ECON0028

Economics of Growth
LECTURE 3: HUMAN CAPITAL;
GROWTH ACCOUNTING

Luk sz R chel, October 2024

1
a
a
 Overview
• The Solow model considered how the accumulation of physical capital a ects
the level and the growth rate of the standard of living.
• Our treatment of the labor force was… crude.
• Now: think about the characteristics of the labor force: human capital

Y = F(K, HL)
• Also, brie ly: how much do di erent factors (K, H, L) and productivity A
contribute to growth over time and income di erences across countries?
• This is called growth accounting.
f
ff
ff
ff
 What is human capital?
• Qualities of people that are productive (i.e. useful in producing more output)
• Health
• Education
• Job experience
• …
• Characteristics of human capital
• Productive
• Produced through investment
• Earns a return to the worker if he/she works
• Depreciates over time
• Note that these characteristics are similar to what we discussed last time in the context
of K — hence human capital.
Human capital is increasingly important

Google Ngram: Capital

Das Kapital by Karl Marx

Google Ngram: Human Capital

Human Capital by Gary Becker

Health as Human Capital
6
7
 Health as human capital
• Health and income
• Health has intrinsic value in itself
• Health also increases productivity → health is a form of human capital
• Clearly, a 2-way interaction between health and income
• This will give rise to multiplier e ects
ff
Two way interaction between health and income
• Healthier workers function more e iciently and enjoy higher income
• Richer people can a ord better inputs into health (vaccines, clean water, doctors, etc.)
• 2.2 doctors/1000 people in OECD countries; 0.3 in sub-Saharan Africa
ff
ff
 How do health and income interact?

• Equilibrium. Unique?
• Consider an increase in income
unrelated to health, such as an
improvement in production
technology, how would this a ect the
equilibrium health and income?
ff
Effects of an exogenous shift in income
 Effects of an exogenous shift in income

• Multiplier e ects — in case not obvious:

• Workers of any given health level can now produce more output
• The initial increase in output will in turn improve health
• The improved health will in turn increase output
• …
• The initial income shift is ampli ied by the positive feedback between health and
income
ff
f
 Effects of an exogenous shift in health
• Improvements in health signi icant in the
20th century

• We’ll learn how to model innovation later on

in the course!
f
Education as Human
Capital
 Education as human capital

• In many (most?) economies, cognitive ability is far more important than physical
ability in determining a person’s wage

• Investment in education a key form of investment in human capital

• It entails costs…
• … and earns a return.
 Educational attainment of US population

• Educational attainment increasing

• Limits to how far this can go?

17
 Cost of educational investment
• In 2010, total educational spending in the US 6.2% of GDP – these are direct costs
(e.g. tuition), both public (about 3/4) and private (about 1/4).

• Opportunity cost (foregone income while in school) estimated (how?) as almost as

large as direct costs

• Total cost 12.4% of US GDP in 2010

• for comparison, investment in physical capital 19% of GDP — same order of
magnitude
 Measuring the returns to education
• To measure productivity of physical capital, simply look at the rent it commands in the
market (the identity of its owner doesn’t matter)
• Measuring productivity of human capital is not as obvious
• A worker’s wage consists of two parts:
• One part due to raw labor (the person without human capital)
• The other part due to his/her human capital
• This is a simpli ication. Why?
• The human capital is inalienable from its owner
• Consider the same person with and without extra years of schooling (“counterfactual”)
• Wage increase due to the extra years of schooling = return to human capital
• 1000s of papers in economics measure returns to education / schooling
• Challenge?
f
Effects of education on wages
 Returns to schooling

• 1-4 yrs: 13.4% per year; 5-8 yrs: 10.1% per year; beyond 8 yrs: 6.8% per year

• Worker A – zero years of schooling; Worker B – ive years of schooling

• How does worker B’s wage compare with worker A’s?

4
(1 + 0.134) × (1 + 0.101) = 1.82 times as much

f
 Human capital’s share of wages
• We learned that physical capital’s share of national income is 1/3. Labor’s share of
national income is 2/3.

• Decompose the payment to labor into two parts

• Payment to the human capital that workers possess
• Payment to raw labor (what workers would have earned if they did not possess
any human capital)
 Human capital’s share of wages: example
• 1-4 yrs: 13.4% per year; 5-8 yrs: 10.1% per year; beyond 8 yrs: 6.8% per year

• Worker A – zero years of schooling earns $1

• Worker B – ive years of schooling earns $1.82
• Worker B’s wage
= $0.82 (payment to human capital) + $1 (payment to raw labor)
• 0.82/1.82 = 45% attributed to human capital

• Can conduct similar analysis for the country level

f
Breakdown of the population by schooling and wages
Share of human capital in wages in developing countries
Share of human capital in wages in developed countries
 Share of human capital in GDP

• Share of human capital in wages:

• 58% in developing countries; 68% in developed countries
• Recall share of wages in GDP: 2/3
• Share of human capital in GDP (recall capital share about a third)
• 39% (58% x 2/3) in developing countries; 45% (68% x 2/3) in developed
countries

• Note: these are 2010 igures. Up to date numbers would be higher.

f
 An aside: 2 slides on
within-US developments

28
 College wage premium
• Supply of highly educated workers: UP
• Price of highly educated workers: UP
• What’s the story?
• Skill biased technical change

29
Perhaps more than SBTC? Absolute
deterioration for the less educated
on some measures:
Growth / Development
Accounting

31
 Development accounting
• How much of the variation in income per capita across countries is ex-
plained by productivity di erences and di erences in factors of production?
• How much does productivity di er among countries?
• Break down di erences in income levels into two parts:
α 1−α
Y = AK (hL) ⇒
α 1−α
y = A(k h )
output = productivity × (factors of production)
ff
ff
ff
ff
 Development accounting
• α 1−α
Country 1: y1 = A1 × (k1 h1 )

• α 1−α
Country 2: y2 = A2 × (k2 h2 )
• Ratio of output per worker in Country 1 relative to Country 2:
α 1−α

( k2αh21−α )
y1 A1 k1 h1
= ×
y2 A2
• Ratio of output = ratio of productivity × ratio of factors of productions
 Development accounting
• Previous slide implies:
• ratio of productivity = ratio of output / ratio of factors of productions

( y2 )
y1
A1
=

( k2αh21−α )
A2 k1αh11−α

observables: y, k, h; unobservable: A
 Development accounting: example

• Can you get the ratio of productivity in Country 1 to productivity in Country 2?

Use α = 1/3

( y2 )
y1
24
A1 1 24
= = = =2

( 2 2 )
A2 k1αh11−α 271/382/3 3×4
1
k αh 1−α
Development accounting
 Growth accounting
• Growth accounting is similar, but asks about the drivers of growth rather than
levels

• We have y α 1−α
= Ak h . Taking logs and di erentiating we obtain:

gy = gA + αgk + (1 − α)gh
• This is what this decomposition looks like in the US:

ff
 Takeaways and next week

• Human capital is important. Several angles: covered health and education

• Rising educational attainment has been a source of growth
• Unlikely to drive growth forever.

• Next week: more of a economic and historical perspective on the origins of

modern growth.

38