# Microfoundations Matter for Keynesian Unemployment

Lecture 7 of Advanced Macroeconomics,

26th September 2018. by Dr Asad Zaman VC PIDE

WEBSITE: https://sites.google.com/site/az4macro

To understand Keynes, we must FIRST Understand Say’s Law: “the aggregate demand price of output as a whole is equal to its aggregate supply price for all volumes of output”

Production of goods has a COST = Aggregate SUPPLY price = Payment to factors of production + Profits.

Total Money generated as factor income is ALL the money that is available as demand. ALL goods are for sale. Money will be exactly enough to buy all the goods (because prices will adjust to clear the market).

Consider Thought Experiment in Simple Model

Two factors of production: 40 Acres land and 40 laborers

Land Rent is 40xR, Payment to labor is 40×100=4000

Factor Income = 4000 + 40R, Total Product = 400 Corn.

Price of Corn = (4000+40R)/400=10+R/10

Demand = Supply; Price equilibriates.

This is Say’s Law – Factor Income is exactly the amount needed to purchase the entire amount produced – price creates equality

Crucial Question: What determines Rent & Wages

Share of Capital and Share of Labor?

Classical Answer: Rent is Marginal Product of Land, Wage is Marginal Product of Labor.

The two sum to total income from production under zero profit long run equilibrium.

Keynes agrees with classical answer: W=MPL; this determines capital share as well.

In order to compare with our own answer/model, we first REVIEW the neoclassical theory

According to Neoclassical Theory, Shares are Determined by Marginal Product – Why?

Assume 40 Mini-Landlords, each has one acre of Land (K).

è K=1, L=1 produces 10 Units of Corn.

Suppose W=100 Rp, this is the numeraire. One price can be set arbitrarily, all others will adjust proportionately.

SOLVE for competitive equilibrium in this economy.

Fix K=1, every landlord has K=1 è

◉ First Order Conditions for Profit Maximization

For equations — see slides, or video lecture, linked above

Firms are price takers, p,w fixed in above equation, which gives the demand for labor by firms. At competitive price p, demand MUST come out to 1 unit of labor, otherwise excess or deficit demand for labor. Set L=1 to solve for competitive price. Fix value of b arbitrarily, say b=0.8. Solving for p in above we get:

For equations — see slides, or video lecture, linked above

◉ Competitive Solution

Each LL hired 1 unit of L at wage 100 Rp, produces 10 units of corn worth 12.5/unit, sells for 125 units, makes profit of 25 Rp, enough to buy 2 units of corn. Laborers buy 8 units of corn with 100 Rp. Factor Payments exactly equal demand. Say’s Law holds.

For equations — see slides, or video lecture, linked above

◉ Disequilibrium Dynamics: What Happens if p=15?

Revenue from Sales is 150=15×10, 100 goes to Labor, 50 to LL.

Laborer buys 6.66, LL buys 3.33, Demand=Supply for Corn.

BUT MPL > w.

Competition for additional laborers will drive up the wage, until the shares of capital and labor are 20-80

◉ Cobb-Douglass First Order Conditions

For equations — see slides, or video lecture, linked above

Competitive Equilibrium exists ONLY with Constant Returns to Scale. If , then total revenue from sales will not equal payments to factors.

**Other Production Functions Lead to Different Results**

Key Point of this Lecture: Micro-Structure Matters

Little, Apparently Un-important Details, Make Huge Difference

Cobb-Douglass, CRTS, lead to W=MPL, R=MPK.

If α+β>1 IRTS (Increasing Returns to Scale), then MPL+MPK is less than Total Revenue.

If α+β<1 DRTS, than revenue is less than sum of shares to capital and labor. Both cannot be paid.

Competitive Equilibrium fails in either case.

As Usual, Economists adopt Ostrich Strategy: Assume it all works out — SO LET US ASSUME CRTS!

This assumption reflects an Ideological commitment to Free Markets & Competitive Equilibrium, we freely make assumptions, as required, to get optimality of free market.

In Leontief Production: Rents,Wages not equal to MPL

In constant proportions production function, both factors have FULL output as marginal product. When K=1, L=0, adding L=1 gives 10 units of output. When L=1,K=0, adding K=1 also gives 10 units of output. So BOTH factors have FULL output, factor payments CANNOT be made according the marginal products.

**Central Question**: What Determines Wage/Rent Shares?

Neoclassical Answer is Ideological Propaganda: Conflicting Technical Details are suppressed.

Impression: S&D holds everywhere: Reality: ONLY in Comp Eq

Impression: W=MPL, R=MPK; Reality: ONLY in CE with CRTS

What if NO CE or NO CRTS?

If we reject neoclassical views, THEN what determines share of capital and labor?

Marxist View: Labor Gets Subsistence Wage

Note that FULL output is Marginal Product of Labor = Labor Theory of Value (of course Full output is ALSO MPK).

To understand Marx, Start at arbitrary 50-50 share; determined by historical context and social norms.

Capitalism is competition among capitalists for profits.

Profits come from exploitation of labor. This means capitalists will use power to increase their share.

Exploitation increases until laborers revolt.

So: Is Wage Equal to MPL? Data shows clear divergence

Data strongly conflicts with the idea. To explain these differences, Economists List Six Theories

- Cheaper Capital lead to rise in investment in capital
- Increased Capital Mobility – substitution of cheap labor
- Resultant Reduced Bargaining Power. Fewer Unions.
- “Superstar Effect”: Externality gives massive returns to top
- Rents: Excess profits from patents, non-market effects
- Tax Policy

Above reasons are rationalizations – attempts to save neoclassical theory.

**Real Reason: Financial De-Regulation in 1970s**

Rise of FIRE: Finance, Insurance, Real Estate

Massive increase in payments of rents to FIRE sector (Hudson)

All productivity Gains are captured by the powerful – Marxist hypothesis: share is NOT according to marginal product, but according to relative POWER of the two classes.

Reagan-Thatcher revolution unchained finance, and created power necessary to capture gains in productivity.

Economics & Theories – driven by power struggles between classes.

**Back to MAIN question: **

How to Understand Unemployment after Great Depression?

Answer: use models which capture essential elements –

Models are necessary, because world is complex.

All intuitions cannot be put together in any simple way.

Models ensure coherence and consistency of entire structure.

Experience cannot be understood without models.

Models cannot be understood without experience.

◉ How to Understand Unemployment?

◉ What caused Unemployment after Great Depression?

Say’s Law – if producers produce, process of production will generate enough income for factors to enable purchase of product. No danger of insufficient demand.

OBSERVATION: Producers had CAPACITY to produce – laborers were there, factories were there. BUT they did not produce – WHY? If Say’s Law is VALID, then there was no danger from producing an excess. When we supply, the process will create its own demand.

This is why Keynes STARTS by rejecting Say’s Law.

Keynes’ Intuition: Producers thought they would be unable to sell

That is, Say’s Law does not hold. BUT WHY?

Intuition: Producers think there will be insufficient aggregate demand. (and they are right!)

Intuition: What happens at the aggregate level is DIFFERENT from a scaled up version of what happens at MICRO level.

IDEA: Work solely at Aggregate Level – Unique Innovations of Keynes – invented MACROECONOMICS.

Translate Intuition into Model

Keynes painted broad details – Intuition about Complexity: Behaviors of Aggregates not scaled up Micro Behavior.

Omitted the Micro Picture, concrete details of what happens at the micro level which leads to this Macro.

This led to the critique: Keynesian Economics LACKS Micro-Foundations.

Responses to Micro-Foundations Critique of Keynes

One Response: Macro Picture does not depend on Micro – Chemical properties of molecules have no relation to properties of sub-atomic particles.

Another Response: Create Micro Picture – New Keynesian/ Neo-Keynesian

Crucial Issue: Micro Details MATTER – Is production function Cobb-Douglass CRTS, or Substitutable with IRTS, or Leontief? This MATTERS for theory of effective demand.

Important Point: Neoclassical Micro-foundations lead to REJECTION of Keynes (contrary to Keynes own ideas). This is why Post-Keynesians often resist demand for micro-foundations.

Massive Confusion About Effective Demand: Different Ways of Filling out Micro Details lead to Different Theories

How should we fill out the micro detail?

Try for REALISM – closest approximation to hidden reality.

Approximation is good (enough) if it reproduces observed phenomena. [This is ONLY Criteria: Friedman’s Folly – match between model results and observation is important, but NOT the sole criterion for validity of model]

Critical Realism: Trying to find the HIDDEN structures of reality. In addition to matching empirical evidence, model should provide a PLAUSIBLE explanation of the hidden causal structure which leads to the observed phenomena.

Questions we need to answer – in terms of match between HIDDEN structur

◉ What is the Production Function?

◉ Do We Have Perfect Competition? S&D Theory Holds?

◉ Cobb-Douglass with CRTS required for S&D

◉ Sraffa’s Critique: IRTS & DRTS are not compatible with S&D

◉ CRTS leave SCALE indeterminate.

◉ C-D with CRTS implies MPL is below subsistence level.

So – to find out what model is suitable, we must ask: How did the real wage behave in GD?

Keynes speculates that it increased.

◉ Bordo, Michael D., Christopher J. Erceg, and Charles L. Evans. “Money, sticky wages, and the Great Depression.” American Economic Review 90.5 (2000): 1447-1463.

GRAPH (see slides or video lecture linked above) show steady decline in prices and wages during the great depression, while the real wage remains roughly constant.

Cobb-Douglass: As more labor is hired MPL should decrease.

◉ Links to Relevant Materials

Website for course: Advanced Macroeconomics

Related Post Subtleties of Effective Demand

Information about Author: About Me section of asadzaman.net