Addis Ababa University

College of Business and

Economics
Econ 605: Mathematics for
Economists
2020/21 AY
Tewodros Negash Kahsay (PhD)

Tewodros Negash (PhD), Addis Ababa

1
University, Department of Economics
5. Differential Equations
Definitions and concepts
• A differential equation is an equation that expresses
relationship between a function 𝒚 = 𝒇(𝒕) and one or
more of its derivatives or differentials
Examples:
(𝒂) 𝒚′(𝒕) = 𝟓𝒕 + 𝟕 ; (𝒃) 𝒚′(𝒕) = 𝟗𝒚(𝒕);
(𝒄) 𝒚′′(𝒕) – 𝟗𝒚′(𝒕) = 𝟎
• The solution of a differential equation is an equation
without derivative or differential
• The order of a differential equation is the order of the
highest derivative in the equation
• The degree of a differential is the highest power to which
the derivative of highest order is raised
Tewodros Negash (PhD), Addis Ababa
2
University, Department of Economics
The order and degree of differential equations
𝑑𝑦
• = 2𝑡 + 3 first – order, first – degree
𝑑𝑡
𝑑𝑦 4
• − 2𝑡 5 = 0 first – order, fourth – degree
𝑑𝑡
𝑑2𝑦 𝑑𝑦 3
• + + 𝑡 2 = 0 second – order, first –
𝑑𝑡 2 𝑑𝑡
degree
7 5
𝑑2𝑦 𝑑3𝑦
• + = 7𝑦(𝑡) third – order, fifth –
𝑑𝑡 2 𝑑𝑡 3
degree

Tewodros Negash (PhD), Addis Ababa

3
University, Department of Economics
5.1. Linear, First – Order Differential equations
Autonomous equations
• These are first – order differential equations in which the
variable 𝒕 does not enter the equation explicitly
• Definition: The general form of the linear, autonomous,
first – order differential equations is 𝑦ሶ + 𝑎𝑦 = 𝑏, where 𝒂
and 𝒃 are known constants
– The differential equation is linear because 𝑦ሶ and 𝑦 are
raised to a power of one
– It is of the first – order because one is the highest –
order derivative in the equation
– It is of first – degree because the power of the
derivative of the highest order is one (𝑦ሶ is raised to one)
– It is autonomous because the coefficient 𝒂 and the term
𝒃 are constant
Tewodros Negash (PhD), Addis Ababa
4
University, Department of Economics
If 𝒂 or 𝒃 vary with 𝒕 (i.e. are explicitly functions of 𝒕), the
equation becomes nonautonomous
The solution method
• Finding the general solution relies on the technique of
separating the problem of finding the general solution for
the complete differential equation into two simpler
subproblems
• If 𝑦ℎ denotes the general solution to the homogeneous
form (obtained by setting 𝑏 = 0) and 𝑦𝑝 denotes a
particular solution (as opposed to the general solution) to
the complete equation, we can use 𝑦 = 𝑦ℎ + 𝑦𝑝
• That is, the general solution, 𝑦, for the complete equation is
the sum of the general solution to the homogeneous form
and a particular solution to the complete equation (such as
the steady state solution)
Tewodros Negash (PhD), Addis Ababa
5
University, Department of Economics
The Homogeneous Solution
• Definition: The homogeneous form of the linear,
autonomous, first – order differential equation is
𝑦ሶ + 𝑎𝑦 = 0 , 𝑎 ≠ 0
• The homogeneous form can be solved by direct
𝑦ሶ
integration as 𝑓𝑜𝑙𝑙𝑜𝑤𝑠 = −𝑎
𝑦
• Integrating both sides,
𝑑𝑦
𝑦ሶ ൗ𝑑𝑡
→ ‫׬ = 𝑡𝑑 𝑦 ׬‬−𝑎 → ‫ = 𝑡𝑑 𝑦 ׬‬−𝑎𝑡 + 𝐶1
1
→ ‫ = 𝑦𝑑 𝑦 ׬‬−𝑎𝑡 + 𝐶1 → 𝑙𝑛𝑦 + 𝐶2 = −𝑎𝑡 + 𝐶1
→ 𝑙𝑛𝑦 = −𝑎𝑡 + 𝐶1 − 𝐶2
Taking the antilog of both sides,
→ 𝑦 = 𝑒 −𝑎𝑡+𝐶1 −𝐶2 = 𝑒 −𝑎𝑡 𝑒 𝐶1−𝐶2 = 𝐶𝑒 −𝑎𝑡
• Where 𝐶 = 𝑒 𝐶1 −𝐶2 is still an arbitrary constant of
integration Tewodros Negash (PhD), Addis Ababa
University, Department of Economics
6
The general solution
• Theorem 1: The general solution to the homogenous form of
the linear, autonomous, first – order differential equation is
𝑦ℎ 𝑡 = 𝐶𝑒 −𝑎𝑡 . The subscript 𝒉 in 𝑦(𝑡) refers to
homogenous.
• To check if the solution is correct, substitute the solution
𝑦ℎ 𝑡 into the differential equation, 𝑦ሶ + 𝑎𝑦 = 0, and check
that it equals 0.
−a𝐶𝑒 −𝑎𝑡 + 𝑎𝐶𝑒 −𝑎𝑡 = 0
Example
• Given the differential equation 𝑦ሶ = 3𝑦 + 2 ,
solve the homogenous form of the differential equation
Tewodros Negash (PhD), Addis Ababa
7
University, Department of Economics
The particular solution
A particular solution is the steady – state equilibrium value
of 𝒚
• Definition: A steady – state value of a differential equation
is defined by the condition 𝑦ሶ = 0. It is the value of 𝑦,
which we call 𝑦ത , at which 𝑦 is stationary.
• To find the steady – state value of 𝑦, set 𝑦ሶ = 0 in the
complete differential equation 𝑦ሶ + 𝑎𝑦 = 𝑏
𝑏
i.e, 0 + 𝑎𝑦ത = 𝑏 → 𝑦ത = is the steady – state value of 𝒚
𝑎
• As long as 𝑎 ≠ 0, there does exist a steady – state value.
• In the case of an autonomous, linear, first – order
differential equation with 𝑎 ≠ 0, we will always use the
steady –state value as the particular solution. That is, we
use 𝑦𝑝 = 𝑦ത
Tewodros Negash (PhD), Addis Ababa
8
University, Department of Economics
• To confirm that 𝑦ത is a particular solution to the complete
differential equation, note that the time – derivative of the
𝑏
solution, 𝑦ത , is 0 and substitute back 𝑦ത = into the
𝑎
complete differential equation.
𝑏
0 + 𝑎𝑦ത = 𝑏 → 0 + 𝑎 = 𝑏 → 𝑏 = 𝑏
𝑎
• Thus, 𝑦ത satisfies the complete differential equation and is
therefore a solution
The general solutions
• Theorem 2: The solution to any linear differential equation
is equal to the sum of the homogeneous solution and any
particular solution to the complete differential equation.
That is, 𝒚 = 𝒚𝒉 + 𝒚𝒑
• Using the results for 𝒚𝒉 and 𝒚𝒑 , the general solution to
the complete equation is given by the following theorem.
Tewodros Negash (PhD), Addis Ababa
9
University, Department of Economics
• Theorem 3: The general solution to the complete
autonomous, linear, first – order differential equation is
−𝑎𝑡
𝑏
𝑦 𝑡 = 𝐶𝑒 +
𝑎
• We begin by solving the homogeneous form, and then
find a particular solution to the complete equation.
Example
• solve the differential equation
(a ) 𝑦ሶ + 2𝑦 = 8
(b) The differential for capital in an industry is given by
𝐾ሶ = 𝐼 ҧ − 𝛿𝐾
where 𝛿 is the rate of depreciation of capital and 𝐼 ҧ is a
constant rate of investment in the industry. Derive 𝑲(𝒕),
the quantity of capital available in the industry at time 𝒕.
Tewodros Negash (PhD), Addis Ababa
10
University, Department of Economics
Nonautonomous equations
• If the coefficient, a or the term 𝒃, in the linear
differential equation 𝑦ሶ + 𝑎𝑦 = 𝑏 are functions of time,
the equation is nonautonomous and the solution
method discussed before doesn’t work
• We need a solution technique that works for any linear,
first – order differential equation
• Definition: The general form of the linear, first – order
differential equation is 𝑦ሶ + 𝑎(𝑡)𝑦 = 𝑏(𝑡) where 𝒂(𝒕)
and 𝒃(𝒕) are known, continuous function of 𝒕.
• In this case, we cannot use the approach of solving the
homogeneous form first, and then try to find a particular
solution to the complete equation.
• The problems is we can no longer use the steady – state
solution as a particular solution because when 𝒂(𝒕)
and/or 𝒃(𝒕) are not constant, any value of 𝑦 at which
𝑦ሶ = 0 doesn’t exist in general.
Tewodros Negash (PhD), Addis Ababa
11
University, Department of Economics
• An alternative solution strategy is to use the concept of an
integrating factor that converts any linear, first – order
differential equation into an equation that can be directly
integrated to obtain the general solution
• The solution technique is based on the fact that any linear,
first – order differential equation can be multiplied by a
known factor which converts it into a form that can be
integrated directly.
• Multiply through the differential equation 𝑦ሶ + 𝑎(𝑡)𝑦 = 𝑏(𝑡)
𝑑𝐴(𝑡)
by the term 𝑒 𝐴(𝑡) where 𝐴 𝑡 = ‫ 𝑡𝑑 𝑡 𝑎 ׬‬or = 𝑎(𝑡)
𝑑𝑡

𝑒 𝐴(𝑡) 𝑦ሶ + 𝑎 𝑡 𝑦(𝑡) = 𝑒 𝐴 𝑡 𝑏 𝑡 (∗)

𝑑
• The left hand –side is equal to 𝑒 𝐴 𝑡 𝑦(𝑡) because
𝑑𝑡
𝑑
• 𝑒 𝐴 𝑡 𝑦(𝑡) = 𝑒 𝐴(𝑡) 𝑦ሶ + 𝑎 𝑡 𝑦(𝑡)
𝑑𝑡 Tewodros Negash (PhD), Addis Ababa
12
University, Department of Economics
• Therefore, (∗) can be written as
𝑑 𝐴𝑡
𝑒 𝑦(𝑡) = 𝑒 𝐴 𝑡 𝑏 𝑡 (∗∗)
𝑑𝑡
• The differential equation can thus be solved by direct
integration of (∗∗)

𝑒𝐴 𝑡 𝑦 𝑡 = න 𝑒𝐴 𝑡 𝑏 𝑡 + 𝐶

• Dividing by 𝑒 𝐴 𝑡 yields,

𝑦 𝑡 = 𝑒 −𝐴 𝑡 න 𝑒𝐴 𝑡 𝑏 𝑡 +𝐶

Theorem 4: The general solution to any linear, first – order

differential equation is
𝑦 𝑡 = 𝑒 −𝐴 𝑡
‫ 𝐴 𝑒 ׬‬Tewodros
𝑡 𝑏 𝑡 + 𝐶 where 𝐴 𝑡 = 𝑎 𝑡 𝑑𝑡
Negash (PhD), Addis Ababa
‫׬‬
13
University, Department of Economics
The integrating factor
• In (∗) above, we multiplied both sides of the
differential equation by a specific term.
• This made the differential equation amenable to direct
integration
• Terms that serve this purpose are called integrating
factors.
Theorem 5: The general form of the integrating factor for
the linear, first – order differential equation is 𝑒 𝐴 𝑡
where 𝐴 𝑡 = ‫𝑡𝑑 𝑡 𝑎 ׬‬
• In the special case where the coefficient 𝑎 𝑡 = 𝑎, a
constant, the integrating factor becomes 𝑒 𝑎𝑡
Example:
• solve the differential equation 𝑦ሶ − 2𝑡𝑦 = 𝑏𝑡
Tewodros Negash (PhD), Addis Ababa
14
University, Department of Economics
5.2. Nonlinear, First – Order Differential Equations
• We can apply a single solution technique to solve any
linear, first – order differential equation
• But when the differential equation is nonlinear no single
solution technique will work
– In fact, only a few special classes of nonlinear, first –
order differential equations can be solved at all
– And it is rare in economic applications to be able to
find explicit solutions to nonlinear differential
equations
– Commonly used alternative is to do a qualitative
analysis with the aid of phase diagrams
Tewodros Negash (PhD), Addis Ababa
15
University, Department of Economics
Autonomous equations and qualitative analysis
Qualitative Analysis
• It is common to conduct a qualitative analysis with the
aid of a phase diagram
• Consider the following nonlinear, first – order differential
equation
𝑦ሶ = 𝑦 − 𝑦 2 (2.3)
• Let’s try a qualitative analysis of the solution to equation
(2.3) using phase diagram
• Rearranging the equation yields,
𝑦ሶ = 𝑦(1 − 𝑦) (2.4)
Tewodros Negash (PhD), Addis Ababa
16
University, Department of Economics
• For an autonomous, first – order differential equation, we
can find the steady – state value of 𝒚 by setting 𝑦ሶ = 0
• Applying this to equation (2.4), 𝑦ሶ = 0 at 𝑦 = 0 𝑎𝑛𝑑 𝑦 = 1
• This information is used to construct the phase diagram for
equation (2.4)
𝑦ሶ
Phase diagram for equation (2.4)

0 1
𝑦

• The phase diagram shows 𝑦ሶ as a function of 𝑦.

Tewodros Negash (PhD), Addis Ababa
17
University, Department of Economics
• To draw the diagram we use
(1) The points 𝑦, 𝑦ሶ = 0,0 and 𝑦, 𝑦ሶ = 1,0
(2) The value of 𝒚 at which the curve reaches an extreme
value is determined by setting the derivative of 𝑦ሶ w.r.t 𝒚
equal to zero.
𝑑 𝑦ሶ
= 1 − 2𝑦 = 0 → 𝑦 = 1Τ2
𝑑𝑦
Does the curve reach a maximum or a minimum at 𝑦 = 1Τ2 ?
Check the second derivative of
𝑑 2 𝑦ሶ
= −2 → the curve reaches a maximum at 𝑦 = 1Τ2
𝑑𝑦 2
Thus, we have the shape of the graph depicted in the Figure
above.
• A phase diagram helps us to determine the qualitative
properties of a solution to a differential equation without
actually having to solve it.
Tewodros Negash (PhD), Addis Ababa
18
University, Department of Economics
• It is clear from the diagram that,
– If 𝒚 ever reaches zero or one, it will remain there
forever since 𝑦ሶ = 0
– For 0 < 𝑦 < 1, 𝑦ሶ > 0 → 𝒚 tends to increase over
time till it reaches one where it remains there.
– If 𝑦 > 1, 𝑦ሶ < 0 → 𝒚 will tend to decrease to toward
one
– If 𝑦 < 0, 𝑦ሶ < 0 → 𝒚 will continue to decrease to
−∞
• Our analysis is summarized in the arrows of motion in
the above figure. It indicates that 𝒚 tends to converge to
one if it starts at a value where 𝑦 > 0 and diverges to
− ∞ if it starts at a value where 𝑦 < 0
Tewodros Negash (PhD), Addis Ababa
19
University, Department of Economics
Stability Analysis
• In the above qualitative analysis we required two points
– The steady – state values of 𝒚 (i.e. the values at which
𝑦ሶ = 0)
– The arrows of motion or the motion of the system
around the steady – state values are used to
determine whether or not the system converges to
one of these steady – state values
• In equation (2.4), the system converged to one of the
steady – state value (𝑦 = 1) but not the other (𝑦 = 0)
• Thus, 𝑦 = 1 is a stable equilibrium and 𝑦 = 0 is an
unstable equilibrium
• What makes one stable and the other unstable?
Tewodros Negash (PhD), Addis Ababa
20
University, Department of Economics
Theorem 2.2: A steady – state equilibrium point of
nonlinear, first – order differential equation is stable if
𝑑 𝑦ሶ
the derivative ൗ𝑑𝑦 is negative at that point and
unstable if the derivative is positive at that point.
• Applying theorem 2.2 to the differential equation 2.4,
– At 𝑦 = 0, 𝑑𝑦ሶൗ𝑑𝑦 = 1 − 2 0 = 1
𝑑 𝑦ሶ
– At , 𝑦 = 1, ൗ𝑑𝑦 = 1 − 2 1 = −1
We conclude that 𝑦 = 1 is a stable equilibrium and 𝑦 = 0
is an unstable equilibrium
• The stability condition 𝑑𝑦ሶൗ𝑑𝑦 < 0 at equilibrium values
applies not only to nonlinear, first – order differential
equations but also to linear first – order differential
equations with constant coefficients.
Tewodros Negash (PhD), Addis Ababa
21
University, Department of Economics
Example: Given 𝑦ሶ + 𝑎𝑦 = 𝑏 , the steady – state
equilibrium is 𝑏Τ𝑎
𝑑𝑦ሶ
• The derivative ൗ𝑑𝑦 = −𝑎 is a constant
• Thus the value of 𝒚 will converge to the
equilibrium value
𝑏Τ only if 𝑎 > 0
𝑎

(a) Find the steady – state points and determine

their stability properties for 𝑦ሶ = 3𝑦 2 − 2𝑦

Tewodros Negash (PhD), Addis Ababa

22
University, Department of Economics
5.3. Separable Equations
• Suppose that 𝑥ሶ = 𝐹(𝑡, 𝑥), where 𝐹(𝑡, 𝑥) can be written
as a product of two functions, one of which depends only
on 𝒕 and the other only on 𝒙
• Specially suppose that 𝑥ሶ = 𝑓 𝑡 𝑔 𝑥 (3.1)
– we say that this differential equation is separable
Example:
• 𝑥ሶ = −2𝑡𝑥 2 is apparently separable
• 𝑥ሶ = 𝑡 2 + 𝑥 is not separable
• Separable differential equations are among those that
can be solved in terms of integrals of known equations
Tewodros Negash (PhD), Addis Ababa
23
University, Department of Economics
• Method for Solving Separable Differential Equations
𝑑𝑥
(a) Write equation (3.1) as = 𝑓 𝑡 𝑔 𝑥 (∗)
𝑑𝑡
𝑑𝑥
(b) Separate the variables =𝑓 𝑡 𝑑𝑡
𝑔(𝑥)
𝑑𝑥
(c) Integrate each side ‫)𝑥(𝑔 ׬‬ = ‫𝑡𝑑 𝑡 𝑓 ׬‬
(d) Evaluate the two integrals (if possible) to obtain a
solution of (∗) (possibly in implicit form). Solver for 𝒙, if
possible.
Example:
• Solve the differential equations
𝑑𝑥
(a) = −2𝑡𝑥 2
𝑑𝑡
𝑑𝑥 𝑡3
(b) = 6
𝑑𝑡 𝑥 +1
𝑑𝑥
(c) = 𝑡 2 Tewodros Negash (PhD), Addis Ababa
𝑑𝑡 University, Department of Economics
24
• Determine which of the following differential
equations are separable
a) 𝑥ሶ = 𝑥 2 − 1
b) 𝑥ሶ = 𝑥𝑡 + 𝑡
c) 𝑥ሶ = 𝑥𝑡 + 𝑡 2
d) 𝑥𝑥ሶ = 𝑒 𝑥+𝑡 1 + 𝑡 2
4
e) 𝑥ሶ = 𝑡 2 + 𝑥
f) 𝑥ሶ = 𝐹 𝑡 + 𝐺(𝑥)

Tewodros Negash (PhD), Addis Ababa

25
University, Department of Economics