Chap 18: Overall View of the Economic Approach Applied to

Equilibrium Systems and the Rate of Return Equilibrium Model

1) Set Up the Problem

a) Endogenous Variables
Variables that are determined by the forces in the system.

b) Exogenous Variables
Variables that are constant and unchanging; they are unaffected by
the forces in the system.

Often, simplifying assumptions have to be made. Everything, in a

sense, is endogenous. For the model to be tractable, however, we often
assume a certain cut-off point and decree certain variables to be
exogenous. For example, we might say that the weather is an
exogenous variable in a predator-prey population model, but growth of
a particular plant-eating species might very well affect the weather.
We assume that away when we make weather exogenous.

c) Structural Equations
Tell how the variables are related to each other.

d) Equilibrium Conditions
Are statements telling us when the system is at rest.

2) Finding the Equilibrium Solution

There are a variety of ways to find the repeated values of an equilibrium system.

a) Pencil and paper—use algebra to find the values of the endogenous

variables where the structural equations meet the equilibrium conditions.
Problems can be solved “concretely” or more “generally” as reduced
forms, equilibrium value of an endogenous variable as a function of
exogenous variables alone.

b) Graphs—evaluate the structural equations at a variety of different values

of the endogenous variables in order to see where the equilibrium conditions
are met.
i) An endogenous variable over time
Visually inspect to see if and where it settles down.
ii) An endogenous variable against another endogenous variable
Intersections are often important here.
iii) A phase diagram
The intersection(s) of the phase line and the slope of +1 line
immediately reveal(s) the equilibrium solution(s).

c) Computer—use Excel’s Solver to find the values of the endogenous

variables where the structural equations meet the equilibrium conditions.
The Target Cell is the equilibrium condition and it is set to a value of
0. The Changing Cells are the endogenous variables.

[Link] 1
In equilibrium systems, we are often interested in an exploration of the equilibration
process. Thus, before going to step 3) Comparative Statics, we tackle

2A) Understanding the equilibration process

a) Three Questions about the equilibration process:
i) Is the equilibrium stable or unstable?
ii) How will the system get to equilibrium (oscillatory versus direct)?
iii) How fast will the system get to equilibrium (slowly or quickly)?

b) Graphs used to answer these questions:

i) An endogenous variable over time
ii) An endogenous variable against another endogenous variable
iii) A phase diagram
1) The ALGEBRAIC SIGN of the slope of the phase line at the equilibrium
point reveals information about the type of equilibration process.
2) The MAGNITUDE of the slope of the phase line at the equilibrium point
reveals information about the speed of the equilibration process.

3) Comparative Statics
1) Method of Actual Comparison—∆exo-->∆y e
Find an equilibrium solution expressed as a numerical value of an endogenous
variable, impose a finite-size change in an exogenous variable, recalculate the
numerical value of the equilibrium solution, and then compare the initial and new
equilibrium values.
This comparison can be
1) Qualitative (sign)
2) Quantitative (magnitude)
a) Own units (slope)— ∆ye/∆exo
b) Percentage change (elasticity)—% ∆ye/%∆exo

Excel’s Solver and the Comparative Statics Wizard are ideal for this. You can
quickly recalculate the equilibrium solution for a variety of exogenous variables
and have the results displayed in a nice format. A few simple formulas are needed
to find own unit or percentage change responses and drawing presentation graphs
is a snap.

2) Method of the Reduced Form—dexo-->dye

If you have a reduced form expression for an equilibrium value, then your work in
comparative statics is a snap. Simply take the derivative of the reduced form with
respect to an exogenous variable of interest.

The algebraic sign of the resulting derivative is a qualitative prediction; while its
magnitude is a quantitative prediction in own units. In addition, if the endogenous
variable appears in the derivative expression itself, you know that the relationship
is non-linear.

[Link] 2
Today’s Task . . .

Let’s apply the outline above to another example. Instead of walking you through every
step, we will ask general questions and allow you to structure your answer in your own
way.

The Rate of Return Equilibrium Model

The Behavior to be Analyzed:

Each year, millions of individuals are faced with an important decision: to go to college
or go directly to work. In this lab, we will explore how the decision to get a college
degree versus working straight out of high school can be modeled as part of an equilibrium
system. The individual decision-makers who choose to go or not to go to college affect the
wages of college and non-college jobs. In turn, these wages influence the college choice of
future decision-makers. This is the feedback mechanism that motivates us to think about
the system as an equilibrium problem.

Some Background on the Rate of Return:

Economists believe that the decision to go not to go to college can be seen as an

investment decision—getting a college degree is an investment in yourself. You pay
now (tuition and foregone earnings today) in order to make more later. As a rational
investor, you should decide to get a college degree if the value of an education is greater
than the value of the alternative.

One way to measure the value of college as an investment is based on the rate of return
to college education. The derivation and calculation of the rate of return, R, is too
involved to explain here, so we merely point out two relevant characteristics:

1) The rate of return is measured as a percent of the principal invested per

year. Thus, a rate of return, R, of .25, or, 25% per year means that you are getting
back $25 per year for every $100 invested. This is usually a very good rate of return.
A rate of return of say 0.02 or 2% per year is an unattractive, low rate of return.

2) The investment you make when you go to college is tuition and foregone
earnings; the return on that investment is higher wages after you graduate
from college. The rate of return to a college degree, Rc, therefore, depends upon
how much higher college graduates’ wages are than high school graduates’ wages.
We will use a variable called WageDiff, the percentage difference between the
wages of jobs done by college graduates versus those done by high school graduates.
If WageDiff is high, then the rate of return to a college degree will be high. On the
other hand, a low wage percentage differential results in a low rate of return. If the
wages of college and high school graduates are the same, the rate of return will be
negative because the costs of college are not compensated for by higher wages.

[Link] 3
 An Important Equation:

We will greatly simplify things by assuming a linear relationship between the rate of
return to a college degree and the college/high school wage differential. We can represent
the two points discussed above by the following equation:

R c = m A Wage Diff + b
where:
Rc = rate of return to a college degree (units are percent per year)
m = exogenous slope coefficient (units are 1/year)

college wage − high school wage

Wage Diff = A 100%
high school wage (units are percent)
b = exogenous intercept coefficient, b < 0 (units are percent per year)

For example, if m = 2, b = - 20%/year, and WageDiff = 25%, then Rc = 30%/year. That’s

pretty good. Probably lots of graduating high school seniors would then choose to go to
college. There would then be lots of people available for jobs that require college degrees
and this would drive the WageDiff down, thus, driving down the rate of return for college.

Making the Decision to Go or Not Go to College:

Obviously, a college education costs a lot these days in money and time, resources that
could be invested elsewhere (such as opening your own business or even just putting that
money in the bank and letting it accumulate interest). Suppose every individual had the
opportunity to invest in only one alternative to college that yielded a 10% rate of return
(Ra = Rate of return for the alternative investment = 10%). We assume that Ra is some
given, fixed constant.

College would be a good investment if the Rate of return to college, Rc, were greater than
10%; college would be a bad investment if the rate of return to college were less than 10%.

So the rule is very simple:

Rc > Ra --> go to college after high school

Rc < Ra --> go to work after high school (and take the rate of return of the alternative
investment)

If Rc = Ra, then it doesn’t matter, the two paths (going to college versus going to
work) are exactly equivalent. Flip a coin.

[Link] 4
The Equilibrium System:

With this background, we can now set up the equilibrium system. The idea is that
the excess rate of return (or ExcessRc = Rc - Ra) is the signal that informs the
decision of whether or not to go to college which, in turn, determines how many
people get college degrees and, thus, the college minus high school wage differential
(which is the result).

The feedback mechanism is captured by the fact that as more people get college
degrees, the supply of college grads increases, their wages fall, and the rate of
return to a college degree falls. Conversely, if very few people get degrees, the
supply curve of college grads shifts left (because the number of people exiting the
market, as they retire or die, are greater than those entering it), wages rise, and Rc
rises.

Decision
go to college
or go to
work

Signal ExcessRc (Rc - Ra) WageDiff Result

When we examine the equilibration process in this model, we’ll vary the rate at
which high school grads respond to the signal. That is the adjustment parameter
that will determine the type of equilibration process.

From Simple Model to Reality:

In the real world, what happens is that different people face different alternative
rates of return. That’s why we do not see every single high school student making
the same decision. In addition, the real world is complicated by the fact that
students act on perceived rates of return (it seems people tend to over estimate the
costs of college) and they may be hampered by capital constraints (with a marked
distaste for borrowing).

Even so, labor economists believe that the decision to go to college (and also what to
major in) is influenced by the prospective return. In the United States today, the
rate of return to a college degree is around 12-15% or so. That number has varied
somewhat, but it has not decreased as would have been expected. We will let your
Labor Economics professor fill you in on the details and turn our attention back to
applying the Economic Approach to our simple Rate of Return Equilibrium Model.

[Link] 5
Part 1) Finding the Equilibrium Solution

Let’s begin this part by examining a Concrete Problem. We set up the problem by
dividing the variables into endogenous and exogenous variables. The variables are
Rc, m, WageDiff, b, and Ra.

Endogenous Variables Exogenous Variables

R c , the rate of return to the m, the slope coefficient in the rate of
investment of obtaining a college return model
degree
b, the intercept coefficient in the rate
WageDiff, the % difference between of return model
wages of jobs employing college grads
and those employing high school R a , the rate of return in an alternative
grads investment

The Structural Equations for a Concrete Problem are given below:

Rc = 2 WageDiff + (- 20%/year)
Ra = 10%/year

It’s a “Concrete Problem” because the exogenous variables have been given explicit
numerical values.

The Equilibrium Condition is:

Rc - Ra = 0

Launch Excel and open the file called [Link]. There you will complete
Part 2: Finding the Equilibrium Solution and Part 3: Comparative Statics

Use the information provided in this handout to help you construct your own
analysis of the Equilibrium Rate of Return Model.

Good luck!

[Link] 6