Graduate Macro Theory II:

Notes on Log-Linearization
Eric Sims
University of Notre Dame

Spring 2011

The solutions to many discrete time dynamic economic problems take the form of a system of
non-linear difference equations. There generally exists no closed-form solution for such problems.
As such, we must result to numerical and/or approximation techniques.
One particularly easy and very common approximation technique is that of log linearization.
We first take natural logs of the system of non-linear difference equations. We then linearize the
logged difference equations about a particular point (usually a steady state), and simplify until we
have a system of linear difference equations where the variables of interest are percentage deviations
about a point (again, usually a steady state). Linearization is nice because we know how to work
with linear difference equations. Putting things in percentage terms (that’s the “log” part) is nice
because it provides natural interpretations of the units (i.e. everything is in percentage terms).
First consider some arbitrary univariate function, f (x). Taylor’s theorem tells us that this can
be expressed as a power series about a particular point x∗ , where x∗ belongs to the set of possible
x values:
00
f 0 (x∗ ) f (x∗ ) f (3) (x∗ )
f (x) = f (x∗ ) + (x − x∗ ) + (x − x∗ )2 + (x − x∗ )3 + ...
1! 2! 3!
00
Here f 0 (x∗ ) is the first derivative of f with respect to x evaluated at the point x∗ , f (x∗ ) is the
second derivative evaluated at the same point, f (3) is the third derivative, and so on. n! reads “n
factorial” and is equal to n! = n(n − 1)(n − 2) · ... · 1. In words, the factorial of n is the product
of all non-negative integers less than or equal to n. Hence 1! = 1, 2! = 2 · 1 = 2, 3! = 3 · 2 · 1 = 6,
and so on.
For a function that is sufficiently smooth, the higher order derivatives will be small, and the
function can be well approximated (at least in the neighborhood of the point of evaluation, x∗ )
linearly as:

f (x) = f (x∗ ) + f 0 (x∗ ) (x − x∗ )

Taylor’s theorem also applies equally well to multivariate functions. As an example, suppose
we have f (x, y). The first order approximation about the point (x∗ , y ∗ ) is:

1
 f (x, y) ≈ f (x∗ , y ∗ ) + fx (x∗ , y ∗ ) (x − x∗ ) + fy (x∗ , y ∗ ) (y − y ∗ )

Here fx denotes the partial derivative of the function with respect to x and similarly for y.
Suppose that we have the following (non-linear) function:

g(x)
f (x) =
h(x)
To log-linearize it, first take natural logs of both sides:

ln f (x) = ln g(x) − ln h(x)

Now use the first order Taylor series expansions:

f 0 (x∗ )
ln f (x) ≈ ln f (x∗ ) + (x − x∗ )
f (x∗ )
g 0 (x∗ )
ln g(x) ≈ ln g(x∗ ) + (x − x∗ )
g(x∗ )
h0 (x∗ )
ln h(x) ≈ ln h(x∗ ) + (x − x∗ )
h(x∗ )

d ln f (x) f 0 (x)
The above follows from the fact that dx = f (x) . Now put these all together:

f 0 (x∗ ) g 0 (x∗ ) h0 (x∗ )

ln f (x∗ ) + (x − x ∗
) = ln g(x ∗
) + (x − x ∗
) − ln h(x ∗
) − (x − x∗ )
f (x∗ ) g(x∗ ) h(x∗ )

Group terms:

f 0 (x∗ ) g 0 (x∗ ) h0 (x∗ )

ln f (x∗ ) + (x − x ∗
) = ln g(x ∗
) − ln h(x ∗
) + (x − x ∗
) − (x − x∗ )
f (x∗ ) g(x∗ ) h(x∗ )

But since ln f (x∗ ) = ln g(x∗ ) − ln h(x∗ ), these terms cancel out, leaving:

f 0 (x∗ ) ∗ g 0 (x∗ ) ∗ h0 (x∗ )

(x − x ) = (x − x ) − (x − x∗ )
f (x∗ ) g(x∗ ) h(x∗ )
To put everything in percentage terms, multiply and divide each term by x∗ :

x∗ f 0 (x∗ ) (x − x∗ ) x∗ g 0 (x∗ ) (x − x∗ ) x∗ h0 (x∗ ) (x − x∗ )

= −
f (x∗ ) x∗ g(x∗ ) x∗ h(x∗ ) x∗
(x−x∗ )
For notational ease, define x
e= x∗ , or the percentage deviation of x about x∗ . Then we
have:

2
 x∗ f 0 (x∗ ) x∗ g 0 (x∗ ) x∗ h0 (x∗ )
x = x − x
f (x∗ ) g(x∗ ) h(x∗ )
e e e

The above discussion and general cookbook procedure applies equally well in multivariate con-
texts. To summarize, the cookbook procedure for log-linearizing is:

1. Take logs

2. Do a first order Taylor series expansion about a point (usually a steady state)

3. Simplify so that everything is expressed in percentage deviations from steady state

A number of examples arise in economics. I will log-linearize the following four examples: (a)
Cobb-Douglass production function; (b) accounting identity; (c) capital accumulation equation;
and (d) consumption Euler equation.

(a) Cobb-Douglass Production Function: Consider a Cobb-Douglas production function:

yt = at ktα nt1−α

First take logs:

ln yt = ln at + α ln kt + (1 − α) ln nt

Now do the Taylor series expansion about the steady state values:

1 1 α (1 − α)
ln y ∗ + ∗
(yt − y ∗ ) = ln a∗ + ∗ (at − a∗ ) + α ln k ∗ + ∗ (kt − k ∗ ) + (1 − α) ln n∗ + (nt − n∗ )
y a k n∗

As above, note that ln y ∗ = ln a∗ + α ln k ∗ + (1 − α) ln n∗ , so these terms cancel:

1 1 α (1 − α)
∗
(yt − y ∗ ) = ∗ (at − a∗ ) + ∗ (kt − k ∗ ) + (nt − n∗ )
y a k n∗
Now using our definition of “tilde” variables being percentage deviations from steady state, we
have:

yet = e kt + (1 − α)e
at + α e nt

(b) Accounting Identity: Consider the closed economy accounting identity:

yt = ct + it

Take logs:

ln yt = ln (ct + it )

3
 Now do the first order Taylor series expansion:

1 1 1
ln y ∗ + (yt − y ∗ ) = ln (c∗ + i∗ ) + ∗ (ct − c∗ ) + ∗ (it − i∗ )
y∗ (c + i∗ ) (c + i∗ )
Now we have to fiddle with this a bit more than we did for the production function case. First,
note that ln (c∗ + i∗ ) = ln y ∗ , so that these terms cancel out:

1 1 1
∗
(yt − y ∗ ) = ∗ ∗
(ct − c∗ ) + ∗ (it − i∗ )
y (c + i ) (c + i∗ )
Now multiply and divide (so as to leave the expression unchanged) each of the two terms on
the right hand side by c∗ and i∗ , respectively:

1 ∗ c∗ (ct − c∗ ) i∗ (it − i∗ )
(y t − y ) = +
y∗ (c∗ + i∗ ) c∗ (c∗ + i∗ ) i∗
Now simplify and use our “tilde” notation:

c∗ i∗ e
yet = ct + it
y∗ y∗
e

(c) Capital Accumulation Equation: Consider the standard capital accumulation equation:

kt+1 = it + (1 − δ)kt

Take logs:

ln kt+1 = ln(it + (1 − δ)kt )

Do the first order Taylor series expansion:

1 1 (1 − δ)
ln k ∗ + (kt+1 − k ∗ ) = ln(i∗ + (1 − δ)k ∗ ) + ∗ (it − i∗ ) + ∗ (kt − k ∗ )
k∗ (i + (1 − δ)k ∗ ) (i + (1 − δ)k ∗ )

Now simplify terms a bit, noting that ln(i∗ + (1 − δ)k ∗ ) = ln k ∗ , so that again terms cancel:

1 1 (1 − δ)
∗
(kt+1 − k ∗ ) = ∗ (it − i∗ ) + (kt − k ∗ )
k k k∗
Now multiply and divide the first term on the right hand side by i∗ :

1 ∗ i∗ (it − i∗ ) (1 − δ)
(k t+1 − k ) = + (kt − k ∗ )
k∗ k∗ i∗ k∗
Using our “tilde” notation:

4
 i∗
kt+1 = ∗ eit + (1 − δ)e
e kt
k

(d) Consumption Euler equation: Consider the standard consumption Euler equation that
emerges from household optimization problems with CRRA utility:
 σ
ct+1
= β(1 + rt )
ct
σ > 0 is the coefficient of relative risk aversion. Take logs:

σ ln ct+1 − σ ln ct = ln β + ln(1 + rt )

Now do the first order Taylor series expansion:

σ σ 1
σ ln c∗ + ∗
(ct+1 − c∗ ) − σ ln c∗ − ∗ (ct − c∗ ) = ln β + ln(1 + r∗ ) + (rt − r∗ )
c c 1 + r∗
Some terms on the left hand side obviously cancel:

σ σ 1
∗
(ct+1 − c∗ ) − ∗ (ct − c∗ ) = ln β + ln(1 + r∗ ) + (rt − r∗ )
c c 1 + r∗
Note that, in the steady state, 1 + r∗ = β1 , hence ln(1 + r∗ ) = − ln β. Using this, we have:

σ σ 1
∗
(ct+1 − c∗ ) − ∗ (ct − c∗ ) = (rt − r∗ )
c c 1 + r∗
There are two semi-standard things to do with the right hand side. First, since rt is already
a percent, it is common to leave it in absolute (as opposed to percentage) deviations. Hence, we
can define ret = (rt − r∗ ), while, for all other variables, like consumption, we use the tilde notation
∗)
to denote percentage deviations, so e ct = (ctc−c
∗ , as before. Secondly, we approximate the term
1
1+r∗ = 1. If the discount factor is sufficiently high, this will be a good approximation. Then,
simplifying, we can write:

1
ct+1 − e
e ct = ret
σ
This says that the growth rate of consumption is approximately proportional to the deviation
of the real interest rate from steady state, with σ1 interpreted as the elasticity of intertemporal
substitution.