Yonsei University
Chae-Yeon Xon
Yonsei University, School of Economics
Home/Resources/Structural Models and Economic Computation

Structural Models and Economic Computation

This book carries economic primitives through choice, equilibrium, measurement, and welfare calculations. It emphasizes executable model contracts: every counterfactual uses the same utility, state, transition, choice set, normalization, and shock law unless a stated alternative model changes one of those objects.

17 chapters85 focused topics

Demand, Supply, Bundles, and Welfare

IV-01

Random Utility, Choice Sets, and Normalization

A discrete-choice model begins with consumers, alternatives, utilities, a shock law, and a choice set. Normalizations remove utility components that choices cannot reveal. Once these objects are declared, probabilities, simulations, derivatives, and counterfactuals can be generated from the same executable model rather than from disconnected formulas.

  1. Consumers and alternatives
  2. Random utility
  3. Logit probabilities
  4. Normalization
  5. Executable model objects
→
IV-02

Residual-Income Demand and Approximation Domains

When utility depends on income remaining after purchase, price enters through a nonlinear residual-income term. Replacing that term by a linear price coefficient is a local approximation whose error depends on price-income ratios, affordability, and heterogeneity. The approximation domain must therefore travel with any derivative, share, or welfare calculation.

  1. Exact residual-income utility
  2. Linear price branches
  3. Taylor error
  4. Affordability
  5. Income heterogeneity
→
IV-03

Market Shares, Inversion, and Demographic Micro Moments

Market shares integrate individual choices over heterogeneity, while demographic micro moments condition on purchase events. Share inversion recovers mean utilities for a fixed model and parameter vector. Matching aggregate shares and a limited set of purchaser moments does not reconstruct the joint distribution of utilities, incomes, and choices.

  1. Individual-to-market aggregation
  2. Share matching
  3. Purchaser moments
  4. Estimation objectives
  5. Reconstruction boundaries
→
IV-04

From Demand Derivatives to Markups and Prices

Supply-side counterfactuals inherit the entire demand derivative. Multi-product ownership maps the share Jacobian into markups through firms' first-order conditions. Two demand systems that match baseline shares can therefore imply different elasticities, markups, merger effects, and equilibrium prices.

  1. Individual demand derivatives
  2. Aggregate share Jacobians
  3. Ownership matrices
  4. Markup equations
  5. Price counterfactuals
→
IV-05

Consumer Welfare and Compensating Variation

Consumer welfare must be computed from the same utility model that generated demand. A log-sum formula has a closed monetary interpretation under quasilinear utility with a constant marginal utility of income. Nonlinear residual-income utility generally requires a consumer-specific compensating-variation equation and an explicit treatment of shocks and choice sets.

  1. The valuation object
  2. Log-sum welfare
  3. Nonlinear income effects
  4. Ex ante and ex post welfare
  5. Coherent counterfactuals
→
IV-06

Viewing-Time Optimization and Set-Valued Demand

A bundle's value can arise from an internal allocation problem rather than a sum of item values. With concave viewing benefits and a time constraint, optimal attention depends jointly on all available channels. Reference-bundle increments approximate this set value only under restrictions that make marginal contributions portable across sets.

  1. Time allocation
  2. KKT systems
  3. Proportional allocation rules
  4. Reference bundles
  5. Purchase and choice sets
→
IV-07

Bargaining, Disagreement States, and Choice-Law Consistency

A bargaining calculation requires a feasible agreement set, disagreement payoffs, bargaining weights, and a rule linking fees to consumer choices and firm profits. Target-specific substitutions in the choice kernel can change both surplus and disagreement values. Coherent counterfactuals therefore propagate one declared choice law through demand, bargaining, and welfare.

  1. The bargaining object
  2. Nash bargaining
  3. Choice kernels
  4. Target-specific computation
  5. Welfare consistency
→

Dynamic Choice, Aggregation, and Policy Capacity

IV-08

Dynamic Choice and Forward Simulation

A dynamic choice model links current rewards, state transitions, expectations, and policy rules. Solving the Bellman equation identifies a policy for the declared state; forward simulation then maps that policy into employment, participation, or other paths. Compressing the state to counts is valid only when transition and reward objects close on those counts.

  1. States, transitions, and rewards
  2. Bellman equations
  3. Policies and deviations
  4. State compression
  5. A labor-market example
→
IV-09

Corporate Control and Policy-Invariant Aggregation

Corporate groups can coordinate financing, pricing, or investment across legal entities. Aggregates built from entity-level balance sheets may conceal control links that govern joint adjustment to policy. A policy-invariant state must preserve the directions through which the relevant corporate organization changes equilibrium responses.

  1. Economic organization
  2. Joint adjustment
  3. Comparative statics
  4. Identification limits
  5. Empirical operationalization
→
IV-10

Heterogeneous States, Transition Economies, and Additive Summaries

A distribution of households, firms, or cases is often the true dynamic state. Moments or counts provide useful summaries when their transitions close under the policy and equilibrium law. Fixed-price transitions and equilibrium feedback must be separated, because a statistic sufficient for logistics can fail once prices or incentives respond to the distribution.

  1. Distribution states
  2. Fixed prices and equilibrium
  3. Resolution
  4. Moment closure
  5. Transition and logistics examples
→
IV-11

Stable Baselines and Policy Response Capacity

A stable baseline describes what happens under the prevailing transition law. Policy capacity describes the set of response paths reachable through available interventions. These objects depend on different matrices and constraints, so a quiet system can contain substantial unused response directions while a volatile system can offer few controllable directions.

  1. Response matrices
  2. Transition accounting
  3. Action menus
  4. Mechanism separation
  5. Value of instruments
→

Policy and Empirical Computation

IV-12

Complementary Inputs, Option Production, and Supply Incentives

Households may need several inputs jointly to make an option feasible: time, care, transport, scheduling, or service availability. The resulting opportunity set is a production object distinct from preferences over feasible options. Bundle experiments and provider incentives can then be analyzed through enabling sets, set functions, and cost certificates.

  1. Options and preferences
  2. Alternative enabling paths
  3. Time and care constraints
  4. Package experiments
  5. Supply incentives
→
IV-13

Policy Transmission and the Actual Unit of Randomization

A policy announcement travels through administrative stages before a person receives a treatment. Assignment, notification, eligibility review, enrollment, and service delivery define different interventions and populations. Estimands must be indexed by the actual randomized unit and the observed transmission path.

  1. Intervention units
  2. Administrative dynamics
  3. Handoff chains
  4. Sample boundaries
  5. Interpretation of effects
→
IV-14

Monetary Policy Experiments and Structural Model Families

Comparing monetary models requires a common shock unit, horizon, observable target, conditioning information, and loss. Structural families can then enter the comparison by their implied response maps. Calibration, profiling, and estimation are separate routes to model-specific inputs and should be labeled accordingly.

  1. A common experiment protocol
  2. Four structural families
  3. Response objects
  4. Calibration and profiling
  5. Counterfactual comparison
→
IV-15

Price Baskets, Bond Cash Flows, and Measurement Certificates

A price or value is defined by its source fields, units, aggregation rule, valuation date, and environment. Price baskets transport item-level records into an index; bond prices transport cash flows through discount factors. Certificates preserve enough of the source to recompute the target under declared environment changes.

  1. Basket measurement
  2. Source fields and revaluation
  3. Bond cash-flow value
  4. Sensitivity records
  5. Information budgets
→
IV-16

Gas Records, Event Time, and Storage Decisions

Recorded event time can differ from physical delivery, market response, and the decision deadline. When timing is uncertain, the value of a clock depends on the storage or scheduling action it informs. Measurement quality is therefore target-specific and must be evaluated against the loss of the actual decision.

  1. Physical and recorded events
  2. Observed responses
  3. Comparison designs
  4. Decision loss
  5. Timing uncertainty
→
IV-17

Bank Records, Reporting Systems, and Chronological Validation

Administrative bank records are produced by definitions, reporting incentives, revisions, and acquisition choices. A learning system combines those records with a prior and a likelihood, then evaluates forecasts in chronological order. Discrete, rounded, censored, or point-mass variables require likelihoods that respect their support.

  1. Field meaning
  2. Posterior construction
  3. Record acquisition
  4. Prediction baselines
  5. Calibration and validation
→