DAVIS, HAROLD T. The Analysis of Eco nomic Time Series. Pp. xiv, 620. Bloom ington, Ind.: The Principia Press, Inc., 1941. $5.00

• Author: Harold Hotelling
• Published date: 01 May 1942
• Date: 01 May 1942
• DOI: http://doi.org/10.1177/000271624222100159
• Subject Matter: Articles

219

sugar,

dentrifices,

sheeting,

oranges,

wal-

nuts,

lettuce,

shoes,

and

mechanical

re-

frigerators.

Other

representative

fields,

of

greater

importance,

receive

less

detailed

treatment

because

of

less

adequate

data.

On

many

phases

of

the

subject

the

find-

ings cannot

be

justly

summarized.

How-

ever,

the

author

and

his

advisory

committee

agree

in

pointing

out

among

the

beneficial

effects

of

advertising

the

following:

that

it

has

been

a

powerful

force

in

raising

the

standard

of

living;

that

it

has

accelerated

the

improvement

of

products;

and

that

it

has

subsidized

education

and

entertainment

through

radio

and

in

periodicals.

Perhaps

it

was

inevitable

that

emphasis

should

be

placed

on

values

susceptible

of

measurement

in

monetary

terms.

Thus

the

indirect

contribution

of

advertising

to

enter-

tainment

and

education

through

radio

and

periodicals

is

estimated

to

be

one-fifth

of

the

total

cost,

or

nearly

$400,000,000.

No

way

has

been

found

to

measure

the

value

of

more

important

direct

contributions

to

edu-

cation

and

entertainment

in

advertisements

themselves.

A

slight

inconsistency

appears

in

the

dis-

cussions

of

product

differentiations

and

price

differentials.

The

study

justly

points

out

that

product

differentiations

at

any

given

moment

are

usually

small,

and

that

their

value

depends

not

on

the

judgment

of

an

individual

but

on

that

of

the

public

as

a

whole;

moreover,

that

progress

in

the

de-

velopment

of

most

products

results

from

an

accumulation

of

these

small

differentiations.

For

consistency

it

might

seem

that

price

differentials

due

to

consumer

preferences

should

have

been

regarded

as

similarly

jus-

tified.

The

practical

usefulness

of

the

book

in

determining

business

policies

is

limited,

temporarily

at

least,

by

the

fact

that

these

are

no

longer

subject

to

the

individual’s

free

decision.

However,

a

contrasting

back-

ground

may

be

helpful

in

gaining

a

new

perspective

on

the

place

of

advertising

in

a

system

of

free

enterprise

and

in

an

economy

of

abundance.

G.

B.

HOTCHKISS

New

York

University

DAVIS,

HAROLD

T.

The

Analysis

of

Eco-

nomic

Time

Series.

Pp.

xiv,

620.

Bloom-

ington,

Ind.:

The

Principia

Press,

Inc.,

1941.

$5.00.

In

addition

to

the

problems

of

statistics

in

general,

time

series

present

many

special

questions.

To

meet

the

peculiar

problems

of

observations

ordered

in

time,

whether

these

be

of

economic,

biological,

astronomi-

cal,

or

meteorological

character,

a

great

variety

of

statistical

methods

have

been

de-

vised,

chiefly

by

persons

who

knew

little

about

the

general

theory

of

statistics

but

were

faced

by

a

burning

desire

to

draw

in-

ferences

from

observations

of

some

particu-

lar

kind.

These

varied

methods,

some

good,

some

bad,

have been

described

in

articles

scattered

through

journals

dealing

with

all

sorts

of

subjects,

and

several

distinct

litera-

tures

have

grown

up,

largely

independent

of

one

another,

without

the

benefit

of

mutual

criticism

and

articulation

with

more

funda-

mental

considerations

common

to

all

of

them.

It

is

therefore

a

distinct

service

to

have

brought

together

in

one

volume

a

great

part

of

the

statistical

methods

that

have

been

applied

to

time

series.

The

impressive

volume

by

Professor

Davis

does

this,

and

much

more.

It

is

concerned

not

only

with

the

technique

of

analysis

but

also

with

the

results

of

actual

analyses

of

many

historical

series,

and

fur-

ther,

with

the construction

of

an

economic-

historical-political

cosmogony

based

on

the

statistical

findings.

This

is

an

argument

of

almost

Miltonian

dimensions,

soaring

far

above

the

Aonian

mount

on

the

multi-cylin-

der

engine

of

harmonic

analysis

while

it

pursues

a

synthesis

of

history

and

eco-

nomics

by

a

combination

of

higher

mathe-

matics,

historical

references,

and

imagina-

tive

speculation

unattempted

hitherto

by

gods

or

men.

Some

idea of

the

scope

of

the

work

may

be

gained

from

the

index,

which

lists

references

to

Dinsmore

Alter,

Oskar

Anderson,

J.

W.

Angell,

Thomas

Bayes,

D.

Bernoulli,

Sir

William

Beveridge,

A.

L.

Bowley,

Tycho

Brahe,

D.

Brunt,

A.

F.

Burns,

J.

M.

Clark,

Christopher

Columbus,

W.

L.

Crum,

Hugh

Dalton,

Paul

Douglas,

Sir

Francis

Drake,

Irving

and

R.

A.

Fisher,

Galileo

and

Galton,

the

Gracchi

and

U.

S.

Grant,

W.

S.

Jevons

and

Andrew

Jackson,

Kelvin

and

Kepler,

Keynes

and

Koopmans,

Lincoln

and

Lotka,

Malthus,

Marschak

and

Marx,

J.

S.

Mill

and

F.

C.

Mills,

Wesley

C.

at SAGE PUBLICATIONS on December 4, 2012ann.sagepub.comDownloaded from