Week 1 Notes: Single Variable

Differentiation

Slopes of Curves
ˆ the derivative is the most important con-
cept you will learn in the course.
ˆ many of the crucial concepts in economics
can be expressed as derivatives.
ˆ a derivative
– a generalization of the slope
is the rate of change of one variable due
to a change in another
ˆ the slope is defined for linear functions
ˆ derivatives are defined for linear and non-
linear functions

1
ˆ some examples include
– marginal cost
– marginal product
– marginal revenue
– marginal utility
– marginal social utility
– marginal benefit
– demand elasticity
– supply elasticity
– government spending multipliers
– investment multipliers
– tax multipliers
ˆ all these describe how one variable changes
due to changes in another

2
Outline
i. The slope of a tangent and the derivative
ii. The derivative
iii. Increasing and decreasing functions re-
visited
iv. Rates of change and their economic Sig-
nificance

3
i. The slope of a tangent and the derivative

ˆ with a linear function,

y = f (x) = α + βx
the slope, call this m, is defined as the
rate of change
ˆ let (x1, y1) and (x2, y2) are arbitrary points
along the graph of (x, f (x))
ˆ note that
y1 = α + βx1, y2 = α + βx2

y2 − y1 = β(x2 − x1)

4
ˆ hence, the slope is equivalent to:
y2 − y1
m=
x2 − x1
=β
rise
≡
run
ˆ or, if you know just the slope and one
point (e.g. x1, y1) on the line you have

y − y1 = β(x − x1)
ˆ this is also referred to as the equation of
the line with slope β which runs through
the point (x1, y1)
ˆ here the slope is

y − y1
m=
x − x1
=β

5
ˆ we want to define the rate of change of a
function, f (x), at a point a
ˆ a logical way to do this would be to con-
sider the rate of change of a linear func-
tion which “ just touches” f (x) at a

6
ˆ consider the function in the figure
ˆ let (x1, y1), (x2, y2) denote the coordi-
nates of the points P and L
ˆ the slope, m, of this curve at a is
y2 − y1 y2 − f (a)
=
x2 − x1 (a + 4) − a
f (a) + 2 − f (a)
=
a+4−a
2
=
4
1
=
2

7
ii. The derivative
ˆ the slope of the “tangent” to f (x) at P
is called the derivative of f at point a
ˆ this definition of a tangent as the line
that “just touches” f (x) is a little vague
ˆ let’s make this a bit more rigorous
ˆ consider a function f (x) and suppose we’re
interested in its rate of change at P
ˆ let Q be another point close to P and
consider a straight line cutting through
P and Q
– this line is known as a secant

8
ˆ suppose P has coordinates (a, f (a)) and
Q has coordinates (a + h, f (a + h))
ˆ the slope of the secant is:
f (a + h) − f (a) f (a + h) − f (a)
=
(a + h) − a h
ˆ this is clearly not exactly equal to the
rate of change of f (x) at a

9
ˆ now, consider a sequence of points such
that
– as we move along the curve towards
P from Q, we define a sequence of
secants defined as P Q1, P Q2 etc . . .
– the slopes of these secants approach
the rate of change of f (x) at P

10
 f (a + h) − f (a)
mP Q =
a+h−a
f (a + h) − f (a)
=
h
f (a + h1) − f (a)
mP Q1 =
a + h1 − a
f (a + h1) − f (a)
=
h1

11
ˆ similarly
f (a + h2) − f (a)
mP Q2 =
h2

f (a + h3) − f (a)
mP Q3 =
h3

12
ˆ as QN approaches P, hN gets smaller
ˆ these “Newton quotients” or “differential
quotients” get closer and closer to the
rate of change of f (x) at a

13
14
ˆ in fact, we define the derivative of f (x)
at a as:
0 f (a + h) − f (a)
f (a) = lim
h→0 h

15
ˆ but, this is not the same as
f (a + 0) − f (a)
6= f 0(a)
0
ˆ the left hand side is actually undefined
ˆ we read limh→0 g(h) as the limit of the
function g(h) as h tends to 0

16
ˆ in fact, the equation for the tangent to
the graph of y = f (x) is defined as that
which has a slope equal to f 0(a)
ˆ the equation for a straight line passing
through (x1, y1) with slope β is given by
y − y1 = β(x − x1)
ˆ therefore, the tangent to the function f (x)
at a passes through (a, f (a)) and has
slope f 0(a)
y − f (a)
f 0(a) =
x−a

17
ˆ therefore, as above, its equation is:

y − f (a) = f 0(a)(x − a)
y = f (a) + f 0(a)(x − a)
ˆ we’ve used “limit” in an intuitive sense
and will continue to do so for a while
ˆ it’s not necessary to get completely rig-
orous at this point

18
ˆ let’s continue to think of the derivative
of a function f (x) at a as the limit
– as we simultaneously allow h to ap-
proach zero
* in the numerator
* and the denominator
of the Newton quotient
f (a + h) − f (a)
h

19
EXAMPLE:: f (x) = x2.
ˆ compute f 0(a) at a = 0, 1, 21
ˆ find the equation for the tangent func-
tion at ( 21 , 14 )

f (a + h) − f (a) (a + h)2 − a2
=
h h
a2 + 2ah + h2 − a2
=
h
2ah h2
= +
h h
= 2a + h

20
ˆ thus

f (a + h) − f (a)
lim = lim 2a + h
h→0 h h→0
= 2a
f 0(0) = 0
f 0(1) = 2
 
1
f0 =1
2

21
ˆ the equation of the tangent line at
 
1 1
(a, f (a)) = ,
2 4
is given by

y − f (a) = f 0(a)(x − a)
y − a2 = 2a(x − a)
y − ( 12 )2 = 2( 12 )(x − 21 )
y − 14 = 1(x − 21 )
y = x − 14

22
iii. Steps for finding derivative of f (a)

1. Add h to a and compute f (a + h)

2. Compute f (a + h) − f (a)
3. For h 6= 0 compute the Newton quotient
f (a + h) − f (a)
h
4. Simplify this as much as possible, that
is, where possible cancel h from the nu-
merator and the denominator
5. Calculate f 0(a) as
f (a + h) − f (a)
lim
h→0 h

23
EXAMPLE:: f (x) = x3
(Step 1)
f (a + h) = (a + h)3
= a3 + 3a2h + 3ah2 + h3
(Step 2)
f (a + h) − f (a) = (a3 + 3a2h + 3ah2 + h3) − a3
= 3a2h + 3ah2 + h3
(Step 3)
f (a + h) − f (a) 3a2h + 3ah2 + h3
=
h h

24
(Step 4)
f (a + h) − f (a)
= 3a2 + 3ah1 + h2
h
(Step 5)
0 f (a + h) − f (a)
f (a) = lim = 3a2
h→0 h

25
EXAMPLE:: f (x) = x2

ˆ f 0(a) = 2a ∀ a, i.e. for all a

ˆ it makes sense to write
f 0(x) = 2x
ˆ consider
f (x) = x3
f 0(x) = 3x2

26
ˆ other notation for f 0(x) given y = f (x)
y 0 = f 0(x)
y 0 = f (1)(x)
ˆ or, in Leibnitz’s notation

dy d
y 0 = f 0(x) = = dy/dx = f (x)
dx dx

27
EXAMPLE:: Given y = x2
dy
= 2x
dx

28
EXAMPLE:: notation often changes, for
examples

P (t) = t2 ⇒ P 0(t) = 2t
Y = K3 ⇒ Y 0 = 3K 2
dA
A = r2 ⇒ = 2r
dr

29
iii. Increasing and decreasing functions re-
visited
ˆ let x1 < x2 be points in an interval I
ˆ if
– f (x2) ≥ f (x1): f is increasing in I
– f (x2) > f (x1): f is strictly increas-
ing in I
– f (x2) ≤ f (x1): f is decreasing in I
– f (x2) < f (x1): f is strictly decreas-
ing in I

30
31
ˆ these conditions sometimes are difficult
to verify
ˆ we also have if, ∀ x ∈ I
– f 0(x) ≥ 0 ⇔ f is increasing in I
– f 0(x) > 0 ⇒ f is strictly increasing
in I
– f 0(x) ≤ 0 ⇔ f is decreasing in I
– f 0(x) < 0 ⇒ f is strictly decreasing
in I

32
ˆ note the uni-directional arrows of the strictly
statements
– this is due to functions such as x3
which is strictly increasing around x =
0 but yet f 0(0) = 0
ˆ we also have if ∀ x ∈ I
– f 0(x) = 0 ⇔ f is constant in I

33
 iv. Rates of change and their economic
significance
ˆ there are many applications of the deriva-
tive in economics
ˆ let f (x) be a function evaluated at two
points, a and a + h
ˆ we often interpret
f (a + h) − f (a)
h
as the average rate of change of f over
the interval from a to a + h

34
EXAMPLE
ˆ let f denote miles traveled as a function
of time
ˆ thus, the Newton Quotient
f (a + h) − f (a)
h
is average mph over the interval [a, a+h]
ˆ therefore
f (a + h) − f (a)
f 0(a) = lim
h→0 h
and f 0(a) is the instantaneous rate of
change of f at a

35
ˆ in contexts in which x represents time,
we often use t to denote the variable of
interest and write

f 0(t) = f˙(t) or x0(t) = ẋ(t)

36
ˆ we are often interested in proportional
rates of change or relative rates of change

f 0(a)
f (a)
ˆ an example would be growth rates of in-
come and population
ˆ also elasticities
a
0
Elf a ≡ f (a) ·
f (a)
∆y
y
≈ ∆x
x
∆y x
=
∆x y

37
Growth Rates
ˆ suppose N (t) is population at time
ˆ N (t + h) − N (t) is the change in popu-
lation
N (t+h)−N (t)
ˆ h is the average rate of change
in pop per unit of time
N (t+h)−N (t)
ˆ limh→0 h is the instantaneous
rate of change in the population

38
EXAMPLE
ˆ suppose P (t) = 6.4t + 641 population of
Europe, in millions
ˆ d
dt P (t) = 6.4 million per year

39
Economic examples
ˆ C(x) cost of producing x units
ˆ R(x) revenue from producing x units
ˆ Π(x) = R(x) − C(x) profits from pro-
ducing and selling x units
ˆ C 0(x) marginal cost at x
ˆ R0(x) marginal revenue at x
ˆ Π0(x) marginal profit

40
ˆ economists often refer to the marginal
propensity to consume (MPC).
ˆ let C = f (y) where C is consumption
and y is income
ˆ C 0 is the marginal propensity to consume
where

C 0 = f 0(y)
typically with 0 < C 0 < 1

41
EXAMPLE:
ˆ for a cost function C(x) the marginal
cost is given by

C(x + h) − C(x)
C 0(x) = lim
h→0 h
ˆ if x is large, 1 is relatively small:
C(x + 1) − C(x)
C 0(x) ∼
=
1
≈ C(x + 1) − C(x)
= ∆C(x)
ˆ often, however, you will see incremental
costs as marginal costs in your economics
classes as your professors choose to avoid
using calculus terminology

42
Differentiability and empirical functions
ˆ some functions are not differentiable, but
the derivatives can be approximated.
EXAMPLE
ˆ let K(t) denote capital stock at time t

K 0(t) = K̇(t)
0 d
K (t) = K(t)
dt
≡ I(t)
ˆ I(t) is investment
ˆ for actual accounting purposes
I(t) = K(t) − K(t − 1)

43