Phân tích mức độ tiêu thụ Pizza của học sinh, sinh viên ở Hà Nội

ASSIGNMENT
PIZZA CONSUMPTION OF STUDENTS IN HANOI

September 25, 2015


TABLE OF CONTENT


I.

ABSTRACT

Bumble-bee team is a group of second-year students at Foreign Trade University, apart
from worrying about the high degree, has considerable concerns about health and money.
Especially, food has been a matter of great controversies perplexing our team whether
how we can pay a reasonable price for their goods and ensure the food safety. Started
from the actual needs of the students, the assignment is carried out through a detailed
survey of the behavior of consuming pizza of the students in Hanoi, in September 2015.
We hope that after this survey we will find what factors mainly affecting preferences to
type of pizza of the young generation at the moment to explain the habit of eat fast food
of the young in Hanoi at present. The survey in forms of multiple-choice questions
presented and filling information. As well as this goal, our result may hopefully be a
reference to the restaurant’s managers and chef in Hanoi to implement some policies and
make their products become one of the most attractive dishes. After collecting the data,
we generalize and classify data on base of the knowledge that we have learned. Softwares used are Typeform, Excel, Stata and Word to accomplish the assignment.

3


II.

1.

INTRODUCTION

Rationale

Until French colonization on the mid-19th century, Vietnam’s economy was mainly
agrarian and village-oriented. In 1986 Vietnam launched a political and economic
renewal campaign (Doi Moi) that introduced reforms intended to facilitate the transition
from a centrally planned economy to form of market socialism officially termed
"Socialist-oriented market economy”. Doi Moi combined economic planning with freemarket incentives and encouraged the establishment of private businesses in the
production of consumer goods and foreign investment, including foreign-owned
enterprises.
The Gross Domestic Product (GDP) in Vietnam expanded 6.44 percent in the second
quarter of 2015 over the previous quarter. GDP Growth Rate in Vietnam averaged 6.15
percent from 2000 until 2015, reaching a high of 8.46 percent in the fourth quarter of
2007 and a record low of 3.14 percent in the first quarter of 2009.
Moreover, Vietnamese people’s living standard has been increasing in the past few years.
A proof for this is that CPI of 2009 is 6.88% while in 2011, the CPI jumped to 18.12%
and exceeded the target. Therefore, they are able to afford their demand for food and
services.
In this report, we carried out to analyse an aspect of that field that is consumer behaviour
about eating pizza. We mainly focused on analysing influences of many factors such as
student’s income, quantity of pizza consumption, the ability of compete with spaghetti,…
2.

Objectives

- Surveying the satisfaction level of students about pizza
- Pinpointing different crucial elements influencing young generation uses in the opting

of pizza
4


- Investigating precise relationship in vital components reckoned by consumers for the
choice between pizza and spaghetti.
3.

Subject and Scope

-

Subject: students aged 18-24 in universities and colleges in Hanoi

-

Scope: data are collected from all students studying in universities and colleges, in

September 2015.
4.

Method of Research

-

Data source:

+ Primary source: information collected from 60 students through online surveys.
+ Second source: information collected from the Internet and textbook
-

Form: surveys in form of multiple-choice questions and filling information.

-

Support tools: Excel, Word, Stata.

-

Quantity: 59, in which 60 are valid.

After collecting the data from some different sources above, we summarized and analyse
information through statistical techniques to show the final result then write a report as an
overview of our team’s working process.

III.

LITERATURE REVIEW

1. History
Pizza is a baked pie of Italian origin consisting of a shallow bread-like crust covered with
seasoned tomato sauce, cheese, and often other toppings such as sausage or olive. The
pizza could have been invented by the Phoenicians, the Greeks, Romans, or anyone who
learned the secret of mixing flour with water and heating it on a hot stone.

5


In one of its many forms, pizza has been a basic part of the Italian diet since the Stone
Age. This earliest form of pizza was a crude bread that was baked beneath the stones of

the fire. After cooking, it was seasoned with a variety of different toppings and used
instead of plates and utensils to sop up broth or gravies. The first major innovation that
led to flat bread pizza was the use of tomato as a topping. It was common for the poor of
the area around Naples to add tomato to their yeast-based flat bread, and so the pizza
began.
While it is difficult to say for sure who invented the pizza, it is however believed that
modern pizza was first made by baker Raffaele Esposito of Naples. In fact, a popular
urban legend holds that the archetypal pizza, Pizza Margherita, was invented in 1889,
when the Royal Palace of Capodimonte commissioned the Neapolitan pizzaiolo Raffaele
Esposito to create a pizza in honor of the visiting Queen Margherita. Of the three
different pizzas he created, the Queen strongly preferred a pie swathed in the colors of the
Italian flag: red (tomato), green (basil), and white (mozzarella). Supposedly, this kind of
pizza was then named after the Queen as Pizza Margherita.
Later, the dish has become popular in many parts of the world:
•
•

The first pizzeria, Antica Pizzeria Port'Alba, was opened in 1830 in Naples.
In North America, The first pizzeria was opened in 1905 by Gennaro Lombardi at

53 1/3 Spring Street in New York City.
• The first Pizza Hut, the chain of pizza restaurants appeared in the United States
during the 1930s.
We take an obvious example about a group of professors researching customer’s
behaviors of Pizza Hut. The research they conducted is descriptive as well as theoretical
in nature. They use the easiest method to reach a lot of pizza customers such as data
collection, scaling technique, questionnaire development, pre – testing, sample
techniques and field work.

6

2. Customer satisfaction
It’s official. The American Customer Satisfaction Index (ACSI) Report on Airlines,
Hotels, Fast Food, Restaurants, and Express Delivery Services was released this week
and shows that Papa John's posted the top customer satisfaction score for both the pizza
business and the fast-food restaurant sector overall.
The ACSI uses a 100-point scale to measure customer expectations, perceived quality,
value, complaints, and loyalty and Papa John's posted the top score of 80 points, 7% up
from last year. Little Caesar and Pizza Hut tied just behind the leader at 78 points each
followed by Domino’s with 77 points.
The largest mainstream fast food brands all scored below their pizza chain competitors
with Wendy’s being the burger-chain leader at 77 points. Burger King showed the largest
improvement in score from the year prior by gaining 7.2% to score 74 points this year.
Taco Bell tied the King at 74 points while McDonald’s lost 4.3% to rank at the bottom of
this year’s ACSI study with just 64 points overall.
The ACSI report acknowledged that 2009 was a difficult year for restaurants of all types
and fast food sales were down 2.9% overall for 2009 after falling 1.2% in 2008. The
tough economy meant that consumers ate out less often and spent less when they did. The
pizza and fast food segment fared better than general restaurants due to their lower menu
prices, but both categories (fast food and restaurant) lost some business in 2009.
3. Price:
Selection of a Pizza outlet depends upon price value. The genre of restaurant is judged by
customer through Pizza selling price; with the view that a costly restaurant will provide a
better quality of both service and quality. The kind of restaurant, type of occasion,
profession and age group. The relative vitality of the restaurant decision vary extensively
by restaurant sort, eating event, age and occupation.
Rich clients select feel and ambience level as their determinant determination variables.
Different researchers have demonstrated price as client's first choice (Kara, 1995), (Park,
7

2004), (Andaleeb, 2006), (Tse, 2001) (Palazon, 2009). Introduction of novelty items and
limited time deals reap fruitful and recurrence sales. (Consuegra et al, 2007)
demonstrated that recognized price impartiality impacts buyer's contentment and loyalty.
Nonetheless, client fulfillment and loyalty are two paramount predecessors of value
acknowledgement. In the meantime, (Iglesias and Guillen, 2004), agreed that cost can
influence consumer satisfaction. Furthermore (Cater and Cater, 2009) proposed that
customer satisfaction is contrarily influenced by cost. It could be characterized as "the
procedure by which buyers translate cost and ascribe quality to a product or service". It
has intrigued specialists for some years. It is a well-known reality that cost and quality
are two significant elements of value. They both accelerate client gratification and
additionally client upkeep, which help increment the benefits of any business. So for a
chief of fast food restaurant it is significant to know customers discernment of value and
price.
Past studies analyzing the effect of cost on perceived value have inferred a negative
connection: the higher the value, the lower the product value is discerned. This is a
general phenomenon that when clients go out for shopping they have a tendency to
purchase items which have lower costs so they get a better value. This is upheld by
(Hutton, 1995) asserting that now more purchasers are attempting to boost quality for
cash used, requesting better quality at lower level costs. Even though this may not be
fully right for all the consumers on the grounds, since a few customers are ready to pay
more if they truly like a product. Higher recognized quality brings about a more amazing
eagerness by the purchaser to receive another item (Mcgowan & Sternquist, 1998).
Clients who are eager to pay higher costs for an item or service have a tendency to be
brand cognizant and renown touchy. They likewise accept cost is an indicator of quality
or status. When clients are persuaded that they are getting the best quality product or
service, they will have a tendency to improve reliability to it in the long run. Research led
by (Kandampully & Suhartanto, 2003) on hospitality industry revealed a positive
relationship between cost and client loyalty. Past studies looking at the effect of cost on

8


observed worth have proposed a negative connection: the higher the value, the bring
down the item quality is discerned (Dodds et al. 1991; Grewal et al. 1998). This is a
general wonder that when clients go out for shopping they have a tendency to purchase
items which have more level costs so they improve worth. This is underpinned by
(Hutton, 1995) guaranteeing that now more purchasers are attempting to boost worth for
cash used, requesting better quality at easier costs. In spite of the fact that this may not be
completely accurate for all the clients since a few clients are eager to pay more assuming
that they truly like an item. Higher discerned worth brings about a more amazing
eagerness by the customer to embrace another item (Mcgowan & Sternquist, 1998).

IV.

THEORETICAL FRAMEWORK

This entire research rests based on the theoretical framework. Since the theoretical
framework offers the conceptual foundation to proceed with the research, and since a
theoretical framework is none other than identifying the network of relationships among
the variables considered important to the study of any given problem situation, it is
essential to understand what a variable means in this study. Based on the literature
review, this research concentrates on conceptual framework of pizza consumption and its
impact on consumers‟ mind. This framework emphasizes variables such as price of pizza
(Pp), price of spaghetti (Ps), quantity of pizza consumers are willing to buy (W_t_b_p),
quantity of spaghetti consumers are willing to buy (W_t_b_s) and income. Comparing
pizza to spaghetti, we can assess the relationship between two kinds in fast food market.
Outline of the model structure:
The situation in which economic correlations involves only two variables are very rare.
Rather we have a situation where a dependent variable, Y, can depend on a whole series

of factorial variables or regressions. For example, the demand for pizza depends not only
on price but also on the prices of substitutes goods (spaghetti), the general level of

9


consumer prices and resources (income, satisfaction level). Thus, in practice, there are
normally correlations as:
Y = β1X1+β2X2 + β3X3 +...+ βkXk + ε
where values Xj (j = 2, 3, ..., n) represents the variable factor or regressors, the values
βj (j = 1, 2, 3, ...,k) represents the parameters of the regression and ε is theresidual factor
factor . Residual factor reflects the random nature of human response and any other
factors other than Xj, which might influence the variable Y. Note that we have adopted
the usual notation, we assigned to the first factor, notation X2, the second, notation X3
etc. In fact, as we shall see, it is sometimes convenient that a parameter β to be
considered a coefficient to a variable X1 whose value is always equal to unity. Then it
becomes possible to rewrite the equation in the form:
Y = β1X1+β2X2 + β3X3 +...+ βkXk + ε
In the case of pizza consumption of students in Hanoi, Conventional Linear Demand
Model: Qx = β0 + β1X1 + β2X2 + β3X3 + β4X4 + ε
QX – Index of pizza quantities (base year = 2015)
X1 – Own price of a given type of pizza
X2 – Price of related good (spaghetti)
X3 – Disposable Income
X4 – Trend
Ei – Error term
Without Trend

10

Source

SS

df

MS

Model
Residual

15.8871138
38.8247506

5
53

3.17742277
.732542464

Total

54.7118644

58

.943308007

Qp

Coef.

pp
Ps
W_t_b_p
W_t_b_s
income
_cons

6.00e-06
-.0000112
.3312113
-.0255937
2.47e-09
.4066745

Std. Err.

t

2.73e-06
4.80e-06
.1016999
.0804771
3.97e-09
.5175505

2.20
-2.34

3.26
-0.32
0.62
0.79

Number of obs
F( 5,
53)
Prob > F
R-squared
Adj R-squared
Root MSE

P>|t|
0.032
0.023
0.002
0.752
0.537
0.436

=
=
=
=
=
=

59
4.34

0.0022
0.2904
0.2234
.85589

[95% Conf. Interval]
5.25e-07
-.0000209
.1272272
-.1870104
-5.50e-09
-.6314003

.0000115
-1.61e-06
.5351954
.1358229
1.04e-08
1.444749

(Marshallian) Demand Function: Maximize U(X,Y) you have optimal x*=x*(px,py,I)
CES utility function: U(x, y) = where 0 = δ ≤ 1
Setting up the Lagrangian: L =
First-order conditions:

First-order conditions imply:
Substituting into the budget constraint:

Solving:
where

where

11


V.

EMPIRICAL FRAMEWORK

Based on theoretical framework analysed above (multiple regression), we can use the
method of least squares (OLS)
Ordinary Least Squares (OLS)
The parameters in econometric models are unknown constants. There are many methods
for estimating these parameters. Here, we will use the most common method that is
Ordinary Least Squares. The purpose of testing hypothesis is to determine the appropriate
level of models, the form of the model and sent out signals which violate the classical
assumptions of econometric models. We will use the appropriate model to evaluate the
relationship between the dependent variables and the explanatory variables, whereby, we
can assess, predict and make decisions on related issues.
If we assume, as in the case of two-variable regression that E(ε) = 0, then,
by substitution results:
Y = β1X1+β2X2 + β3X3 +...+ βkXk + ε
Estimate the parameters of the model:
The real value = + +
The estimated value =
The deviation = - = - Finding , so that the sum of the squares of deviation is min
Solving extreme function of two variables, we get:

With
12

is the average value of X and
is the average value of Y and
The assumptions of OLS:
Assumption 1: The relationship between Y and X is linear. The value Xi is given and is
not random.
Assumption 2: The deviation Ui is random variable with the average value of

Assumption 3: The deviation Ui which was random variable has variance unchanged.

Assumption 4: There is no correlation between the U1

Assumption 5: There is no correlation between Ui and Xi.

Theorem Gauss - Markov: When this assumption is ensured, the estimates calculated by
the method of OLS is the linear unbiased estimation, the most effective of the overall
regression.
Assumption 6: The deviation Ui has the normal distribution

The coefficient of determination of the model:
Total Sum of Squares (TSS):

13


Explained Sum of Squares (ESS):

Residual Sum of Squares (RSS):

We have:

The coefficient of determination:

Note:

R2 = 1: The model is perfectly suitable with the research sample
R2 = 0: The model is not perfectly suitable with the research sample

14


VI. DATA
Online survey software is a powerful survey tool for designing and administering online
surveys, collecting and managing accurate data, and facilitating advanced analysis and
reporting. Therefore, we used an online survey to examine the students’ behavior. In an
online survey, the respondent does not have the benefit of asking someone for
clarification. All they have to go on is the following information:
•

The Question Text – Words and Formation of the Sentence That Asks the

Question of the Respondent
• The Answer Set – The Range of Offered Answers for a Closed Question
• The Context – Prior Questions, Instructions, Guidance within the Questionnaire
 That survey questions are clearly understood and have the best chance of a
truthful, accurate response.
We dug in and analysed the data by starting to export the data in form of an excel table.
Then, we ran Stata, crunched the numbers to get the final results.
The reason for changing topic: The previous topic was cake consumption behaviour of

the students of FTU, but later we changed the topic because we felt that we should focus
more on a specific product. Moreover, the topic that we are doing is no longer restricted
to FTU students but to all students. That also makes it easier for us to gather data.
The process of making the survey form: determine what information we need
•
•
•
•
•

Income (subsidy from parents)
Pizza and spaghetti consumption per month
Pizza and spaghetti prices
To what extend do they love pizza? (from 1 to 5)
Time to complete the online survey: 2 days

There are 60 people answer the questions. The result we collected:
Price
Pizza
Spaghetti

35000
5/60
(8%)
27/60
(45%)

50000

70000

Other

13/60 (22%)

22/60 (37%)

20/60 (33%)

16/60
(27%)

11/60
(18%

6/60
(10%)

15


Frequency

1
25/60
(42%)
29/60
(48%)

Eat Pizza

Eat spaghetti

Preference
Score

0

Pizza

0

Spaghetti

4/58
(7%)

Place to
eat

2

3
9/60
(15%)
6/60
(10%)

12/60 (20%)
13/60
(22%)

Other
14/23 (23%)
12/60
(20%)

1

2

3

4

5

3/59
(5%)
3/58
(5%)

6/59
(10%)
4/58
(7%)

16/59
(17%)
16/58
(29%)

17/59
(29%)
17/58
(24%)

17/59
(29%)
14/58
(28%)

Pizza Hut

Pepperonis

Pizza Box

20/60
(33%)

17/60 (28%)

7/60
(12%)

Spaghetti
Box
7/60
(12%)

The purpose of the survey:
-

Find information about students’ consumption.
Students’ satisfaction about 2 products

VII. RESULTS AND DISCUSSION
The output is shown below, followed by explanation of the output.

16

Other
9/60
(15%)


Source

SS

df

MS

Model
Residual

15.8871138
38.8247506

5
53

3.17742277
.732542464

Total

54.7118644

58

.943308007

Qp

Coef.

pp
Ps
W_t_b_p
W_t_b_s
income
_cons

6.00e-06
-.0000112
.3312113
-.0255937
2.47e-09

.4066745

Std. Err.
2.73e-06
4.80e-06
.1016999
.0804771
3.97e-09
.5175505

t
2.20
-2.34
3.26
-0.32
0.62
0.79

Number of obs
F( 5,
53)
Prob > F
R-squared
Adj R-squared
Root MSE

P>|t|
0.032
0.023
0.002

0.752
0.537
0.436

=
=
=
=
=
=

59
4.34
0.0022
0.2904
0.2234
.85589

[95% Conf. Interval]
5.25e-07
-.0000209
.1272272
-.1870104
-5.50e-09
-.6314003

.0000115
-1.61e-06
.5351954
.1358229

1.04e-08
1.444749

a) Source - Looking at the breakdown of variance in the outcome variable, these are

the categories we will examine: Model, Residual, and Total. The Total variance is
partitioned into the variance which can be explained by the independent variables
(Model) and the variance which is not explained by the independent variables
(Residual, sometimes called Error).
b) SS - These are the Sum of Squares associated with the three sources of variance,

Total, Model and Residual.
c) df - These are the degrees of freedom associated with the sources of variance. The
total variance has N-1 degrees of freedom.

The model degrees of freedom

corresponds to the number of coefficients estimated minus 1.

Including the

intercept, there are 6 coefficients, so the model has 6-1=5 degrees of freedom.
The Residual degrees of freedom is the DF total minus the DF model, 58-5=53.
d) MS - These are the Mean Squares, the Sum of Squares divided by their respective
DF.
e) Number of obs - This is the number of observations used in the regression
f)

analysis. Our number of observations is 59
F( 5, 53) - This is the F-statistic is the Mean Square Model (3.17742277) divided

by the Mean Square Residual (0.732542464), yielding F=4.34.
17


g) Prob > F - This is the p-value associated with the above F-statistic. It is used in

testing the null hypothesis that all of the model coefficients are 0.0022.
h) R-squared - R-Squared is the proportion of variance in the dependent variable
(Qp) which can be explained by the independent variables (pp, Ps, W_t_b_p,
W_t_b_s and income). This is an overall measure of the strength of association
and does not reflect the extent to which any particular independent variable is
associated with the dependent variable. . You can see from our value of 0.2904
that our independent variables explain 29.04% of the variability of our dependent
variable
i) Adj R-squared - This is an adjustment of the R-squared that penalizes the addition
of extraneous predictors to the model. Adjusted R-squared is computed using the
formula 1 - ((1 - Rsq)((N - 1) /( N - k - 1)) where k is the number of predictors.
j) Root MSE - Root MSE is the standard deviation of the error term, and is the
square root of the Mean Square Residual (or Error).
k) Qp - This column shows the dependent variable at the top (Qp) with the predictor
variables below it (pp, Ps, W_t_b_p, W_t_b_s , income and _cons). The last
l)

variable (_cons) represents the constant or intercept.
Coef. - These are the values for the regression equation for predicting the
dependent variable from the independent variable.
The regression equation is presented in many different ways, for example:
Ypredicted = β0 + β1X1+β2X2 + β3X3 + β4X4 + ε

The column of estimates provides the values for β0, β1, β2, β3 and β4 for this equation.

•

pp (price of pizza) - The coefficient is 6.00e-06. So for every unit increase in pp, a

6.00e-06 unit increase in Qp is predicted, holding all other variables constant.
• Ps (price of spaghetti) - For every unit increase in Ps, we expect a 0.0000112 unit
•

decrease in the Qp , holding all other variables constant.
W_t_b_p (willingness to buy pizza) - The coefficient for W_t_b_p is 0.3312113.
So for every unit increase in W_t_b_p, we expect an approximately 0.33 unit
increase in the Qp holding all other variables constant.

18


•

W_t_b_s (willingness to buy spaghetti) - The coefficient for

W_t_b_s

is

-0.0255937. So for every unit increase in W_t_b_s, we expect a 0.025 point
•

decrease in Qp.
Income – The coefficient for Income is 2.47e- 09. So for every unit increase in

Income, we expect a 2.47e-09 unit increase in Qp
• Std. Err. - These are the standard errors associated with the coefficients.
• t - These are the t-statistics used in testing whether a given coefficient is
significantly different from zero.
• P>|t| - This column shows the 2-tailed p-values used in testing the null hypothesis
•

that the coefficient (parameter) is 0. Using an alpha of 0.05:
The coefficient for pp is statistically significantly because its p-value is 0.032,

which is smaller than 0.05.
• The coefficient for Ps (-0.0000112) is statistically significant at the since the pvalue (0.023) is smaller than 0.05.
• The coefficient for W_t_b_p (0.3312113) is statistically significantly different
•

from 0 because its p-value (0.002) is definitely less than 0.05.
The coefficient for W_t_b_s (-0.0255937) is not statistically significant because its

p-value (0.752) is larger than 0.05.
• The coefficient for Income is not statistically significant because its p-value
(0.537) is larger than 0.05
• The constant (_cons) is not statistically significantly at the 0.05 alpha level
•

because p-value (0.436) is larger than 0.05
p. [95% Conf. Interval] - These are the 95% confidence intervals for the
coefficients. The confidence intervals are related to the p-values such that the
coefficient will not be statistically significant at alpha = 0.05 if the 95%
confidence interval includes zero. These confidence intervals can help to put the
estimate from the coefficient into perspective by seeing how much the value could

-

vary.
 Conclusion:
The coefficient of price of pizza is 6.00e-06 (>0), the coefficient is 0.3312113 (>0)
so when Pp increases Qp increases => Pizza does not follow the demand rule =>
Customers researched (students in Hanoi) consider pizza as a Giffen Good

19


-

P-value of variables (Ps, W_t_b_s, _cons) is larger than 0.05 so we cannot analyse
and show any results
Own Elasticity:
Cross – price elasticity:
 Pizza and spaghetti are complement goods

VIII. IMPLICATION AND LIMITATION
1. Implication:
-

This can help customers to increase their likeability towards products
This survey can give opportunities to marketers to that understand students’ needs

-

and wants by producing more tasty food.

Analyze students’ preference to food at the moment

2. Limitations: In this study, there are several limitations
-

No interviewer: A lack of a trained interviewer to clarify and probe can possibly

-

lead to less reliable data
Survey Fraud: Respondents may not feel encouraged to provide accurate, honest

-

answers
Inability to reach challenging population: This method is not applicable for

-

surveys that require respondents who do not have an access to the Internet.
Do not analyze students’ income

IX. ASSESSMENT OF OUR GROUP
No.

Member

Tasks
Leader

1

Nguyễn Thị Phương Ly

Theoretical Framework
Edit assignment
Introduction

2

Nguyễn Thị Duyên

Data
Implication and Limitation
20

Comment
Complete the task
well and responsibly

Complete the task
well and responsibly


3

Lê Thùy Giang

4

Vũ Cao Quỳnh Chi

5

Đặng Ngọc Phương
Thảo

Results and Discussion
Data Analysis
Abstract
Empirical Framework
Literature Review
Data Analysis

21

Complete the task
well and responsibly
Complete the task
well and responsibly
Complete the task
well and responsibly


X.

REFERENCES

/> /> /> /> /> /> /> />sa=t&rct=j&q=&esrc=s&source=web&cd=7&cad=rja&uact=8&ved=0CF4QFjAGahUK
EwjEkpeNnpLIAhUh26YKHQeeA7o&url=https%3A%2F%2Fideas.repec.org%2Fp

%2Fags%2Faaea13%2F150777.html&usg=AFQjCNGkWVePqP2_rqHAhrtOyaKgvnWjg&sig2=FdP2Lm1kgqqrzEWRG7FPMA
/> /> /> /> />
22