Economic Growth, Production Functions and Convergence

Learning objectives

By the end you can:

1. Write the neoclassical production function Y = A · K^α · L^(1−α), state what constant returns to scale and diminishing marginal returns each require of it, and estimate α from published factor shares rather than assuming it.
2. Decompose measured output growth into capital, labour and residual contributions, and verify the result by the independent per-worker route.
3. Explain why total factor productivity is a residual rather than a measurement, and quantify how much of a published TFP figure is driven by the capital-share convention alone.
4. Distinguish capital deepening from technological progress, and predict which of two economies with identical output growth will decelerate first.
5. Solve a Solow steady state for capital, output and consumption per effective worker; derive and apply the golden-rule saving rate; and state precisely what a permanently higher saving rate does and does not do to the growth rate.
6. Distinguish absolute, conditional and club convergence, run the cross-sectional regression that separates them, convert a regression slope into an annual convergence speed and a half-life, and read the middle-income slowdown as a conditional-convergence statement.
7. State the endogenous-growth correction: why non-rival ideas break the diminishing-returns argument, why the AK limiting case removes the steady state altogether, and what the semi-endogenous critique does to the policy conclusion.
8. Quantify how human capital, institutions, demographics and resource dependence enter growth, and locate each correctly inside or outside the production function's A term.
9. Decompose an equity index's total return into dividend yield, real earnings growth, inflation, share-count change and repricing, and explain from it why cross-country GDP growth and equity returns correlate weakly.
10. Write a defensible ten-year country growth assumption with a stated band and three falsifiers, and reconcile it against the terminal growth rate in your own most recent valuation.
11. (Productivity objective: R10 duality.) Build the decomposition, the perpetual-inventory capital series and the convergence regression in code, while running the verification that makes the output admissible: the per-worker identity closes, the capital stock reproduces its own investment series, and the regression's slope reproduces the implied convergence speed.

The duality, stated once (R10). Objectives 1 to 10 are what the gate rewards, because a growth-accounting table is exactly the kind of small, clean, confident-looking output a language model produces fluently and wrongly. A residual of 1.48 and a residual of 0.82 look equally reasonable on a page, and only the capital share behind them says which is right. Objective 11 is the payoff you keep. A tool can write the loop; it cannot tell you that the residual you just published is an artefact of how the National Accounts treat the income of the self-employed.

Prerequisites & connections

Builds on. M7.01 gave you the economy as a machine of transactions, the credit cycle, and the Solow intuition in words: diminishing returns to capital, catch-up growth, and technology as the only durable source of rising living standards. That intuition gets solved here rather than described. M7.07 gave you the India dashboard and the indicators that read a regime; the growth trend under those indicators is what gets built here. EC1.01 gave you the firm's production and cost curves, and the aggregate production function is the same object one level up, with the same diminishing returns doing the same work. QM1.02 gave you regression, the standard error on a slope and the discipline of reading a t statistic before a point estimate, all of which the convergence test needs. M3.02 gave you the equity risk premium three ways, which the index-return decomposition here reconciles against.

Deliberate non-overlaps (R1). M7.01 owns the cycle machine, money creation, the two banking models and the long-term debt cycle; nothing here re-derives them, and where growth and the cycle interact the treatment stays on the trend side. M7.07 owns India's macro dashboard, the FIT framework, the reserve-adequacy tests and the twin-deficit arithmetic; this node uses India's growth data and leaves the policy dashboard alone. EC1.01 owns firm-level production, cost curves and market structure. M3.02 owns the equity risk premium, historical and implied, and the country-risk stack. M3.05 owns the terminal-value cap rule and the reverse DCF; the macro justification for that cap is supplied here and the mechanics are not repeated. EC1.04 owns exchange rates and trade policy, so the growth effects of openness are named here and priced there.