L ABOUR E CONOMICS

E DUCATION AND T RAINING

Costanza Naguib
(University of Geneva)
H UMAN C APITAL – 1

I What makes workers more productive?

I their ability to do things

I Economists call this ability human capital.

I It is the combination of innate intellectual ability and

accumulated knowledge/capabilities.
H UMAN C APITAL – 1

I What makes workers more productive?

I their ability to do things

I Economists call this ability human capital.

I It is the combination of innate intellectual ability and

accumulated knowledge/capabilities.
H UMAN C APITAL – 1

I What makes workers more productive?

I their ability to do things

I Economists call this ability human capital.

I It is the combination of innate intellectual ability and

accumulated knowledge/capabilities.
H UMAN C APITAL – 2

I How can one accumulate human capital?

1. informally in the family;

2. through formal schooling;

3. training after schooling;

4. learning by doing.
H UMAN C APITAL – 3

Source: OECD 2016.

I NVESTMENTS IN S CHOOLING : C OSTS

Economic agents choose how much schooling to acquire by

comparing costs and benefits:

I The costs of education:

I direct monetary costs (school and university fees, books

and other learning materials, et.);

I forgone earnings (incomes that could have been earned

working instead of studying);

I physical and psychological costs (studying is hard!).

I NVESTMENTS IN S CHOOLING : C OSTS

Economic agents choose how much schooling to acquire by

comparing costs and benefits:

I The costs of education:

I direct monetary costs (school and university fees, books

and other learning materials, et.);

I forgone earnings (incomes that could have been earned

working instead of studying);

I physical and psychological costs (studying is hard!).

I NVESTMENTS IN S CHOOLING : B ENEFITS

I The benefits of education:

I higher incomes (also higher probability of employment);

I better jobs (more interesting and more fun);

I complementarity between schooling and leisure (higher

ability to enjoy certain types of consumption goods, etc.).

I Both costs and benefits need to be evaluated at the time of

making the investment decision. . .
I NVESTMENTS IN S CHOOLING : B ENEFITS

I The benefits of education:

I higher incomes (also higher probability of employment);

I better jobs (more interesting and more fun);

I complementarity between schooling and leisure (higher

ability to enjoy certain types of consumption goods, etc.).

I Both costs and benefits need to be evaluated at the time of

making the investment decision. . .
T HE P RESENT D ISCOUNTED VALUE (PDV)

I I want to give you 1 CHF. Do you prefer that I give it to you

today or tomorrow?

I 1 CHF today is better than 1 CHF tomorrow for two

reasons:

1. Investment: if I had 1 CHF today I could invest it and get

(1 + i) CHF tomorrow;

2. Preferences: consuming today is better than consuming

tomorrow (other things equal).

I Every agent discounts the future at different rates r .

T HE P RESENT D ISCOUNTED VALUE (PDV)

I I want to give you 1 CHF. Do you prefer that I give it to you

today or tomorrow?

I 1 CHF today is better than 1 CHF tomorrow for two

reasons:

1. Investment: if I had 1 CHF today I could invest it and get

(1 + i) CHF tomorrow;

2. Preferences: consuming today is better than consuming

tomorrow (other things equal).

I Every agent discounts the future at different rates r .

T HE P RESENT D ISCOUNTED VALUE (PDV)

I I want to give you 1 CHF. Do you prefer that I give it to you

today or tomorrow?

I 1 CHF today is better than 1 CHF tomorrow for two

reasons:

1. Investment: if I had 1 CHF today I could invest it and get

(1 + i) CHF tomorrow;

2. Preferences: consuming today is better than consuming

tomorrow (other things equal).

I Every agent discounts the future at different rates r .

C OMPUTING THE PDV – 1

I If one is indifferent between consuming today or tomorrow

than one should be indifferent between having 1 CHF
today and (1 + i) CHF tomorrow;

I Or between 1 CHF today and (1 + i)(1 + i) = (1 + i)2 the

day after tomorrow;

I Or between 1 CHF today and

(1 + i)(1 + i)(1 + i) = (1 + i)3 in three-day time;

I And so on...
C OMPUTING THE PDV – 2

I If we introduce time preferences, each of us will discount

future CHF at an individual-specific rate r , usually larger
than i;

I So, (1 + r ) CHF tomorrow are worth 1 CHF today;

I How much is 1 CHF tomorrow worth today? 1

(1+r )
C OMPUTING THE PDV – 2

I If we introduce time preferences, each of us will discount

future CHF at an individual-specific rate r , usually larger
than i;

I So, (1 + r ) CHF tomorrow are worth 1 CHF today;

I How much is 1 CHF tomorrow worth today? 1

(1+r )
C OMPUTING THE PDV – 3

I Consider the decision to spend one additional year in

school;

I If I decide to go to school I would have to pay the cost of

schooling for one year and then get wage w e=1 later on.

I If I decide not to go to school, then I start working

immediately and get a wage equal to w e=0 .
C OMPUTING THE PDV – 4

I Let us compare the two resulting streams of income

discounted at the time of making the decision:

h i
PDV = c − w1e=0 +
" # " #
w1e=1 − w2e=0 w2e=1 − w3e=0
+ +
(1 + r ) (1 + r )2
" #
wTe=1
−1 − wT
e=0
... +
(1 + r )T −1
C OMPUTING THE PDV – E XAMPLE

Costs:
I What is the total direct cost of the bachelor?

I 500 CHF per semester = 500 × 6 = 3000 CHF;

I Estimated living living costs 2000 CHF per month =

2000 × 36 = 72000 CHF;

I Total = 75, 000 CHF or 25, 000 CHF per year.

C OMPUTING THE PDV – E XAMPLE

Benefits:
I How much more can you make with a bachelor?

I Available estimates say around 50% more

I around 9,000CHF against around 6,000CHF per month.

I What is your discount rate?

I Interest rates are pretty low now (say 1%), add some
preference for early consumption...4%?

Total PDV = 480,000CHF approx.

C OMPUTING THE PDV – E XAMPLE

Benefits:
I How much more can you make with a bachelor?

I Available estimates say around 50% more

I around 9,000CHF against around 6,000CHF per month.

I What is your discount rate?

I Interest rates are pretty low now (say 1%), add some
preference for early consumption...4%?

Total PDV = 480,000CHF approx.

C OMPUTING THE PDV – E XAMPLE

Benefits:
I How much more can you make with a bachelor?

I Available estimates say around 50% more

I around 9,000CHF against around 6,000CHF per month.

I What is your discount rate?

I Interest rates are pretty low now (say 1%), add some
preference for early consumption...4%?

Total PDV = 480,000CHF approx.

S TUDYING OR W ORKING ?

Earnings
Income Stream B: Studying & Working

Gross Benefits
Income Stream A: Working

A
Forgone
Earnings
19
Age
22
Costs

Costs
S TUDYING OR W ORKING ?

Earnings
Income Stream B: Studying & Working

Gross Benefits
Income Stream A: Working

A
Forgone
Earnings
19
Age
22
Costs

Costs
S TUDYING OR W ORKING ?

Earnings
Income Stream B: Studying & Working

Gross Benefits
Income Stream A: Working

A
Forgone
Earnings
19
Age
22
Costs

Costs
S TUDYING OR W ORKING ?

Earnings
Income Stream B: Studying & Working

Gross Benefits
Income Stream A: Working

A
Forgone
Earnings
19
Age
22
Costs

Costs
S TUDYING OR W ORKING ?

Earnings
Income Stream B: Studying & Working

Gross Benefits
Income Stream A: Working

A
Forgone
Earnings
19
Age
22
Costs

Costs
S TUDYING OR W ORKING ?

Earnings
Income Stream B: Studying & Working

Gross Benefits
Income Stream A: Working

A
Forgone
Earnings
19
Age
22
Costs

Costs
S TUDYING OR W ORKING ?

Earnings
Income Stream B: Studying & Working

Gross Benefits
Income Stream A: Working

A
Forgone
Earnings
19
Age
22
Costs

Costs
W HO S TUDIES M ORE

I Young people!
I who have more time ahead to accumulate the benefits of
higher earnings [large T]

I Those who are less impatient

I small r

I Those who expect high wage returns to schooling

I what type of schooling?
R ETURNS FROM E DUCATION

Source: the Economist,

http://www.economist.com/blogs/graphicdetail/2015/03/daily-chart-2.
R ETURNS FROM E DUCATION

Source: the Economist,

http://www.economist.com/blogs/graphicdetail/2015/03/daily-chart-2.
R ETURNS FROM E DUCATION

Source: the Economist,

http://www.economist.com/blogs/graphicdetail/2015/03/daily-chart-2.
R ETURNS FROM E DUCATION

Source: the Economist,

http://www.economist.com/blogs/graphicdetail/2015/03/daily-chart-2.
R ETURNS FROM E DUCATION

Source: the Economist,

http://www.economist.com/blogs/graphicdetail/2015/03/daily-chart-2.
R ETURNS FROM E DUCATION

Source: the Economist,

http://www.economist.com/blogs/graphicdetail/2015/03/daily-chart-2.
T HE S IGNALLING M ODEL OF E DUCATION

I The model was elaborated by Prof. Michael Spence (Nobel

Prize Winner, 2001);

I The model was published in 1973 in the Quarterly Journal

of Economics.

I It is a model of “useless education”.

I In the model education is a pure signaling device.

U SELESS E DUCATION

I Imagine a world in which education has no effect on

people’s human capital.

I People are born good or bad, productive or not-productive;

I ...but nobody is born with a label on the forefront saying

whether one is good or bad.

I Employers want to hire good workers but cannot tell who is

good and who is bad...
E DUCATION AS A S IGNALLING D EVICE

I Employers cannot tell who is good and who is bad.

I But they can see who is educated and who is not.

I If, for some reason, the good people end up studying more
than the bad ones, then employers may use the
observable level of schooling to tell the two groups apart.
S IGNALLING T HEORY OF E DUCATION – 1

Consider the following simplified framework:

I Good workers produce as independently of how much they
have studied;

I Bad workers produce au < as independently of how much

they have studied.

I Studying is more costly for the bad ones:

e
c(e, ai ) =
ai

with e being the level of education.

S IGNALLING T HEORY OF E DUCATION – 2

I Let us simplify even more and assume there are only two
possible levels of schooling, e = 0 or e = 1.

I Acquiring e = 0 costs zero to both good and bad students:

0
c(e = 0, ai ) = =0
ai

I Acquiring e = 1 costs:

1
c(e = 1, as ) = for the good students
as
1
c(e = 1, au ) = for the bad students
au
S IGNALLING T HEORY OF E DUCATION – 2

I Let us simplify even more and assume there are only two
possible levels of schooling, e = 0 or e = 1.

I Acquiring e = 0 costs zero to both good and bad students:

0
c(e = 0, ai ) = =0
ai

I Acquiring e = 1 costs:

1
c(e = 1, as ) = for the good students
as
1
c(e = 1, au ) = for the bad students
au
S IGNALLING T HEORY OF E DUCATION – 2

I Let us simplify even more and assume there are only two
possible levels of schooling, e = 0 or e = 1.

I Acquiring e = 0 costs zero to both good and bad students:

0
c(e = 0, ai ) = =0
ai

I Acquiring e = 1 costs:

1
c(e = 1, as ) = for the good students
as
1
c(e = 1, au ) = for the bad students
au
S IGNALLING T HEORY OF E DUCATION – 3
The employer offers a higher salary to those who are educated with
e = 1.

Wage

Wh

Wl

Education
0 1
S IGNALLING T HEORY OF E DUCATION – 4

I Given this situation, is it the case that the good workers

choose e = 1 and the bad ones choose e = 0?

I Sometimes yes (separating equilibrium) sometimes no

(pooling equilibrium).
S IGNALLING T HEORY OF E DUCATION – 4

I Given this situation, is it the case that the good workers

choose e = 1 and the bad ones choose e = 0?

I Sometimes yes (separating equilibrium) sometimes no

(pooling equilibrium).
S IGNALLING T HEORY OF E DUCATION – 5

Wage

Cu

Ws B

D
A
Wu Benefit

Cost

Education
0 1
S IGNALLING T HEORY OF E DUCATION – 5

Wage

Cu

Ws B

D
A
Wu Benefit

Cost

Education
0 1
S IGNALLING T HEORY OF E DUCATION – 5

Wage

Cu

Ws B

D
A
Wu Benefit

Cost

Education
0 1
S IGNALLING T HEORY OF E DUCATION – 5

Wage

Cu

Ws B

D
A
Wu Benefit

Cost

Education
0 1
S IGNALLING T HEORY OF E DUCATION – 6

Wage

Cu

Ws B

Benefit-Cost (if e = 1)

D
A
Wu

Benefit-Cost (if e = 0)

Education
0 1
S IGNALLING T HEORY OF E DUCATION – 6

Wage

Cu

Ws B

Benefit-Cost (if e = 1)

D
A
Wu

Benefit-Cost (if e = 0)

Education
0 1
S IGNALLING T HEORY OF E DUCATION – 6

Wage

Cu

Ws B

Benefit-Cost (if e = 1)

D
A
Wu

Benefit-Cost (if e = 0)

Education
0 1
S IGNALLING T HEORY OF E DUCATION – 7

Wage

Ws B

Cs

A
Wu Benefit
F

Cost

Education
0 1
S IGNALLING T HEORY OF E DUCATION – 7

Wage

Ws B

Cs

A
Wu Benefit
F

Cost

Education
0 1
S IGNALLING T HEORY OF E DUCATION – 7

Wage

Ws B

Cs

A
Wu Benefit
F

Cost

Education
0 1
S IGNALLING T HEORY OF E DUCATION – 7

Wage

Ws B

Cs

A
Wu Benefit
F

Cost

Education
0 1
S IGNALLING T HEORY OF E DUCATION – 8

Wage

Ws B

Cs
Benefit-Cost (if e = 1)
A
Wu
F
Benefit-Cost (if e = 0)

Education
0 1
S IGNALLING T HEORY OF E DUCATION – 8

Wage

Ws B

Cs
Benefit-Cost (if e = 1)
A
Wu
F
Benefit-Cost (if e = 0)

Education
0 1
S IGNALLING T HEORY OF E DUCATION – 8

Wage

Ws B

Cs
Benefit-Cost (if e = 1)
A
Wu
F
Benefit-Cost (if e = 0)

Education
0 1
S EPARATING E QUILIBRIUM

I For the unskilled, the optimal level of schooling is zero

(A − 0 is larger than B − D);

I For the skilled the optimal level of schooling is 1 (B − F is

larger than A − 0);

I In this case the employer can use schooling as a signal of

ability.

I In the data we would see a positive correlation between

education and earnings even if education has no real
value.
P OOLING E QUILIBRIUM

I This result depends on how we have drawn the curves

(particularly the cost);

I There could very well be situations in which separation is

not possible (pooling equilibrium).

I What would employers do?

I Offer the same wage to everybody (presumably some

average between ws and wu ).

I This is more likely to happen if: cu and cs are similar,

and/or as and au are similar.
P OOLING E QUILIBRIUM

I This result depends on how we have drawn the curves

(particularly the cost);

I There could very well be situations in which separation is

not possible (pooling equilibrium).

I What would employers do?

I Offer the same wage to everybody (presumably some

average between ws and wu ).

I This is more likely to happen if: cu and cs are similar,

and/or as and au are similar.