Analysis of Influencing Factors of Tourism Revenue Based on
Multiple Linear Regression Model
Na Li and Gang Fang
*
Business School, Beijing Institute of Fashion Technology, Beijing, China
Keywords: Tourism Development, Economic Model, DPI.
Abstract: The report of the 20th National Congress of the Communist Party of China points out that "accelerating the
high-quality development of culture and tourism, and constantly meet the people's new expectations for a
better life". In the new era, people are more eager to live a better life, and tourism is gradually becoming an
important channel to enhance people’s sense of well-being. Tourism, as an important part of the tertiary
industry, plays a major role in enhancing the overall economy of China. In order to explore the economic
factors affecting the development of tourism in China, this paper uses computer technology to conduct
numerical simulations to analyse the relationship between the total income of the tourism industry and the per
capita disposable income of residents, the number of travel agencies, the number of tourists and other factors
in China from 1994 to 2019. Multiple regression analysis, tests and corrections are performed by
Econometrics Views. The final model regression results are used to make informative recommendations for
the recovery of the tourism industry.
1 INTRODUCTION
In the report of the 20th Party Congress, "promoting
cultural self-confidence and self-improvement,
forging new glories of socialist culture", it is clearly
stated that "insisting on shaping tourism with culture,
highlighting culture with tourism, and promoting the
deep integration and development of culture and
tourism" is an important requirement for the
prosperous development of cultural undertakings and
cultural industries. The cultural construction in a
prominent position, the work of culture and tourism
to make important arrangements, fully reflects the
Party Central Committee with Comrade Xi Jinping as
the core to the cultural construction and tourism
development of high importance. Although the
epidemic in China has gradually stabilized, there is
still a huge gap between the overall development of
the industry before the outbreak. Then, the
development of the tourism industry in the context of
the normalization of the epidemic is also a major issue
that the country urgently needs to address at present.
By analyzing the factors affecting the development of
China's tourism industry, we put forward targeted
suggestions for the development of China's tourism
*
Corresponding author
industry, thus promoting the healthy development of
China's tourism industry and at the same time
promoting the development of China's overall
economy.
This paper uses the computer software
Econometric Views (EViews), which refers to the
observation of quantitative patterns of socio-
economic relations and economic activities, using
econometric methods and techniques. The core of
EViews is through designing models, collecting
information, estimating models, testing models, and
applying models (structural analysis, economic
forecasting, and policy evaluation). In this paper
EViews processes time series data and performs
multiple linear regression using least squares to
assign real economic meaning to the regression
results. Based on the model regression results,
recommendations can be made in the context of the
current market environment on the one hand, and
forecasting the future development of the industry on
the other.
270
Li, N. and Fang, G.
Analysis of Influencing Factors of Tourism Revenue Based on Multiple Linear Regression Model.
DOI: 10.5220/0012029200003620
In Proceedings of the 4th International Conference on Economic Management and Model Engineering (ICEMME 2022), pages 270-275
ISBN: 978-989-758-636-1
Copyright
c
 2023 by SCITEPRESS – Science and Technology Publications, Lda. Under CC license (CC BY-NC-ND 4.0)

2 REVIEW OF THE
LITERATURE
Many domestic and foreign scholars have conducted
many analyses of the impact of tourism development
in recent years. SHI, P. H. et al (2020) selected panel
data from 60 tourism cities. Using the double
difference method to analyze the impact of tourism
model on the level of economic development of
tourism economy and constructing a region-wide
tourism model can increase tourism investment and
promote tourism development in the region. Zhou Li
(2019) constructs tourism consumer price index for
urban and rural residents in the process of tourism
consumption forecasting research found that tourism
consumer price index fluctuations in the short term
for urban and rural residents have an impact, while
long-term fluctuations only have an impact on rural
residents. DING Y.F. (2021) selected the tourism data
of Anhui Province from 2000-2018 for the
development pattern of tourism in Anhui Province,
the relationship between tourism revenue and per
capita GDP, road mileage, analyzed empirically using
EViews, and it was finally obtained that per capita
GDP and railroad mileage have a significant impact
on domestic tourism in Anhui Province. CAI L.J.
(2015) analyzed the data of domestic tourism industry
from 1994 to 2013 and modeled that the economic
influences that significantly affect the development
status of domestic tourism industry are the per capita
income of urban and rural residents and the number
of domestic population.
It can be seen that many scholars have done a lot
of research on the development of the tourism
industry, most of them have done research on the
development of tourism in a certain provincial area,
and there is a great limitation on the number of
samples selected. Therefore, this paper selects data
from 1994-2019 based on the research of many
scholars for the factors influencing the overall level
of development of the domestic tourism industry in
China, and proposes reference suggestions for
tourism development through the results of empirical
analysis combined with China's national conditions.
3 CURRENT STATUS OF
DOMESTIC TOURISM
DEVELOPMENT
In recent years, with the development of China's
social economy and the continuous improvement of
people's living standards, people's demand for
tourism is also increasing, the number of domestic
tourism shows a steady increment, and the total
income level of the tourism industry is also increasing.
Figure 1: Gross domestic tourism receipts and tourism
growth rate in the last decade
According to the analysis of China's tourism
industry revenue from 2010-2019, it can be seen that
China's tourism revenue has shown a rapid growth
trend in the last decade, from 1257.977 billion yuan in
2010 to 5,7250.92 billion yuan in 2019, an increase of
about 4.56 times. In 2009, the "Opinions of the State
Council on Accelerating the Development of
Tourism" for the first time clearly positioned tourism
as "a strategic pillar industry of the national economy
and a modern service industry that is more satisfying
to the people", resulting in a growth rate of 23.53% in
tourism revenue in 2010 and up to 53.46% The
tourism industry was hit hard by the sudden
worldwide epidemic in 2020, with a total domestic
tourism revenue of 222.8630 billion yuan, and since
then China's tourism industry has entered a moment of
silence.
4 ECONOMETRIC ANALYSIS OF
DOMESTIC TOURISM
REVENUE
The empirical part of this paper is divided into
identifying variables, modeling, regression parameter
estimation, economic significance and statistical
inference
4.1 Model Variables and Modeling
For the analysis of the factors influencing domestic
tourism income studied in this paper, the national
disposable income, the number of domestic tourists,
Analysis of Influencing Factors of Tourism Revenue Based on Multiple Linear Regression Model
271

and the number of travel agencies are used as
influencing variables, so as to explore the main
factors of domestic tourism income.
4.1.1 Analysis of Model Variables
The variables are set as in the table 1. Domestic
tourism industry revenue is also affected by other
unpredictable factors, let's say unquantifiable
variables such as consumption perceptions, so they
are included in the random disturbance term, denoted
by μ.
Table 1: Variable settings.
Y X1 X2 X3 X4 X5 X6 X7
Time
Domestic
tourism revenue
(billion)
DPI
(RMB)
Number of
travel agencies
(pcs)
Number of
domestic tourists
(million)
Total number of
star-rated hotels
(pcs)
Inbound
Visitors
(million)
Private car
ownership
(million)
4.1.2 Model Assumptions
To better explain the economic variables, the model
(1) is therefore treated as logarithmic.
Y=β
0
+β
1
lnX
1
+β
2
lnX
2
+β
3
lnX
3
+β
4
lnX
4
+β
5
lnX
5
+β
6
lnX
6
+β
7
X
7
+μ (1)
4.1.3 Data Sources
Data in this paper are from the National Statistics
Office and are recorded from 1994 to 2019 for
completeness. Inflation using CPI to eliminate
economic variables, i.e. Y/CPI (Xi/CPI), and the data
obtained after processing are shown in Table
4.2 Regression Parameter Estimation
The OLS regression estimation of the data in Table 2
using EViews yielded the results as in Table.
Table 2: EViews regression results.
Dependent Variable: LOG(Y); Method: Least Squares;Sample: 1994 2019;Included observations: 26
Variable Coefficient Std. Erro
r
t-Statistic Prob.
C -5.15429 5.144394 -1.001924 0.3297
LOG
(
X1
)
-3.513656 1.838875 -1.910764 0.0721
LOG
(
X2
)
0.066679 0.349255 0.190919 0.8507
LOG
(
X3
)
1.363423 0.531704 2.564252 0.0195
LOG(X4) -0.052115 0.227301 -0.229277 0.8212
LOG(X5) 0.067663 0.560575 0.120703 0.9053
LOG
(
X6
)
1.521488 0.771876 1.971155 0.0643
X7 0.03299 0.134708 0.2449 0.8093
R-s
q
uare
d
0.993102 Mean de
p
endent va
r
0.480515
Adjusted R-square
d
0.99042 S.D. dependent va
r
1.05365
S.E. of regression 0.10313 Akaike info criterion -1.457995
Sum squared resi
d
0.191444 Schwarz criterion -1.070888
Lo
g
likelihoo
d
26.95394 Hannan-Quinn criter. -1.346522
F-statistic 370.2196 Durbin-Watson stat 0.580444
Prob
(
F-statistic
)
0
4.3 Economic Significance and
Statistical Inference
4.3.1 Economic Significance Test
According to the regression results, the sign of the
coefficients of X1 and X4 is negative, which is
contrary to the economic significance and fails the
economic significance test. All other variables have
positive signs, which are in line with economic
theory, and therefore pass the economic significance
test.
4.3.2 Statistical Significance Tests
(1) Goodness-of-fit test
From Table 2, we can obtain R2 =0.993129, Adjusted
R-squared=0.990457. This shows that the estimated
ICEMME 2022 - The International Conference on Economic Management and Model Engineering
272

regression equation is a good fit for the sample
observations.
Significance test of the regression equation. In
order to test the significance of the linear relationship
between the explanatory variables and the explained
variables in the model from the overall, the original
hypothesis (2) tested H
0
:
H
0
：β
1=
β
2
=β
3
=β
4
=β
5
=β
6
=β
7
(2)
Given significance level, ɑ=0.1, n=26, k=7,
distribution side quantileF
0.1
(6, 18)=2.13, F
0.1
(8, 18)
=2.04, Take its average value, F
0.1
(7, 18)=2.085. F
=371.6717 > F
0.1
(7, 18)=2.085. Therefore, H
0
is
negated and there is a significant linear relationship
in the overall regression equation.
(2) Significance test of explanatory variables
The coefficients of the model explanatory
variables were tested for significance at the level of
significance α=0.1, and the hypothesis (3) was
formulated.
H
0
：β
i
=0 (i=1,2,3,4,5) (3)
Checking the t-distribution table, when the critical
value t
0.1/2
(24) =1.711 for degree of freedom of 24.
│t
1
│==1.8073> t
0.1/2
(24) = 1.711, Therefore, the
original hypothesis is rejected and β1 is significantly
not 0. After analysis, it can be seen that β3 and β6 are
significantly not 0, while β2, β4 and β7 do not pass
the significance test, and the original hypothesis β
2
=0
、β
4
=0 andβ
7
=0 is accepted.
4.3.3 Econometric Tests
(1) Multicollinearity test
The explanatory variables were tested for covariance
using EViews and the test results are shown in Table
3. From Table 3, it can be seen that there is a serious
multicollinearity between the explanatory variables,
so the stepwise regression method was used to correct
for it. The results obtained by stepwise regression
using EViews are shown in Table 4.
Table 3: Multicollinearity test
X
1
X
2
X
3
X
4
X
5
X
6
X
7
X
1
1.000 0.964 0.946 0.231 0.513 0.979 0.881
X
2
0.964 1.000 0.932 0.425 0.407 0.907 0.844
X
3
0.946 0.932 1.000 0.299 0.227 0.926 0.838
X
4
0.231 0.425 0.299 1.000 -0.338 0.041 0.328
X
5
0.513 0.407 0.227 -0.338 1.000 0.563 0.347
X
6
0.979 0.907 0.926 0.041 0.563 1.000 0.824
X
7
0.881 0.844 0.838 0.328 0.347 0.824 1.000
Table 4: Regression between lnY and each explanatory variable.
lnX
1
lnX
2
lnX
3
lnX
4
lnX
5
lnX
6
X
7
R
2
0.9878 0.8984 0.9854 0.4764 0.7664 0.9849 0.7718
Since the economic significance of the
explanatory variables in the regression process as
well as the p-value test are significant, Therefore, the
explanatory variables with larger R2 values were
selected as the main regressors of the model, and then
stepwise regression was performed to find the best
regression equation, as shown in Table 5
Table 5: Stepwise regression table.
lnX
1
lnX
1,
lnX
2
lnX
1
,lnX
3
lnX
1,
lnX
3
,l
nX
4
lnX
1
,lnX
3
,l
nX
5
lnX
1
,lnX
3
,l
nX
6
lnX
1
,lnX
3
,
X
7
t
√
×
√
××
√
×
Economic significance
√
×
√
√
√
×
√
Adjusted R-square
d
0.9873 0.9880 0.9900 0.9894 0.9897 0.9921 0.9893
The results of the stepwise regression are shown
in Table 5, passing the t-test as well as the economic
significance test for the variables X
1
and X
3.
The adjusted model R2 =0.990778, Adjusted R-
squared=0.98998, F=1235.452, indicating a good fit
of the model. Regression model (4):
LnY=-3.84306+0.87498*lnX
1
+0.56804*lnX
3
+μ
(4)
(2) Heteroskedasticity test
Analysis of Influencing Factors of Tourism Revenue Based on Multiple Linear Regression Model
273

A further test for the presence of
heteroskedasticity in the model using the White's test,
the constructed auxiliary function (5)
λ
2
=ɑ
0
+ɑ
1
LnX
1
+ɑ
2
LnX
3
+ɑ
4
LnX
1
LnX
3
+ɑ
5
(LnX
1
)
2
+ɑ
6
(LnX
3
)
2
(5)
Formulate a hypothesis (6)
H
0
: ɑ
i
= 0, i= 1,…, 6
H
1
: At least one ɑ
i
is not 0
(6)
The results of the White test are shown in Table 6.
Table 6: White test results.
Heteroskedasticity Test: White; Method: Least Squares; Sample: 1994 2019; Included observations:26
F-statistic 2.050364 Prob. F(5,20) 0.1148
Obs*R-square
d
8.81095 Prob. Chi-Square(5) 0.1168
Scaled explained SS 4.424215 Prob. Chi-Square(5) 0.4901
Variable Coefficient Std. Erro
r
t-Statistic Prob.
C -0.323536 3.424603 -0.094474 0.9257
LOG(X1) -0.383148 1.855374 -0.206507 0.8385
(LOG(X1))^2 -0.057874 0.242156 -0.238996 0.8135
(LOG(X1))*(LOG(X3)) 0.100552 0.434789 0.231267 0.8195
LOG(X3) 0.261064 1.633977 0.159772 0.8747
(LOG(X3))^2 -0.039413 0.193569 -0.203611 0.8407
From Table 6, we get nR
2
=8.810950, Tested by
White, α=0.1, ᵪ
2
0.1
(5)=9.24, nR
2
=8.810950<ᵪ
2
0.1
(5)=9.24,Therefore, accepting the original hypothesis
H
0
, it can be assumed that there is no
heteroskedasticity in the model.
(3) Autocorrelation test
The DW test is used to test whether there is
autocorrelation in the error term μ. The hypothesis is
proposed
H
0
: μ is not autocorrelated
H
1
: μ is autocorrelated
(7)
Table 7: Regression results of lnY with lnx1 and lnX3.
Dependent Variable: LOG(Y); Sample: 1994 2019; Included observations: 26
Variable Coefficient Std. Error t-Statistic Prob.
C -3.843062 0.882877 -4.352883 0.0002
LOG(X1) 0.874978 0.239102 3.659432 0.0013
LOG(X3) 0.568045 0.208885 2.719411 0.0122
R-squared 0.990778 Mean dependent var 2.782924
Adjusted R-squared 0.989976 S.D. dependent var 1.053909
S.E. of regression 0.10552 Akaike info criterion -1.551673
Sum squared resid 0.256091 Schwarz criterion -1.406508
Log likelihood 23.17175 Hannan-Quinn criter. -1.509871
F-statistic 1235.452 Durbin-Watson stat 0.485045
Prob(F-statistic) 0
s.e.=0.105520, DW=0.485045, Check the DW
statistics table, K=2, n=26, α=0.1, dl=1.224, Du
=1.553, DW=0.485045<dl =1.224< du =1.553 in this
model. Reject the original hypothesis, Therefore, μ is
autocorrelated Standard error correction is performed
using the newey-west method in EViews software to
eliminate autocorrelation problems.
5 ANALYSIS OF RESULTS AND
RECOMMENDATIONS
Through quantitative analysis, the factors that affect
our tourism revenue are PDI and the number of
domestic tourists. All other things being equal, i.e.,
ICEMME 2022 - The International Conference on Economic Management and Model Engineering
274

for every 1 percentage point increase in per capita
disposable income of residents, domestic tourism
income increases by an average of 0.87 percentage
points. For every 1 percentage point increase in the
number of domestic tourists, domestic tourism
revenue increases by an average of 0.57 percentage
points, according to which the following
countermeasures are proposed.
PDI significantly affects the total income of the
tourism industry, and the steady increase of residents'
income makes residents pay more attention to the
pursuit of spiritual life and choose to travel for such
enjoyment-oriented consumption, this will further
promote the increase of tourism income. With the
established level of per capita income, the
government should strengthen as well as improve the
social security system to increase the marginal
propensity to consume and thus expand the
consumption support (CAI, L. J. 2015). With the
increase of residents' per capita income, people will
pursue high-quality tourism products more, so the
tourism industry can also be transformed to high-end
tourism. For example, business tourism, health care
tourism, cruise ships, Marine leisure tour. (CHEN, Q.
J. & RAO, W. Y. 2019).
The results of the study show that the number of
domestic tourism affects tourism income
significantly and positively, while the main
countermeasures that affect the number of travellers
in China can be proposed in the following aspects.
Firstly, guidance from the ideological level, while
vigorously developing the economy, to improve the
thinking of the residents of tourism consumption, to
encourage and spread positive tourism ideas and
concepts. Secondly broaden the marketing channels
(YU, Tong & YE, Y. L. 2016). the background of the
epidemic, soft marketing more into the hearts of
people, users can read after the empathy and take the
initiative to share a city's tourism information, not
only has a lower cost of publicity, but also in the
current context to play the most effective role.
ACKNOWLEDGEMENTS
Support by: The construction program of innovation
team at Beijing Institute of Fashion Technology
(BIFTTD201901); “The first batch of new liberal arts
research and reform practice projects of the Ministry
of Education” project (Project No.: 2021140009).
REFERENCES
CAI, L. J. (2015). An empirical analysis of the economic
influences on the development status of domestic
tourism. J. Commercial Accounting. 14, 110-112.
CHEN, Q. J. & RAO, W. Y. (2019). Empirical analysis of
influencing factors of tourism in provinces and cities of
China. J. New Finance Economics. 11,51-53.
DING, L. F. & Li, C. S. (2021). Construction and Empirical
analysis of Tourism development Mode in Anhui
Province. J. J. BaojiColl.Arts&Sci. (Nat.Sci.). 41(01),
52-58
SHI, P. H. et al (2020). Evaluation on the effect of the
establishment of all-region tourism demonstration zone
on the development of regional tourism economy-
Empirical analysis based on difference and difference.
J. Macroeconomics. 06,122-132+175.
YU, Tong & YE, Y. L. (2016). An empirical study on
influencing factors of domestic tourism income in
Guangdong Province. J. Journal of Wuhan Business
University. 30(04),19-22.
ZHOU, Li (2019). An empirical analysis of the construction
of tourism consumer price Index on the tourism
consumption forecast of urban and rural residents -Test
based on VAR model. J Prices Monthly. 06,1-5.
Analysis of Influencing Factors of Tourism Revenue Based on Multiple Linear Regression Model
275