Modern Economy, 2010, 1, 112-117
doi:10.4236/me.2010.12011 Published Online August 2010 (http://www. SciRP.org/journal/me)
A Modified Consumer Price Index*
Gonglin Yuan, Xiangrong Li
College of Mathematics and Information Science, Guangxi University, Nanning, China
Received June 2, 2010; revised July 3, 2010; accepted July 8, 2010
Abstract
It is well known that the Consumer Price Index (CPI), as a Laspeyres-type index, attempts to measure the
average change in the prices paid by urban consumers for a fixed market of goods and services, and new
samples for most item categories are routinely introduced over time to keep the CPI sample representative of
consumer spending patterns. The CPI normally overstates the true rate of increase of the cost of living. In
this paper, our main objective is to propose a new measurement in the CPI which combines with the Gross
Domestic Product (GDP). This new method will make the bias effectively decreased.
Keywords: Consumer Price Index, Gross Domestic Product, Fixed Market Basket
1. Introduction
The CPI is defined by




bt
bt
QP
QP
CPI
1
(1)
where t
P is the price of an item in period t, 1t P is the
price in a base period b, and b Q is an index of the qua-
ntity of an item in a base period b. The CPI can provide
an approximation to a cost-of-living index (CLI), measu-
ring the average change in the prices paid by urban cons-
umers for a fixed market basket of goods and services has
many limitations when interpreted relative to a true CLI.
For example, consumers shift spending patterns in respo-
nse to changes in relative prices, items and outlets avail-
able in the original or base period disappear, and new ite-
ms and outlets enter the marketplace. To alleviate some
of these problems, the CPI uses a modified Laspeyres ap-
proach, which allows for product substitution and intro-
duction of new samples of outlets and items [1].
It is important to have an assessment of the magnitude
of the bias in the CPI. First, the CPI is the most widely
followed measure of inflation. Users of all types, includ-
ing members of the general public, policy makers, and
participants in financial markets, should have the best in-
formation available concerning the size of the bias. Seco-
nd, knowledge about the sources and magnitude of the
bias could be important in guiding efforts to improve the
index. Among other things, this type of knowledge is es-
sential for judging the likely costs and benefits of inv-
esting additional resources in the index. Third, the CPI
has a substantial effect on the Federal budget. This link
between the CPI and the Federal budget has generated
considerable political interest in the magnitude of the
bias in the CPI.
Furthermore, price evaluations may be biased by per-
ceptions of price unfairness [2], low purchase frequency
and steep price changes of particular goods [3]. Although
Kemp [4] to some extent dealt with general costs, the
cognitive processes described typically comprise indi-
vidual reactions to price changes of isolated goods and
services, not to reactions of the general public to prices
changes across consumption categories. At the aggregate
level, the divergence between perceived and actual price
changes cannot be fully explained from cognitive process.
Economic data usually capture price changes by using
price indices, which essentially reflect changes in aggre-
gated prices, i.e., weighted averages of a large number of
price changes in different item categories. Hence, infla-
tion perceptions may deviate from price indices due to
differences between perception processes and statistical
procedures in constructing the price indices.
Kemp [4] mentions the possibility that experience with
purchases, i.e., for frequently purchased items such as
stamps, butter and telephone bills, tends to strengthen
these effects. Experience may add to the availability bias
[5], possibly resulting in greater weight of high-freque-
ncy purchases in perceived inflation judgments. Kemp [6]
found almost correctly perceived inflation for the previ-
ous year but again under-estimated perceived inflation
for the previous 15 years. Brachinger [7] assumes as-
This work is supported by China NSF grants 10761001 and the Scie-
ntific Research
,
Foundation of Guan
g
xi Universit
y
(
Grant No. X081082
)
.

G. L. YUAN ET AL.
113
ymmetric inflation perceptions for prices increases ver-
sus price decreases. Due to the asymmetric value func-
tion in the prospect theory (Tversky & Kahneman, 1991),
price increases should influence perceptions more than
price decreases. Hence, items associated with large price
increases should influence general perceived inflation
more heavily than items associated with minor prices
increases or price decreases. This expectation may be
qualified by distinguishing between absolute and relative
price changes. In contrast, a one cent increase of gasoline
prices may be evaluated as quite low (see [8]). Hoffm-
ann, Leifer, and Lorenz [9] seem to favor the role of
relative price changes in consumer price perception.
The GDP was introduced as a monetary measure of wa-
rtime production capacity during the World War II. To-
day, it is widely used by policymakers, economists, and
the media as the primary scorecard of a nation’s econo-
mic health and welfare. However, GDP has some unav-
oidable deficiencies as a measure of economic perform-
ance (see [10-12]), and is incapable of measuring peo-
ples’ well-being. The major problem is that GDP makes
no distinction between economic transactions that add to
welfare and those that diminish it [13]. It includes all
expenditures, assuming that every monetary transaction
adds to peoples’ welfare. Real GDP is often used as a
proxy of a country’s real income, even though official
statisticians warn against such a practice [14]. Thus,
Prescott [15], who singles out Switzerland for its poor
economic performance over the past three decades, fo-
cuses exclusively on real GDP. Yet, unlike a technologi-
cal progress, the beneficial effect of an improvement in
the terms of trade is not captured by real GDP, which
focuses on production per se. In fact, if real GDP is
measured by Laspeyres quantity index, as it is still the
case in most countries, an improvement in the terms of
trade will actually lead to a fall in real GDP [16]. Based
on the nominal GDP(NGDP) and real GDP (RGDP), an
index GDP deflator (GDPD) is defined by
RGDP
NGDP
GDPD  (2)
which reflects the changes of all items in economics.
Usually the GDPD tends to underestimate the inflation
for consumer price [17].
Motivated by the above observations, we propose a
new index which combines the CPI and the GDP to test
inflation. This index will make the bias decreased effec-
tively in the CPI. In the next section, motivation and
method are stated. The data results are reported in Sec-
tion 3. One conclusion is stated in the last section.
2. Motivation and Method
Many proposals have been forwarded to alleviate the bi-
as caused by the rotation of new item and outlet samples
in the CPI. In the interim, there are three ways that have
been systematically investigated in which the current bias
in the CPI sample rotation process may be alleviated [18]:
1) using geometric means to calculate basic item-area pr-
ice relatives; 2) setting base period prices using pre-link
month “initiation” prices; 3) pricing both the old and new
samples for a period of time before introducing the new
sample into the CPI.
It is well known that the CPI is one of the most im-
portant indexes of the inflation. Normally the CPI overes-
timates the inflation [17]. Many authors study this prob-
lem to decrease the CPI (see [18,19] etc.). From the defi-
nition of CPI (1), it is easy to see that the CPI only refers
to the consumer items but other items. When we co-
nsider the CPI of some items, other items are omitted. In
fact, this CPI will be influenced by other items. Then a
reasonable idea is to consider the items’ percent of the
total property, i.e., the GDP should be considered. Mor-
eover, the authors [18,20-23] use geometric means to
calculate basic item-area price relatives in CPI and get
better results. Motivated by their ideas and the above di-
scussions, we present the modified CPI formula as fol-
lows
,
GDP
QP
GDP
QP
MCPI
t
t
N
b
bt
N
t
bt




1
(3)
where t
GDP is the GDP in period t, b GDP is the GDP
in a base period b, and Nt is the number of all items, res-
pectively. In practice, it is not difficult to compute (or
estimate) the quantities Nt By (3), we have
.CPI
GDP
GDP
GDP
QP
GDP
QP
MCPI t
t
t
N
t
b
N
b
bt
N
t
bt
 

1
(4)
In this paper we will use the index MCPI in (4) instead
of CPI in (1). In the next section, we report the practical
data to compare the given Formula (4) with the normal
CPI Formula (1).
3. Data Results
Since reform and open policy, China has one of the high-
est rates of economic growth in the world, especially for
GDP. In this section, we report the detail data to test our
given method including GDP, CPI, all items of CPI since
the year 1990 in China. We list them as the following
tables.
The data of the Table 1-2 is from National Bureau of
Statistics of China (2008) or can be found at the Home-
page:

G. L. YUAN ET AL.
114
Table 1. The data of GDP, Per Capita GDP, CPI, Urban Household CPI, and Rural Household CPI.
Year GDP(100 million Yuan)Per Capita GDP (Yuan)CPI(preceding year=100) Urban Household CPI Rural Household CPI
1990 18667.8223761059 1644 103.1 101.3 104.5
1991 21781.4994107882 1892.8 103.4 105.1 102.3
1992 26923.4764511214 2311.1 106.4 108.6 104.7
1993 35333.9247145462 2998.4 114.7 116.1 113.7
1994 48197.8564447092 4044 124.1 125 123.4
1995 60793.7292113314 5045.7 117.1 116.8 117.5
1996 71176.5916539871 5845.9 108.3 108.8 107.9
1997 78973.0349964914 6420.2 102.8 103.1 102.5
1998 84402.279768922 6796 99.2 99.4 99
1999 89677.0547509045 7158.5 98.6 98.7 98.5
2000 99214.5543084772 7857.7 100.4 100.8 98.5
2001 109655.170558159 8621.7 100.7 100.7 100.8
2002 120332.689274252 9398.1 99.2 99 99.6
2003 135822.756149557 10542 101.2 100.9 101.6
2004 159878.33791739 12335.6 103.9 103.3 104.8
2005 183217.4 14053 101.8 101.6 102.2
2006 211923.5 16165 101.5 101.5 101.5
2007 249529.9 18934 104.8 104.5 105.4
Table 2. Consumer Price Indices by Category (2007) (preceding year = 100).
Item National IndiceItem National Indices
Consumer Price Index 104.8 Health Care and Personal Articles 102.1
Food 112.3 Health Care 102.1
Grain 106.3 Medical Instrument and Articles 98.2
Rice 105.4 Traditional Chinese Medicine 107.9
Flour 107.3 Western Medicine 99.1
Starches and Tubers 106.2 Health Care Appliances and Articles 101.1
Beans and Bean Products 108.0 Health Care Services 102.2
Oil or Fat 126.7 Personal Articles and Services 102.1
Meal, Poultry and Processed Products 131.7 Cosmetics 100.1
Eggs 121.8 Sanitation Articles 100.3
Aquatic Products 105.1 Personal Ornaments 104.5
Vegetables 107.9 Personal Services 103.1
Fresh Vegetables 107.3 Transportation and Communication 99.1
Flavoring 104.1 Transportation 100.8
Carbohydrate 101.6 Transportation Facility 97.7
Tea and Beverages 101.5 Fuels and Parts 103.5
Tea 103.3 Fees for Vehicles Use and Maintenance 102.4
Beverages 100.7 In city Traffic Fare 101.3
Dried and Fresh Melons and Fruits 102.2 Intercity Traffic Fare 103.0
Fresh Fruits 100.1 Communication 97.1
Cake, Biscuit and Bread 103.6 Communication Facility 81.8
Milk and Its Products 102.7 Communication Service 100.6
Dining Out 107.3 Recreation, Education and Culture Articles 99.0
Other Foods and Manufacturing Services 104.2 Durable Consumer Goods for Cultural 93.1
Tobacco, Liquor and Articles 101.7 and Recreational Use and Services 99.6
Tobacco 100.8 Education 99.1
Liquor 103.5 Teaching Materials and Reference Books 99.6
Articles for Smoking and Drinking 100.1 Tuition and Child Care 101.0
Clothing 99.4 Cultural and Recreational Articles 99.5
Garments 99.4 Cultural Articles 100.7
Clothing Material 101.6 Newspapers and Magazines 102.7
Footgear and Hats 99.0 Expenditure on Culture and Recreation 102.3
Clothing Manufacturing Services 102.3 Touring and Outing 104.5
Household Facilities, Articles and Services 101.9 Residence 105.1
Durable Consumer Goods 101.6 Building and Building Decoration Materials 104.2
Furniture 101.9 Renting 107.0
Interior Decorations 100.3 Private Housing 103.0
Bed Articles 99.4
Daily Use Household Articles 101.7
Household Services and Maintenance and Renovation 107.2

G. L. YUAN ET AL.
115
http://www.sei.gov.cn/hgjj/yearbook/2008/indexeh.htm.
In order to show these data of Table 1, we give the hi-
stogram of GDP and diagram of curves of CPI in Figure
1 and Figure 2, respectively. From Table 1 and Figure 2,
we can see that the growth rate of GDP is about 10%
every year. This growth rate is interesting and shows that
the economy of China is healthy. From Table 1 and Fig-
ure 2, it is easy to observe that the CPI from 1993-1995
are the highest in these years, and we can conclude that
China was facing the inflation except for 1998, 1999, and
2002 years. It is not difficult to see that the urban house-
hold CPI was larger than the rural household CPI from
1990 to 2000. However, the rural household CPI sur-
passed the urban household CPI from 2001 year to 2007
year, which shows that the inflation rate of the rural
household was lager than the inflation rate of the urban
household in this period. This case also shows that the
living level of the rural household is becoming better in
some situation and the speed is lager than the urban
household does. Overall China is in the situation of infla-
tion from these data. From 2005 year, the inflation is
arising.
0
50000
100000
150000
200000
250000
300000
1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007
GDP(100 million
Yuan)
Figure 1. Sources of data: various years of the China Statistical Yearbook and China Data online (2008 year).
Figure 2. The data of GDP, Per Capita GDP, CPI, Urban Household CPI, and Rural Household CPI (preceding year = 100).
Sources of data: Various years of the China Statistical Yearbook and China Data Online (2008).

G. L. YUAN ET AL.
116
In the following, we will compute the inflation rate by
the normal CPI Formula (1) and the modified Formula (4)
according to the data in Table 1, respectively. From Ta-
ble 2, it is easy to compute the number of all items is
about sixty. Since it is this category of 2007 year, the nu-
mber of the category may be less than sixty before 2007
year. So we set 50
t
N by (4) in this paper. The nu-
merical results of formulas (1) and (4) are listed in Table
3 and Figure 3.
Michael, Ellen, Robert, Zvi, and Dale (1995) conclude
that the CPI overestimates the inflation rate 0.8 ~ 1.6
percentage points, and the “best estimation” is $1.1$ per-
centage points (see [17]). Then many modified CPI met-
thods are presented (see [19]), but the CPI still overesti-
mates the inflation. Table 3 provides the inflation rates of
these two indices. Before 1998, relatively high inflation
rates were observed, and the CPI overestimated the MC-
PI from 0.19 to 0.61 percentage points. In this period, the
inflation is serious. In 1998, 1999, and 2002, when the
deflationary pressure became stronger and the inflation
rates became negative, the CPI understated the MCPI by
–0.13, –0.12, and –0.19 percentage points, respectively.
Since 2000, the inflation rates are positive except for
2002, and the CPI overestimated the MCPI from 0.2 to
0.33 percentage points. Overall, China is facing the
pressure of inflation.
Table 3. Inflation rates and substitution bias.
Year CPI Inflation (%) MCPI Inflation (%) Bias (%)
1990
1991
1992
1993
1993
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
3.1
3.4
6.4
14.7
24.1
17.1
8.3
2.8
–0.8
–1.4
0.4
0.7
–0.8
1.2
3.9
1.8
1.5
4.8
2.91
3.1
5.98
14.16
23.49
16.64
7.99
2.6
–0.67
–1.28
0.2
0.5
–0.61
0.96
3.57
1.53
1.21
4.47
0.19
0.3
0.42
0.54
0.61
0.46
0.31
0.2
–0.13
–0.12
0.2
0.2
–0.19
0.24
0.33
0.27
0.29
0.33
Figure 3. The CPI and the MCPI denote the consumer price index and the modified consumer price index, respectively.

G. L. YUAN ET AL.
117
Figure 3 presents the CPI and the MCPI. The higher
inflation rate of the CPI than the MCPI is evident. From
the results of Table 3 and Figure 3, it is not difficult to
see that the modified CPI method can make the bias de-
crease in certain extent.
4. Conclusions
In this paper, we only propose a modified CPI formula
which combing with the GDP. This modified CPI form-
ula can make the normal CPI decrease in certain extent.
From the test results, we can see that this formula is in-
teresting in some cases. Based on the model of this paper,
we can get the following conclusions and extensions.
1) According to the data of National Bureau of Statis-
tics of China (2008), it is not difficult to see that China is
facing the pressure of inflation now although the Chinese
government has drew up related policy.
2) The real GDP should be considered in this modified
formula. The use of real GDP maybe make this method
is more loser to the real inflation. We will also be very
interested in researching conducted by other statistical
agencies in this area.
3) The method of the CLI estimated should be studied,
moreover the accordingly method is measured with the
CPI and the MCPI.
5. References
[1] A. M. Aizcorbe and P. Jackman, “The Commodity Sub-
stitution Effect in CPI Data, 1982-1991,” Monthly Labor
Review, Vol. 116, No. 12, December 1993, pp. 25-33.
[2] L. E. Bolton, L. Warlop and J. W. Alba, “Consumer Perc-
eptions of Price (Un)Fairness,” Journal of consumer Re-
search, Vol. 29, No. 4, 2003, pp. 474-491.
[3] E. D’Elia, “Actual and Perceived Inflation,” Working
paper, ISAE and Statistical Office of the municipality of
Rome, 2005.
[4] S. Kemp, “Estimation of Past Prices,” Journal of Eco-
nomic Psychology, Vol. 8, No. 2, 1987, pp. 181-189.
[5] A. Tversky and D. Kahneman, “Judgment under Uncer-
tainty: Heuristics and Biases,” Science, Vol. 185, No.
4157, 1974, pp. 1124-1131.
[6] S. Kemp, “Remembering and Dating Past Prices,” Jour-
nal of Economic Psychology, Vol. 12, No. 3, 1991, pp.
431-445.
[7] H. W. Brachinger, “Measuring Perceived Inflation: A
New Prospect Theory Based Index,” Journal of Eco-
nomic Psychology, Vol. 29, No. 4, 2008, pp. 433-457.
[8] R. Ranyard, F. Del Missier, N. Bonini, D. Duxbury and B.
Summers, “Perceptions and Expectations of Price Chan-
ges and Inflation: A Review and Conceptual Frame
work,” Journal of Economic Psychology, Vol. 29, No. 4,
2008, pp. 378-400.
[9] J. Hoffmann, H. A. Leifer and A. Lorenz, “Index of Per-
ceived Inflation or EU Consumer surveys?” Intereco-
nomics, Vol. 41, No. 3, 2006, pp. 142-250.
[10] C. Hamilton, “The Genuine Progress Indicator Methodo-
logical Developments and Results from Australia,” Eco-
logical Economics, Vol. 30, No. 1, 1999, pp. 13-28.
[11] P. A. Lawn, “A theoretical foundation to support the In-
dex of Sustainable Economic Welfare (ISEW), Genuine
Progress Indicator (GPI), and other Related Indexes,”
Ecological Economics, Vol. 44, No. 1, 2003, pp. 105-
118.
[12] Z. G. Wen, K. W. Zhang and B. Du, “Defedts of the GDP
Accounts System and the Amendatory Approaches-Meth-
Odology and Case Study (in Chinese, with English Ab-
stract),” Journal of China University of Geosciences (So-
cial Sciences Edition), Vol. 14, No. 13, 2004, pp. 43-46.
[13] C. Covv, T. Halstead and J. Rowe, “If the GDP is Up,
why is America Down?” The Atlantic Monthly, Vol. 276,
No. 4, October 1995, pp. 59-78.
[14] D. W. Caves, L. R. Christensen and W. E. Diewert, “The
Economic Theory of Index Numbers and the Measure-
ment of Input, Output, and Productivity,” Econometrical,
Vol. 50, No. 6, 1982, pp. 1393-1414.
[15] E. C. Prescott, “Prosperity and Depression,” American
Economic Review, papers and Proceedings, Vol. 92, No.
2, 2002, pp. 1-15.
[16] K. Ulrich, “Real GDP, Real Domestic Income, and Terms-
of-Trade Changes,” Journal of International Economics,
Vol. 62, No. 1, 2004, pp. 83-106.
[17] N. G. Mankiw, “Macroeconomics,”5th Edition, Worth
Publishers, New York and Basingstoke, 2003.
[18] A. A. Paul, R. M. Brent and J. S. Kenneth, “Improvem-
ents to the Food at Home, Shelter, and Prescription Drug
Indexes in the U.S. Consumer Price Index,” Bureau of
Labor Statistics, Working Paper 263, February 1995.
[19] D. S. Matthew and W. W. David, “Mismeasurement in
the Consumer price Index: An evaluation,” NBER Chap-
ter, in NBER Macroeconomics Annual, National Bureau
of Economic Research, Inc, Vol. 11, 1996, pp. 93-154.
[20] J. Dalén, “Computing Elementary Aggregates in the Sw-
edish Consumer Price Index,” Journal of Official Statis-
tics, Vol. 8, No. 1, 1992, pp. 129-147.
[21] B. Moulton, “Basic Components of the CPI: Estimation
of Price Changes,” Monthly Labor Review, Vol. 116, No.
12, December 1993, pp. 13-24.
[22] B. J. Szulc, “Price Indices below the Basic Aggregation
Level,” In: R. L. Turvey, et al., Eds., Consumer Price
Indexes: AnILO Manual, International Labour office,
Geneva, 1989, pp. 167-178.
[23] R. Turvey, “Consumer Price Indices: An ILO Manual,”
International Labour Office, Geneva, 1989.