analytical propositions, and uses calibrated simulations with heterogeneous agents to explore distributional implications and regime behavior, which is methodologically rigorous for theory-driven work. SampleNo empirical sample; a formal task-based model with heterogeneous agents (workers and managers) is solved analytically and explored via numerical simulations/calibrations under alternative allocations of coordination benefits (two regimes). Themesorg_design productivity labor_markets inequality IdentificationAnalytical comparative statics from a formal task-based model that endogenizes organizational structure; causal claims follow from model assumptions and derived propositions, with numerical simulations (calibrations) used to illustrate comparative statics across parameter regimes rather than empirical identification. GeneralizabilityNo empirical validation—predictions rely on model structure and parameter choices rather than measured firm-level adoption data., Calibration parameters and agent heterogeneity may not reflect real-world firm distributions and institutional details., Abstracts from dynamic industry entry/exit, competition, and capital ownership details that could alter outcomes., Focuses on coordination-compression channel; other AI channels (task substitution, task creation outside firms) are stylized or omitted., Assumes clear mapping from AI adoption to coordination costs and from coordination rents to bargaining outcomes, which may vary across contexts and institutions.

Claims (12)

Claim	Direction	Confidence	Outcome	Details
Introducing ‘agent capital’ (AI that lowers coordination costs) reduces coordination costs inside firms (coordination compression). Organizational Efficiency	negative	medium	coordination costs (firm-internal coordination friction parameter)	Agent capital assumed to reduce firm-internal coordination costs (model assumption) 0.01
Lower coordination costs expand managers’ spans of control (managers can supervise more subordinates). Organizational Efficiency	positive	high	span of control (number of subordinates per manager)	Lower coordination costs expand managers' spans of control (model-derived positive effect) 0.02
Hierarchy compresses: fewer organizational layers are needed for a given firm output as coordination costs fall. Organizational Efficiency	negative	high	number of hierarchical layers per firm	Hierarchy compresses (fewer layers) as coordination costs fall (model result) 0.02
Aggregate output increases when coordination costs fall because reduced frictions and endogenous task creation raise productive capacity. Firm Productivity	positive	high	aggregate output (economy-wide production)	Aggregate output increases when coordination costs fall (model result) 0.02
The task frontier expands: new tasks become profitable and are created endogenously as coordination costs decline. Task Allocation	positive	high	task frontier (set/number of profitable tasks)	Task frontier expands (new tasks become profitable) as coordination costs decline 0.02
Aggregate demand for managers can increase non-trivially as coordination improvements amplify managerial roles. Employment	positive	medium	aggregate demand for managers (employment/share of managers)	Aggregate demand for managers can increase (non-monotonic; model shows possible positive manager demand response) 0.01
Manager–worker wage gaps widen universally in the model when coordination costs fall, even when overall inequality declines. Wages	positive	medium	manager–worker wage gap (wage premium of managers over workers)	Manager–worker wage gap widens when coordination costs fall (model result) 0.01
There is a regime fork: the same coordination-compressing technology can yield either broad-based gains (widespread wage/output increases) or superstar concentration (concentration of gains among few agents), depending on who captures the coordination rents (who controls organizational elasticity). Inequality	mixed	medium	distribution of gains (e.g., wage and output concentration measures across agents/firms)	Coordination-compressing tech can produce either broad-based gains or superstar concentration depending on rent capture (bimodal/conditional distributional outcome) 0.01
Across both regimes employment expands and economy-wide inequality falls (net effect), but distributional details differ by regime. Employment	positive	medium	employment (aggregate employment) and overall inequality (economy-wide inequality metric)	Across both regimes employment expands and overall inequality falls (model simulation net effects; distribution differs by regime) 0.01
Distributional outcomes hinge on institutional/allocation factors (ownership, bargaining power) that determine who controls organizational elasticity and thus who captures coordination rents. Inequality	mixed	medium	distributional outcomes (wage and income distribution conditional on allocation of coordination rents)	Distributional outcomes depend on institutions/ownership/bargaining that determine who captures coordination rents (conditional effect) 0.01
Simulations with heterogeneous workers reproduce the analytical predictions and show sharp divergence in outcomes across the two regimes. Organizational Efficiency	mixed	medium	simulation outcomes (span of control, manager demand, wage dispersion, task frontier across regimes)	Simulations with heterogeneous workers reproduce analytical predictions and show divergent outcomes across regimes (model validation) 0.01
Observable firm-level and economy-wide moments—changes in spans of control, manager share of payroll, incidence of new tasks, employment growth, and shifts in the wage distribution—can be used to test the model's predictions. Other	null_result	low	empirical testable moments (spans of control, manager payroll share, new-task incidence, employment growth, wage-distribution shifts)	Observable firm- and economy-level moments can be used to test model predictions (identification strategy suggestion) 0.01

Entities

Agent Capital (ai_tool) Coordination Compression (method) Task-Based Model (method) Organizational Elasticity (method) Regime Fork (method) Output Increase (outcome) Hierarchy Compression (Spans of Control) (outcome) Span of Control (outcome) Manager Demand (outcome) Manager–Worker Wage Gap (outcome) Task Frontier Expansion (outcome) Workers (population) Managers (population) Endogenous Organizational Structure (method) Heterogeneous-Agent Simulations (method) Wage Dispersion (outcome) Task Creation (outcome) Employment Expansion (outcome) Overall Inequality Decline (outcome) Coordination Rents (outcome) Heterogeneous Workers (population) Firms (population) Superstar Agents (population) Span of Control (Measured) (outcome) Layers per Firm (outcome) Manager Share of Payroll (outcome) New Task Incidence (outcome) Firm-Level AI Adoption (outcome) Corporate Governance (institution) Labor Bargaining Rights (institution) Ownership Structures (institution) Antitrust (institution) Profit-Sharing (institution) Worker Ownership (institution)

AI that lowers coordination costs lets managers supervise more workers, boosting output and compressing hierarchies while widening manager–worker pay gaps; whether gains are shared or concentrated into superstars hinges on who captures the coordination rents.