Modern Economy, 2011, 2, 203-212
doi:10.4236/me.2011.23026 Published Online July 2011 (http://www.SciRP.org/journal/me)
Copyright © 2011 SciRes. ME
A New Model for AS-AD Analysis Based on Input-Output
Frame
Xinjian Liu
School of Economics and Management, Yanshan University, Qinhuangdao, China
E-mail: lxj6309@126.com, lxj@ysu.edu.cn
Received February 11, 2011; revised March 29, 2011; accepted April 18, 2011
Abstract
This paper has established a new kind of AS-AD models with input-output techniques. The models take the
standard input-output tables as its starting points. We analyze the change effects of government consumption,
direct consumption coefficient, labor productivity, and surplus rate on the equilibrium output. In this paper,
we propose that the aggregate demand function curve should be right upward, and this lead to a series of in-
consistent conclusions with the traditional views. Finally, we also analyze the well-known issue of stagfla-
tion.
Keywords: AS-AD Model, Input-Output, Aggregate Supply, Aggregate Demand
1. Introduction
It was pointed out that the input-output (IO) analysis had
no longer been included in the core of mainstream eco-
nomics since the middle of 1980s. The authoritative
magazines, such as Econometrica, the Review of Eco-
nomics and Statistics, and the Quarterly Journal of Eco-
nomics, didn’t continue to publish input-output papers,
and the best economists seemed seldom to have interest
in the development of input-output analysis field [1]. In
fact, input-output (IO) economics has never had access to
the core of mainstream economics. Leontief’s input-
output economics[2] and Keynesian economics[3] were
both generated in the mid-1930s, but the Keynesian the-
ory is the inheritance and development of traditional
mainstream economics and has the same basic category
to it, whereas the Leontief theory born out of Marxist
economics and is completely different from the tradi-
tional classic economics in the core category. The key
difference between them is that the foundation concept of
Keynesian theory is national income such as GNP or GDP,
and the foundation concept of Leontief theory is total
output which contains intermediate inputs, such that the
Keynesian theory’s research object mainly focused on
the transaction processes and the Leontief theory’s object
mainly focused on the production processes. Although
IO analysis has obtained a huge developing space in the
study of practical economies, it is yet impossible to enter
the mainstream economic theory system. They two can
not be combined together seamlessly. However, Leontief’s
theory is closer to reality than the mainstream one be-
cause some intermediate inputs are necessary for most of
real production processes and any unit of products must
contain some products which were produced previously.
The AS-AD analysis approach is a new development stage
of the mainstream economics in the 1980s to 1990s when
it was used to solve the stagflation problems which had
troubled mainstream economics for more than 20 years.
This paper develops a new AS-AD model with a sin-
gle-sector IO framework and analyses the slopes of new
AD and AS functions at first, and then explores the im-
pacts of government expenditure, direct consumption
coefficient, labor productivity and operating surplus rate
on the equilibrium states in the second part. It discusses
the stagflation issue in a two-sector model in the third
part. The last part gives the conclusion views of the paper
2. Single-Sector AS-AD Model
2.1. Basic Input-Output Relations
The model of a single-sector input-output table shown as
following:
12
X
CGINX
W
TT
Z






X. J. LIU
Copyright © 2011 SciRes. ME
204
where X is intermediate input (or intermediate use), C is
household consumption, G is public consumption(or
government consumption), I is capital formation, NX
is net export, W is total wage, 1
T is net taxes on produc-
tion, 2
T is net taxes on income, Z is operating surplus.
For this economic system, its input-output balance re-
lations are following:
X
CGI NXQ  (1)
1
PX W TZPQ (2)
where P is the price level, Q is the total output.
If let d
Q denote aggregate demand and
s
Q denote
aggregate supply, then Formulas (1) and (2) should be
written as following:
d
X
CGI NXQ (3)
1
s
PX W TZPQ (4)
When the supply and demand are balanced, there is
ds
QQ
Formulas (3) and (4) show that the rows of an IO table
represent the demand sides of an economy and the col-
umns represent the production processes.
2.2. Aggregate Demand (AD) Relations
If we have the behavioral functions of the aggregate de-
mand as following:
s
X
Qa,

1,CfWZP, 12
GTT H ,

2,
frZP,

3,
d
NXfQ e,
where a is the direct consumption coefficient,
H
is a
planned deficit, r is an interest rate, e is an exchange
rate.
The above behavioral functions contain following ba-
sic ideas:
1) In a short term, a is constant. This is not only the
basic usage of input-output techniques, but also con-
forms to the characteristics of short-term economy.
2) For a given economic system, consumption was
mainly influenced by income. The resident income in-
cludes two main parts: one is the total wage which con-
sidered as remuneration for labor, and the other is from
operating surplus which considered usually as capital
income. For a national economy, income also includes
the transfer payments (ignored here) from abroad. The
influences on consumption from wage W or surplus Z are
very different. Z’s main function is to provide the in-
vestment fund for expanded reproduction; W’s main
function is to provide consumption fund.
3) A large part of the government expenditure is a
short-term or annual decision variable, so financial defi-
cit, for a functional finance, may be a short-term deci-
sion-making. Generally, taxation as a long-term factor
cannot change greatly and frequently. The deficit deci-
sion-making which generally takes GDP as a benchmark
is a percentage of GDP and can not be too large.
4) As derived demand, investment is not only affected
by financial interest rate r, but also affected by surplus
level. Obviously, the higher the surplus level is, the big-
ger the power of enterprises investment is. In this paper
we will not discuss the balanced problem of money
market, but directly take r as a kind of exogenous con-
stant. We do not consider its change in the following
analysis, so it will be omitted in the investment function.
Although the bank deposit and loan interest rates may be
endogenous as market variables in microeconomic level,
the modern economic system provides a space for mone-
tary policy. The guidance or benchmark interest rate is
drawn up by the central department of financial man-
agement. The real interest rate fluctuates generally above
or below the guidance interest rate. From the macro-
scopic effect, the benchmark interest rate affects the level
of investment through control rules.
5) What affects the economic import and export level
is not only exchange rate, but also the economic level of
activity as a major contributing factor for import and
export in fact. This factor can be represented by demand
level. The changes of imports and exports induced by
aggregate supply can be embedded in
X
and
I
. This
paper takes exchange rate as an exogenous constant and
does not consider its change, so it will be omitted in the
following analysis.
Substituting the above demand behavior functions into
Formula (3), we will have


123
,
ds d
QQafWZPfZPfQ G
 (5)
where it is assumed that
11 11
22 3
3
0, 0,
0, 10
d
WZ
d
ZQ
ffWffZ
ffZf fQ
 
 
When the price level is given constant and let
s
d
QQ, Formula (3) coupled with various behavior
functions can be equivalent to the
I
Sformula in main-
stream economics.
2.3. Aggregate Supply (AS) Relations
The main development of contemporary mainstream
economics is about aggregate supply theory, and various
schools’ arguments are mainly on the aggregate supply
model. The traditional Keynes theory considers aggre-
gate supply completely elastic in short term and thinks an
economic equilibrium is mainly decided by aggregate
demand. Neoclassic economics assumes that aggregate
X. J. LIU
Copyright © 2011 SciRes. ME
205
supply is decided by the people’s expectation of price
level. If the expectation is rational, the expected price
will be consistent with the actual price. Thus, the aggre-
gate supply is equal to the natural rate of output and
completely inelastic and is not disturbed by random fac-
tors and macroeconomic policies. The new Keynesian
economics believes that the market is not always cleared
up because of the stickiness of wage and price, thus
e
PP. Therefore, the aggregate supply is elastic, but is
not completely elastic as Keynes defined. In the litera-
tures of mainstream economics, there are many ways to
define an aggregate supply function. In the following, we
adopt the way used by Blanchard [4].
2.3.1. AS Fun c ti on Derivation Ste p s of Blanchard
Let’s first show how Blanchard derived AS function.
1) Wage decision function. Assuming
,
e
wPFuz (6)
where w is a wage rate, e
P is an expected price level,
u is an unemployment rate. This relation is decided by
the supply and demand of labor together. For labors, if
w rises and e
P does not change, the supplied labor
will increase and then u will decrease. For enterprises,
the wages that they’re willing to pay when urises will
be reduced and the strength of wage bargaining of workers
will be weaken; thus0
u
F
Fu

.
Formula (6) reflects the behavior characteristics when
labor and capital bargain for employment contract. z
represents the other factors which affect wage besides u
and e
P, such as the unemployment insurance level, the
economic structure regulation and so on. Generally we
assume z changes in the same direction with w.
Formula (6) also means that what the labor and capital
think highly of when they negotiate a labor contract is
real wage(e
wP )rather than nominal wage (w).
2) Production function. Assuming a simple produc-
tion function:
YAN
where N is the employed labor. If measuring the total
output (National income) Y by a suitable unit which
make 1
A
, thus
YN (7)
then 11
NY
uLL
 . Here L is the total labor force.
3) Price decision rule. Assuming

1Pw
 (8)
where
is called price markup. Formula (8) may re-
gard as a cost-plus pricing model. The mainstream eco-
nomics does not consider intermediate inputs, so the only
cost is wage at present. Blanchard pointed out that If the
market is competitive completely, there are 0
,
Pw
. Therefore, it considers a state of incomplete
competition here.
4) Combination. Substitute Formula (6) into Formula
(8), we get
 
1,
e
PP Fuz
 (9)
5) Aggregate supply function. From the above formu-
las, we may obtain the relation ofPand
Y
:

11,
eY
PPF z
L

 


(10)
2.3.2. AS Function in Input-Output System
Similarly to Blanchard’s steps, we can establish the AS
relation under an input-output framework. From Formula
(4), we have 1
s
ss
PQ awQ lTZPQwhere l is the
labor input used to produce one unit product. Assuming
that all labors are homogeneous and the work condition
is the same, then l may be regarded as the labor num-
ber taken by one unit product and wis the remuneration
for one unit labor in certain time, so that
s
lNQ.
Suppose that taxation is a proportion t of added value
or GDP, then
11
s
TtwQlTZ
 (11)
Similarly to Blanchard’s price deciding model, we as-
sume that
s
Z
PQ
(12)
From Formulas (11) and (12), we have


111
s
ss
ttQ
TwQlPQ wlP
tt



Let 1
1
tt
t
, then
1
s
ss ss
PQawQltQwlPPQPQ

 
and then
1
Pawlt wlPPP



1
1
1
11
twl
Pat
 
Let

1
0
1
1
11
t
kat
 , then it gets
00
,0Pkwl k
(13)
Formula (13) is just the price deciding formula in an
input-output system. Then we apply Blanchard’s wage
deciding relation i.e. Formula (6), there is
X. J. LIU
Copyright © 2011 SciRes. ME
206

0,
e
P
klPFu z (14)
Introducing the production function s
QNl, then
11
s
Nl
uQ
LL
 
Therefore we may have the AS function as following,
01,
es
l
PklPF Qz
L




(15)
We will not consider about the change effect of z, so
that z will be omitted in the following context.
2.4. The Slope Signs of AD and AS Curves
For brief, we use the symbol

1e
u
wlulP F
 ,


0
1e
u
PuklPF

 .
2.4.1. AS Cur ve
From Formulas (14) and (15), we have
0
0
d
d
d1
or d
e
u
s
s
e
u
Pl
klPF L
Q
QL
PlklP F








(16)
where u
F
express the partial derivative of function F to
u1 the same below).
Because 00, 0
u
kF, so d0
d
s
Q
P, thus the AS
curve inclines up toward right in the
s
PQ or s
QP
space (see Figures 1-3 below).
2.4.2. AD Curve
From Formula (5), we have

12
1
12
2
3
dd dd
dd dd
d
d
d
ds
ZZ
W
d
Q
ff
f
QQ WZ
a
PPPPPP
ff
Q
fPP
 

dd
dd
s
WQ
PP
0
dd
dd (1)
s
s
s
e
u
ZQ Z
PQ Q
PP uklPF

 


From the above formulas, we may educe


12
1
3
12 12
2
dd
1dd
d
ds
ZZ
W
Q
ZZ
ff
f
QQ
fa P
PP PP
ffZff
P






(17)
AS
Q
AD
E
P
E
AD
Figure 1. The change of G.
AS
AD
(a)
Q
P
AD
ASQ
P
(b)
Figure 2. Practically possible AD and AS curve.
Because d0
d
s
Q
P, so it has dd
0, 0
dd
WZ
PP

, and
1123
0, 0,0,1
d
WZ ZQ
ffff .The symbol of
12 12ZZ
f
fZ ff is not too definite. Using some
data which close to reality and giving price changes in
usual range (for instance P < 0.2) to simulate, we
found d0
d
d
Q
P. This shows that the AD curve also in-
clines up toward right. This conclusion is obviously
1In this paper all of the differential coefficients is partial derivatives,
b
ut we use the ordinary derivative symbols for si
m
plification.
X. J. LIU
Copyright © 2011 SciRes. ME
207
AS
Q
AD
AD
E
AS
E
P
Figure 3. The change of a.
against the traditional theory. Let us see how this hap-
pens.
In mainstream economics, the general derivation
process of AD function starts with


d
d
YCYGIr (for a closed economy)
where d
Y is disposable income, r is the interest rate
decided by money market. In money market, d0
d
r
P.
Because

dd
dd
I
r
Ir
PP
,

d
d
dd
d
d
Y
CCY
PP
,and usually
d
d
YYT (T is tax) and G is considered constant, so

dd1
dd1
d
d
YI
PPCY
. Because

01
d
CY

,
0
Ir
,
so that d0
d
d
Y
P. In fact, there is a subjective mistake of
self-righteousness in the above process of the deduction2.
That is we consider d
d
YYT by mistake. Because
income roots in production process, while the production
process is decided by supply, so actually it should
be

dd
dd
s
d
CY
CY
PP
. Becaused0
d
s
Y
P, so d0
d
C
P. Thus
the symbol of ddd
ddd
d
YCI
PPP
 will be not definite. If we
assume that r is constant, and
I
is the increasing func-
tion of
Z
, obviously we have d0
d
d
Y
P. Such a conclu-
sion also shows the differences between macroeconom-
ics and microeconomics. Besides very few commodities
(such as Giffen commodity), the demand curves in mi-
croeconomics should incline down toward right (this is a
conclusion under given income). But AD curves in mac-
roeconomics generally should incline up toward right
and have the same orientation with AS curves. Looking
from the real economy observation, this point is intui-
tively correct. In macroeconomic level, the rise of price
level shows that economic tends to be prosperous, the
income will be higher and the demand will also increase.
The investment would increase when price level move
up. One reason for that mistake in traditional macroeco-
nomics is that it was forgotten that the income determi-
nation theory was derived out under aggregate supply
equal to aggregate demand when simple Keynesian in-
come determination theory was turned into AS-AD
analysis, i.e. it assumes

ss
YCY IG
in advance.
It must return to
ds
YCY IG
 when deducing
AD function. The total income used in LM model for
money market should be the Y when aggregate supply is
equal to aggregate demand. A proper macroeconomic
analysis framework should be the combination of AD
formula, AS formula and the money market formula, and
an equilibrium state can be solved from these three for-
mulas, but it is not to add the money market (or general
financial market) formula to income determination
model at first. The foreign exchange and the international
money market should be added in for an open economy.
In this paper, we assume the two markets are exogenous,
i.e. r and e are given.
2.5. The Perturbations of AD and AS
With the AS and AD functions given above, we will
make some comparative static analysis based on the per-
turbations caused by several parameters in the following.
Being equilibrium, Q
d = Qs, therefore the above sys-
tem satisfies the following equilibrium formula group:
 
12
3
,, ,
fWZf rZ
QQafQe G
PP
 (5)
0
Pklw
(13)

,
e
wPFuz (6)
1uQlL
Z
PQ
(12)
WQlw
2.5.1. Changes of Government Consumption (G)
Differentiating the above formulas with respect to G, we
have
2Samuelson reminds people not to commit three mistakes in the intro-
duction of his ‘Economics’ [5]. They are post hoc fallacy, fallacy o
f
composition and subjectivity. He said let us alert the assuming condi-
tion our subjective interest didn’t express definitely.
X. J. LIU
Copyright © 2011 SciRes. ME
208

2
0
0
0
ddd
ddd
1d
d
ee
uu
e
u
PlFkl PF
PQQ
kl
GLGLG
uklPFQ
QG

 



(18)

112
1
0
12
3
2
d1
d
1
WZZ
e
u
Q
fff
Qa
GPP
uklPF
ff f
Q
P

 

(19)
Let

112
0
12
3
2
1
1
WZZ
e
u
Q
fff
Ba
PP
uklPF
ff
f
Q
P
 

,
then 1
d
d
QB
G
.
Because the change of G has no influence on AS and it
makes AD increase, thus it should also make the equilib-
rium gross output increase, namely there should be
d0
d
Q
G. Because 0
u
F, according to Formula (18),
there should be d0
d
P
G. Therefore, the expansion of
government consumption can make the equilibrium price
and equilibrium output increase simultaneously. Thus,
according to Formula (19), there should be

112
0
12
3
2
1
1
0
WZZ
e
u
Q
fff
Ba
PP
uklPF
ff f
Q
P

 

(20)
From the above results, we may also have
dd
dd
s
d
QQ
PP
at the equilibrium point, otherwise d0
d
Q
G
and d0
d
P
G which is unreasonable. However, when an
economy is close to potential production, there will be
d0
d
s
Q
Pcertainly and may have dd
dd
s
d
QQ
PP
(See
Figure 1), thus dd
0and0
dd
QP
GG

may appear. We
need introduce the money market and the foreign ex-
change market to explain or solve this problem. In fact a
serious bubble economy appears then and the income no
longer originates from production but from speculation
or the unusual inflow of foreign exchange. On the other
hand, because the economy is close to potential produc-
tion, there should also be d0
d
d
Q
P. Therefore, the pos-
sible AS and AD curves may be as that in Figure 2.
2.5.2. Changes of Direct Consumption Coefficient (a)
Direct consumption coefficient is the characteristic pa-
rameter of input-output analysis. By intuition, the in-
crease of a will enhance the marginal cost, thus the price
will rise, and equilibrium output reduces possibly. This
conclusion is generally correct in microeconomics. We
will observe the microeconomic effects with AS-AD
model in an input-output system in the following.
Differentiating the equilibrium formulas with respect
to a, we have:


0
0
0
1
d
dd
dd
1d
11 d
u
k
Pw
lw k
Eaa a
e
ulkPF
PQ
atQ a






(21)

1
dd d
dd d
u
u
e
uPF
wlQ Q
e
PF
aLaQa



 (22)






12 12
1
2
11
d
d
11 11
ZZ
Q
a
ZffPf f
BQ
Pa tPat


  
(23)
Obviously it is d0
d
Q
a. Substituting Formula (23) into
Formula (21), we might judge immediately that d0
d
P
a.
These results are opposite to the microeconomic situation.
This may be explained as following. Though the increase
of direct consumption coefficient increases the cost, it
also enhances the demand simultaneously. From the
Formula about d
d
w
a, we know that the wage rate will
also enhance, thus the consumer demand will also en-
hance and the investment demand will also increase.
Therefore, the final results are that equilibrium output
may also enhance simultaneously when the price rises.
According to Formulas (16) and (17), the increase of a
will make the slope of AS curve decrease and simulta-
neously the slope of AD curve may increase or decrease
(See Figure 3). If the reduction of AD curve slope is
within a certain range, the equilibrium output still en-
hances. If the slope of AD curve increases, the equilib-
rium gross output must increase.
2.5.3. Changes of Labor Productivity
If labor productivity increases (i.e. l reduces), an enter-
X. J. LIU
Copyright © 2011 SciRes. ME
209
prise’s marginal cost will reduce, thus it will cause the
reduction of market price of that product and the increase
of equilibrium output. Now let us see the macroeconomic
situation.
Differentiating the equilibrium formulas with respect
to l, we have
 
dd
dd
11
u
uu
wlQQ
e
PF
lLlL
ee
uPF uPF
dQ
Qdl l





 
(24)

0
00
dd
dd
d
11
d
e
u
Pw
kwl
ll
lQ
kwukPFQl








(25)


1
00
12 12
1
0
2
d1
d
e
u
ZZ
W
QBkwukPF
l
ffZff
fQk
PP







(26)
According to Formula (26), we can judge that d0
d
P
l
from Formula (25) so long as d0
d
Q
l. This means that
the increase of productivity (i.e. decrease of l) will re-
duce output and price level. Why does this happen? The
reason is that although the enhancement of labor produc-
tivity increases the aggregate supply (this can be proven
from Formula (15)), simultaneously it also causes the
augment of unemployment rate3 and the decrease of
wage rate further. These two kinds of changes will obvi-
ously reduce the income, thus it causes the consumption
demand reduced and makes the equilibrium aggregate
output decrease finally. This conclusion seems to contra-
dict with real situation. This paradox’s solution relies on
the changes of international market. The increase of la-
bor productivity would enhance export competitiveness
and increase export demand obviously, which would
influence investment demand in turn, and then equilib-
rium output was finally augmented. However, the do-
mestic consumption demand would not increase neces-
sarily at the same time4. In the 1990s of the productivity
increase and economic expansion caused by new econ-
omy, for the Expenditure structure of US GDP from
1990 to 2000, the net import proportion increased 2.95
percentage points, the total capital formation increased
1.53 percentage points, the government consumption
dropped 2.93 percentage points, and the household con-
sumption reduced 1.56 percentage points. Looking into
the structure of income distribution, the Compensation of
Employees reduced 2.04 percentage points but the oper-
ating surplus (including depreciation of fixed assets) in-
creased 3.8 percentage points (See Tables 1 and 2 for
details). Therefore, how to transform the benefits of
productivity increase into the income of ordinary resi-
dents and the increase of consumption level needs the
support of government incomes policies (including tax
policy), otherwise an economy would go to serious im-
balance someday.
2.5.4. Changes of Operating Surplus Rate (μ)
The enterprises’ strength would enhance in the labor
market when operating surplus increase, which means
they can extort more surpluses therefore the aggregate
supply would increase. The changes of aggregate de-
mand rely on whether the increase of income caused by
supply increase may compensate the income reduction
caused by the reduction of labour reward share.
Differentiating the equilibrium formulas with respect
to μ, we will deduce

1
dd d
dd d
e
eu
uPF
wlQ Q
PF LQ


 




0
00
02
0
d
dd d
ddd d
1d
d
e
u
k
Pw
kwllkw
ukPF Q
lwk
Q






 






01 2
1
12 0
d1
dZZ
kf f
QBffQkP

From the above, it has d0
d
Q
, d0
d
P
. Therefore,
the increase of surplus proportion causes the equilibrium
output to expand.
3. The Explanation of Stagflation Issues
The mainstream economics generally attributes the stag-
3In Formula (15), it is 01,
l
es
PklPFQz
L




=

0,
e
klPFuz. Let P
= constant, then doing differentiation, thus

d
,0
d
u
u
FuzlF l
. Be-
cause 0,
u
Fso that d0
d
u
l.
4Jorgenson once found, ‘comparing the contribution of intermediate
input with other source of output growth demonstrates that this input is
by far the most significant source of growth. The contribution of inter-
mediate input (for output increase) exceeds productivity growth and the
contribution of capital and labor inputs. If we focus attention on the
contribution of capital and labor inputs alone, excluding intermediate
input from consideration, these two inputs are a more important source
of growth than changes in productivity’ [6].
X. J. LIU
Copyright © 2011 SciRes. ME
210
Table 1. Total expenditure structure of American economy in the 1990s (%).
1980 1990 1995 1997 1998 1999 2000
Government consumption 15.23 15.01 13.60 13.00 12.63 12.25 12.08
Household consumption 57.16 59.56 59.58 58.74 58.58 58.40 58.00
Total capital formation 18.06 15.69 15.97 17.00 17.59 17.77 17.22
Total exports 9.09 8.63 9.84 10.32 9.70 9.23 9.43
Net exports 9.55 9.75 10.85 11.25 11.20 11.59 12.70
Note: This table was calculated with the current price data. Data resource: http://www.stats.gov.cn/tjsj/qtsj/gjsj/.
Table 2. Total income distribution structure of American economy in the 1990s (%).
1980 1990 1991 1992 1993 1994 1995 1996 2000
Indirect tax minus subsidy 7.42 7.45 7.86 7.85 7.92 8.01 7.99 7.73 7.33
Depreciation of fixed assets 13.48 10.84 11.00 10.95 10.65 10.84 10.68 10.47 0.00
Compensation of Employees 61.04 60.35 60.53 60.45 60.17 59.80 60.20 60.05 58.31
Operating surplus 17.58 21.05 20.45 19.99 20.33 20.85 21.14 22.57 35.69
Statistical errors 0.52 0.31 0.18 0.75 0.93 0.52 –0.03 –0.81 –1.33
Note: The operating surplus in 2000 includes the depreciation of fixed assets. Data resource: http://www.stats.gov.cn/tjsj/qtsj/gjsj/.
flation in the 1970s to the oil crisis, and that sort of infla-
tion is called cost-push inflation. Regarding it, Blanchard
pointed out that what we face is that Oil price was nei-
ther in AS relations nor in AD relations because we as-
sumed that productions use labor force only. One way to
deal with this issue is to relax this assumption and to let
production simultaneously use labour and other invest-
ment (including energy), then the relations of price with
wage and oil price can be inferred. However, Blanchard
took a shortcut and used the parameter
to reflect the
rise of the oil prices. His principle was that the rise of oil
price would increase the production cost for given wage
level, which forces enterprises to promote their price
level [4].
We must clarify the vague cognition in Blanchard’s
view at first. Mainstream economics does not assume
that production only uses labor, but its rationales are es-
tablished on the net income principle of “the Trinity”.
From microeconomics to macroeconomics, they have
throughout thought that the essential factors of economic
production are labor, capital and land, and sometimes
Entrepreneurship was added, which is a fallacy actually.
No other than “the Trinity”, the output in macroeconom-
ics can be interpreted as national income or GDP. There-
fore, it is impossible to explain the phenomenon of stag-
flation caused by the rapid rise of oil prices in the interior
of mainstream economics, saying nothing of explaining
the further deep system causes of an economic crisis.
For simplicity, we assume that the oil that an eco-
nomic system uses is completely imported and its do-
mestic output is zero, thus we have an input-output table
as following form.
0
00 00
0
0
0
X
CGI NX
RR
W
T
Z








where
R
represents the oil used as intermediate input.
We assume that oil has no other final uses; moreover, the
oil import has no immediate influence to the net export
of other products. Let
s
bRQ represent the direct
consumption coefficient to oil and the oil price is 0
P,
then it has
0
s
s
RQb
PXP RWTZPQ

Obviously R is decided by
s
Q.
After adding up on the oil, the equilibrium formula set
is as follows
X. J. LIU
Copyright © 2011 SciRes. ME
211




12
3
0
0
1
,, ,
11
,
1
e
fWZf rZ
QQafQe G
PP
Pb
Pklw at
wPFuz
uQlL
ZPQ
WQlw




When being equilibrium, Qd = Qs = Q, then it can be
worked out that

00 0
1
dd d
dd d
e
u
e
u
uPF
wlQ Q
PF
PLPQP

 

 (27)


0
001
1
dd
dd11
e
u
uklPF
PQb
PQPat
   (28)


12 12
1
2
01
d
d11
ZZ
ffZff
Qb
B
PatP

  (29)
According to Formula (29), if the rise of oil price (0
P)
causes the decline of output (Q), it means

12 12ZZ
f
fZ ff < 0. But the total price level
does not rise necessarily according to Formula (28). The
influence factors are quite a lot according to the related
parameter group. At the end of the 1960s in America, the
unemployment rate (u) is relatively low, and the profit
rate (represented by μ) is relatively high, and the labor
productivity(1 /l) and the price expectation (Pe) are also
quite high, therefore, the negative factors may be even
greater and the price level is probable to decline at the
beginning. However, because the output and the profit
rate declined rapidly, the unemployment rate rises rap-
idly, and then the labor productivity may decline. When
the negative factors rapidly reduce and the price move-
ment transferred into inflation very quickly, thus the
stagflation happened. The occurrence of stagflation
should be the result of the enterprises transferring cost. If
big enterprises willed to reduce profit greatly, inflation
might be evaded.
4. Conclusions
Under the single-sector input-output analysis frame, we
established an AS-AD model with total output as the
main variable, and revealed the right upward characteris-
tics of AD curve5. The analysis to some parameters’
changes also indicated that the macro- and micro-effects
of an economy are remarkable different. They are often
opposite6 . For example, the rise of direct consumption
coefficient does not reduce but increases the equilibrium
output. It shows that macro economy is not the simple
superposition of micro individuals. In the analysis of the
labor productivity’s changes, we find that the positive
role of enhancing labor productivity on equilibrium out-
put lies on some conditions, such as good trade condi-
tions or proper income policy, etc. From the analysis of
stagflation, it is found that the oil crisis might not cause
stagflation necessarily, but cause the usual decline.
5. References
[1] B. Los, “Endogenous Growth and Structured Change in a
Dynamic Input-Output Model,” Economic Systems Re-
search, Vol. 13, No. 1, 2001, pp. 3-34.
doi:10.1080/09535310120026229
[2] W. W. Leontief, “Quantitative Input and Output Rela-
tions in the Economic System of the United States,” Re-
view of Economics and Statistics, Vol. 18, No. 3, 1936,
pp. 105-125. doi:10.2307/1927837
[3] J. M. Keynes, “The General Theory of Employment,
Interest and Money,” Macmillan Cambridge University
Press, London, 1936.
[4] O. Blanchard, “Macroeconomics,” Tsinghua University
Press, Beijing, 2001, pp. 126-128.
[5] P. A. Samuelson and W. D. Nordhaus, “Economics,”
China Development Press, Beijing, 1992, pp. 11-15
[6] D. W. Jorgenson, “Productivity, Volume 1: Postwar U.S.
Economic Growth,” MIT Press, Cambridge, 1995, pp. 2-7.
5In Oliver Blanchard’s book, he gave a footnote, “A better name would be ‘the goods and financial market equilibrium relation’. But because this is a
long name, and because the relation looks graphically like a demand curve (that is, a negative relation between output and the price), it has become
traditional to call it the ‘aggregate output demand relation’. Be aware, however, that the aggregate supply and aggregate demand relations are very
different from regular supply and demand curves”. [4] For the reasons what Professor Blanchard said and we said above, I suggest that it is better to
give up the nameAS-AD relation (or model) instead of four market (commodity, money or finance, labor and foreign exchange ) equilibrium rela-
tions in usual textbooks.
6In fact, we have already known that the preconditions of macroeconomics and microeconomics are different. Income is fixed in microeconomics and
variable in macroeconomics; the consumers are competitive in microeconomics and monopolistic as a collective in macroeconomics. Like the supply
function of monopoly in microeconomics, AD and AS are not independent in macroeconomics.
X. J. LIU
Copyright © 2011 SciRes. ME
212
Appendix: A Case of New AS-AD Model
If we assume that:
1) The household consumption is proportional to the
total wage, i.e. 1
CfWP;
2) The investment is proportional to the gross
operational surplus which equals to depreciation
plus net operational surplus, i.e. 2
I
fZ P;
3) The net export is proportional to the gross
output or aggregate demand, i.e. 3
d
NXf Q;
4) The wage rate is inversely proportional to
unemployment rate, i.e. k
wu
;
then we can have the following AD and AS functions:

1
2
3
1/
1
flk
ssss
P
dQaQQl LfQG
Qf

0
skk
L
QlP

Based on the input-output table of China 2007 and the
data from The Yearbook 2010, we get the values of the
parameters as following:
Names of
parameters values Names of
parameters values
L (10,000 capita)78,645 a 0.675104232
f1 0.877373806 l 9.40211E-06
f2 0.944173563μ 0.143465228
f3 0.026281005 G(10,000 yuan) 370,513,349
k0 7.440972761
k 300.7959285
With these values of the parameters, we plot the AS
and AD figures as following:
The Input-Output Table of China 2007 (Unit: 10,000 yuan)
intermediate use
household
consumption
government
consumption
capital
formation export import others gross
output
intermediate input 5,528,151,509 965526184.3 351909186.2 1,109,194,214955,409,910740205546.8 18604162.77 81,88,589,620
payment for labor 1,100,473,000
net production tax 385,187,233
depreciation 372,555,322
net operational
surplus 802,222,556
gross input 8,188,589,620