Examining the Effectiveness of Government Policy for Retail Districts: Evidence from Korea

Abstract

This paper seeks to measure the effects of policy on the self-employed conducting business in traditional retail districts. To verify policy we performed a practical analysis utilizing the multinomial logit model on the relationship between consumption behaviors and the attendant satisfaction level of consumers in the urban commercial districts. We first identified traditional retail districts that had received government policy support and those that had not. We then visited these districts to survey the satisfaction levels of customers. In total, 400 people were surveyed for this study. The results show that political support for the self-employed who conduct business inside the traditional retail districts has a partial effect. Especially, for the policy beneficiaries, the customer’s satisfaction level in specific political support has a very important meaning. The study analyzed the factors necessary for the continuous growth of traditional retail districts, considering consumption behaviors and shop selection attributes. We argue that government support can help sustain regional commercial districts and the individual self-employed through consumer behavior and the magnification of satisfaction levels.

1. Introduction

We first raise some key issues before discussing the effect of government support on sustainable growth of self-employed people: the core issue in this study. How much support might self-employed retailers actually need? Indeed, why should governments support the self-employed at all? Do those self-employed who do not receive government support not suffer discrimination? To answer these basic questions it is necessary to analyze old urban traditional retail districts: some in which the self-employed who receive government support and others where they do not. This analysis will measure the effect of government support on actual consumers and henceforth future policy supplementations can refer to this analysis for guidance. For a systematic analysis, this study analyzes the changes in the consumer purchase patterns that have brought macroscopic and microscopic changes in the global distribution industries. In addition, it will also review the limitations of sustainable growth due to the decline in the competitive power of the self-employed who do not adapt themselves well to ongoing changes in the distribution industry. In addition, it will examine whether the policy efforts of the government to help revive the self-employed have any lasting effects.

Traditional retail formats and thus the competitiveness of self-employed ‘independent’ retailers are constantly declining across the world [1]. New retail formats a rose due to rising consumer demand and because competitive retailers pursued cost reductions: leading the changing trends in retail formats [2]. Most theses on the changes in the distribution industry trends state that new formats such as outlets and online shopping will grow continuously, and the traditional framework will continue to decline [3]. Consumers are increasing in affluence, becoming more discerning, and now own smartphones. New format retailers are meeting new consumer needs by providing a range of formats both in the center of town and out of town [4].

Overall, however, according to statistics from the Organization for Economic Co-operation and Development (OECD) [5], the ratio of the self-employed is not decreasing. As of 2000, the ratio of the self-employed in England was 12%, Netherlands 11%, Denmark 8.6%, and as of 2016, they were 15%, 16% and 9.0%; thus, they have instead increased. One reason for this increase in some OECD nations is that although the new formats increase, small retailers adhering to traditional business methods could not satisfy consumer needs [6]. Another reason is the ‘gig economy’ where workers who were once employees are often now self-employed even when doing much the same work. It follows that, overall, self-employment is itself difficult to measure and it varies sector-by-sector.

As self-employed demographics degrade in certain sectors in some nations, governments may feel political pressure to implement policies to support the self-employed. Government policies include loan guarantee schemes, technology transfers, loan programs for the self-employed, and R&D, etc. [7]. The US Small Business Innovation Research Program provides $2.5 billion to the self-employed each year [8] and the Korea small and medium venture business division provides $1.8 billion each year to support the self-employed [9]. There are approximately 3.98 million self-employed in Korea, which is a very high figure [10]. Korea ranks in fourth place among OECD nations, following Lithuania, Romania, and South Africa.

Relatively, the ratio of the self-employed in Korea is very high, considering its population of only 50 million. Thus, from the Korean government’s perspective, the very high ratio of the self-employed was the obstacle degrading economic growth in the short-and long-term; therefore, the government decided to provide substantial political support for the self-employed. The Korean government supports a policy fund program, labor union new program, entrepreneurship and education program, etc., for management stability and growth for the self-employed. It has been reported that changes in external conditions, such as the opening of distribution markets and changes to their structure, have negative effects on the revitalization of traditional retail districts. Along with these factors, internal problems caused by the decline of traditional retail districts also had a positive effect to the extent of enabling the self-employed to take advantage of the rapidly changing distribution markets that lag behind in facilities. In response to the downturn experienced by traditional retail districts, the Korean government has amended the related legal systems and is providing policy support to various businesses to secure their competitiveness. Some representative applicable laws aimed at supporting traditional retail districts include the 2005 and 2006 special laws pertaining to the promotion of traditional retail districts. The examples of government-supported measures aimed at the revitalization of traditional retail districts through the central government and local autonomous entities include the support toward facility modernization and management innovation. Facility modernization includes the installation of arcades, securing parking lots, improvement of buildings, and the securing of entrances, whereas management innovation includes marketing, educating businesspersons, and the issuance of gift cards. However, there is a lack of robust evidence that these government programs have a positive effect on the growth of the self-employed [11]. However, the long-term effect cannot be measured correctly using conventional evaluation techniques and can only be measured on a short-term basis [12].

Thus, although it is not possible to show the effect of government support policies in the short term, they may appear in the longer term. Thus, there are many difficulties in evaluating such an effect. It is almost impossible to obtain the sales data for 5 or 10 years from any traditional retail district. Thus, this study focuses on short-term support rather than long-term support to detect the effect of government support. There are approximately 3.98 million self-employed in Korea; however, it is assumed that those receiving government support are only a small portion of this figure. Furthermore, there are limitations in judging the political effect on the self-employed who have received government support for their retail businesses in some traditional retail districts [10]. Considering the practical limitations of investigating 1439 traditional retail districts spread across the nation, it is necessary to systematically select regions where government policy has been provided and those that did not receive government support. Especially, government policy support in line with structural changes in the distribution industry in Korea. Thus, the selection of districts should allow for this fact.

In fact, with government support, not only can it attract customers to increase sales in the short term, but it can also contribute to the development of regional and national economies through sustainable growth in the long run. Therefore, self-employed individual retailers, who may have disappeared without government support, remain competitive with that support. Therefore, it is argued that government’s support would result in economic growth and a decline in unemployment. However, governments should exercise caution when intervening as a financial provider because the effects of government intervention can vary by country and region [13]. Hence, it is important to measure the effects of government programs region by region.

As the purpose of this study is to measure the policy-receiving effect of the self-employed conducting business in commercial districts, it uses quantitative methods for a scientific and systematic analysis. This study attempts to perform a practical analysis utilizing the multinomial logit model on the effect relationship between the consumption behaviors and the satisfaction level of consumers in the urban commercial districts, and verify political policy. The purpose of this study is to categorize traditional retail districts into those that have received government policy support and those that have not, and visit these districts to survey the satisfaction levels of customers. This study selects traditional retail districts that did not receive government policy support selectively and surveys the effect of each policy type on the customer satisfaction level to help refine such policy further.

The variables with positive effects on consumer satisfaction levels carry implications for developing countries, which may suffer many trials and errors when (re)vitalizing their traditional retail districts. The differences between this study and conventional studies are as follows. First, as this paper divides, analyzes, and segments districts by type, it reveals political implications with high practical utilization and applications. However, most conventional studies [11,14] have utilized panel data and data provided by institutions such as OECD, etc.

However, such conventional data do not analyze the effects of political policy on consumers in many aspects, or examine the variables with possible effects on consumers. In addition, few conventional studies have verified the political benefits based on the satisfaction level of consumers. Thus, this study intends to propose methods that can magnify the political effect based on the consumer satisfaction level measured in a short period of time. In addition, this study considers the specialty of the retail business and proceeds from the consumer perspective, unlike conventional studies, which have analyzed quantitatively, based on time series data. The premise is that the critical success factors in the retail market are the awareness and behavior of the consumer [15,16].

2. Literature Review

Many studies have examined consumer behavior in order to understand consumer choices. As consumer behavior changes, many studies have examined related customer awareness, attitude, and behavior intentions [17]. Recently, in the retail business, much better services and facilities resulting from fierce competition are the main factors that determine frequency of visits and loyalty of customers. Shop attributes are the variables that work as the dominant factor when a consumer chooses the shop, and over time, many studies have examined shop attributes [18,19,20,21,22,23,24].

Furthermore, over many decades, studies have examined the effect of demographic variables, such as gender, age, and marital status etc. on consumer selection of retail shops [25,26,27,28,29,30,31,32]. In addition, several studies have used socio-economic factors such as academic background and income, etc. [25,28,33]. Other studies focus on consumer expenditure [33,34,35] and number of visits [36,37,38]. Nilsson et al. empirically analyzed how the socio-demographic characteristics and shopping behavior of Swedish consumers relate to shop attributes in order to understand the determining factor in shop selection [39]. Regarding conventional shop selection attributes, studies have applied standardization and integrated it into overall attractiveness attributes, and accessibility attributes.

Findings show that variables such as age, gender, visit frequency, and expenditure size and price range, and quality of the product affected shop selection. A study by Martinez-Ruiz et al. [40] reflected the periodic time difference (2008 and 2013) from the economic crisis onwards, and investigated the effect relationship between the shop attributes and customer satisfaction over unplanned buying in Spain. First, the shop attributes were derived to be three each in 2008 and 2013. According to their regression analysis, the result for 2008 showed that store atmospherics, product assortment, home delivery, free parking, and store accessibility, etc., had an effect. The result for 2013 showed that product and shop mood and accessibility had positive effects on the satisfaction level. The consumer’s selection of shop and commercial district is determined by the shop’s image, which depends on the attributes of the retail shop, leading to expenditure or purchase behavior. In other words, such a result affects customer satisfaction—though differently when considering food versus non-food outlets. In relation, there are many studies on shop attributes and customer satisfaction levels [39,40,41]. When choosing a shop, a consumer considers the facilities/services, environmental factors, accessibility, shop mood, salespersons, and their competencies, and so on, as their main factors. When there is a feeling of satisfaction with the experience, the consumer revisits the shop and may recommend it to acquaintances [29,42,43,44,45,46,47,48,49]. Location factors are also important factors that affect consumer purchase behavior in Korean traditional retail districts [50]. It follows that the support direction of the government should be to improve the environment of traditional retail districts and to modernize their facilities etc.

In other words, if government policy improves locality factors, such as the environment and facilities of the traditional retail market, the number of consumers will increase, leading to the (re)vitalization of the commercial district. From previous studies, such as [51,52] etc., this study deduces four factors: convenience, product, atmosphere, and location. Hence the shop and commercial district selection attributes relevant to government business support are as follows: convenience [20,40,53,54,55], product [20,27,40,55,56,57], store atmosphere [40,48,55,58,59] and location [20,37,39,40,48,55,60,61,62].

3. Method and Model

3.1. Data

Consumer attraction is necessary for the (re)vitalization of traditional retail districts, as fewer visits by consumers will eventually degrade such traditional retail districts. Second, although the prime beneficiaries of the policy are the self-employed in such districts, eventually, consumers should visit and purchase often, for the policy to have real effect. Accordingly, this study focuses on consumer choice according to the types of support for traditional retail districts. We select representative cities per region and classify them into traditional retail districts (management and facility field) that received government support and those that did not, using a questionnaire.

To secure representativeness, five large cities (Figure 1) were chosen for the survey. The regions included Chuncheon city, Gangwon province, Bucheon city, Kyeonggi province, Jaecheon city, Chungcheongbuk province, Yeongju city, Kyeongsangbuk province, and Jeonju city, Jeonlabuk province. For each city, districts that received 1–2 schemes of government support and those that did not receive such support were included as survey objects. One district each in Chuncheon city, Jaechon city, and Yeongju city received both facility and management support, and one district in Bucheon received only facility support. One district in both Bucheon city and Jeonju city received only management support. We conducted a survey for 20 days (1–20 December 2017) through a research company; during this period, the surveyors visited traditional retail districts and conducted individual interviews. In total, 400 people were surveyed for this study and 120 questionnaires were for those that received facility and management support, 20 questionnaires for those that received only facility support and 80 for those that received only management support. Furthermore, 120 questionnaires were for those that did not receive government support at all. For a detailed survey, a total of 40 answers were acquired from the districts that received facility and management support in Chuncheon city, Gangwon province, and 40 answers were acquired from non-supported commercial districts in the same city. In total, 80 answers were acquired in the districts that received only facility support in Bucheon city, Gyeonggi province and 40 in the commercial district that received only management support in the same city. In total, 40 responses were acquired in commercial districts that received both facility and management support: in Jaecheon city, Chungcheonbuk province and 40 in the non-supported commercial districts in the same city. In total, 40 questionnaires were collected in the commercial districts that received both facility and management support in Yeongju city, Kyeongsangbuk province and 40 in the non-supported commercial districts in the same city.

In total, 40 questionnaires were collected from the commercial districts that received only management support in Jeonju city, Jeonlabuk province. This study set the consumer selection per type of government support as the dependent variable, and the independent variables were psychological factor, retail shop selection attributes, and the customer’s visit behavior, personal characteristics, and socio-economic factors, etc. Depending on the type of government support, the consumer-attracting attributes are seen as the effect factor in customer’s selection through maintenance and development. Thus, this study sets the retail shop selection attributes as the main independent variables and the individual demographic characteristics, which include the number of visits, expenditure amount, customer visit behavior, and the socio-economic factors, etc., as the dependent variable group.

3.2. Model

The Multinomial Logit model (MNL) is a selection model used when the categorical dependent variable is polynominal, being three each or more, and is suitable for use in comprehending the influence of the independent variable, which affects each selection model in several choice sets. The details are as follows. According to the McFadden’s random utility theory [63], the usefulness function about the jth alternative of the ith individual is expressed as follows.

$$U_{ij}= V_{ij}+ {ϵ}_{ij}= {\beta }_j^t{X}_i+ {ϵ}_{ij}$$

$V_{ij}$ is the average usefulness and ${ϵ}_{ij}$ is the error clause in the above usefulness function (where $V_{ij}$ is the average utility, ${ϵ}_{ij}$ is an error term). $X_i$ represents the characteristics of the individual ($X_i$ is the set of independent variable), and ${\beta }_j^t$ is the vector of each alternative (${\beta }_j^t$ is a vector of unknown parameter). The probability $P_{ij}$ that an individual I chooses an alternative j is the same as the probability that the usefulness $U_{ij}$ is bigger than all the other usefulness, $U_{ik}$ in the selection set of an individual. Regarding the error term, the probability that individual I chooses the alternative j, under the assumption that the error clauses are independent from one another is as follows.

$$P_{ij}=\frac{e^{{\beta }_j^t}{X}_i}{{\displaystyle \sum _{k\in C}}{e}^{{\beta }_j^t{X}_i}}, all j\in C$$

This study utilizes SPSS (Statistical Package for the Social Sciences) for technical statistics and factor analysis corresponding to basis analysis. For the support type-specific customer characteristics analysis, this study uses the Limdep (limited Dependent) statistics package.

4. Analysis Results

4.1. Characteristics of Samples

The response statistics of the variables used in this study are as follows (

Table 1. Variables.

	Variable	Remark	Mean/Freq (%)	Std.D
Independent Variables	Gender	Male = 1, Female = 0	0.10	0.30
	Age	Younger 20s = 2, 30s = 3, 40s = 4, 50s = 5, 60s or older = 6	2 = 4.8, 3 = 21.0, 4 = 36.0, 5 = 27.8, 6 = 10.5
	Edu	Over University Graduate = 1, Under University Graduate = 0	0.43	0.49
	Marital Status	Married = 1, Single = 0	0.89	0.30
	Income	1391 = 1, 2319 = 2, 3246 = 3, 4174 = 4, 5565 = 5 Monthly Household Income (US$)	1 = 20.5, 2 = 20.5, 3 = 26.5, 4 = 21.8, 5 = 10.8
	Visit	Number	3.26	3.23
	Expen	Expenditure per term (KRW: won)	36,582	22,537
	Fac1	Convenience (5-point Likert scale)	2.43	0.69
	Fac2	Product (5-point Likert scale)	3.09	0.65
	Fac3	Atmosphere (5-point Likert scale)	2.91	0.66
	Fac4	Location (5-point Likert scale)	3.29	1.07
	Satisfaction	Overall Satisfaction (5-point Likert scale)	3.86	0.62
Dependent Variable	Type of policy	Facility = 1 (n= 90), Management = 2 (n = 71), Facility + Management = 3 (n = 120), No Support = 0 (n = 117)

Note: US$1 = 1078 won.

). There were 90% female respondents with the highest distribution of 36.0% in the 40s age group. As to education level, individuals with university level or above were 43%, also 89% were married, and the monthly average income per household was about $3000. The average number of visits in the commercial district under investigation was 3.26 times per week and the expenditure amount per visit was $34. As to the importance of shop attributes, convenience scored 2.43, product 3.09, atmosphere 2.91, and location 3.29.

4.2. Analysis of the Reliability and Validity of Shop Attributes

The reliability and the validity of the shop attributes were verified through factor analysis (

Table 2. Results of the factorial analysis.

		Eigenvalue	Factor Loading	Variance	Cronbach’s α
Convenience Factor	The point system is effective(Coupon)	6.191	0.764	36.419	0.830
	Exchange and refunding		0.763
	Convenience of the payment method		0.717
	Recreational facilities are well provided		0.703
	Price parking system is efficient		0.639
	Parking lot is convenient		0.457
Product Factor	Product Diversity	2.125	0.806	12.500	0.835
	Product Excellence		0.771
	Product Reliability		0.746
	There are many famous brands		0.706
	Product Price		0.698
Atmosphere Factor	Hygienic Product Preservation	1.625	0.800	9.556	0.850
	Externally clean		0.791
	Luxurious mood		0.740
	Good illumination facility		0.720
Location Factor	It is convenient to go on foot.	1.176	0.929	6.916	0.877
	Close from home		0.905
Cumulative variance = 65.390

). The validity analysis of 25 individual variables showed that four factors were extracted among the 17, except for eight variables that degraded the validity, as their value was 0.60 or under. Factors with an Eigenvalue of 1 or over were extracted using the Varimax method of orthogonal rotation. The cumulative variance was around 65.4% and the Cronbach’s α coefficients were all 0.6 or over. The factor analysis ranks the shop attributes as follows: first is the customer’s shopping convenience, such as payment-related variables, point system, recreational facilities, and parking facility. Second is product-related attributes such as variety, superiority, reliability, and price of the product. Third is atmosphere factor, such as the status and mood of the shop, and the last is the location factor related to accessibility [20,27,51,52,54,55,59,60,62].

4.3. Analysis Result

4.3.1. Examination of Differences in Consumer Behavior Depending on Provision of Government Support

Analysis of results on the differences in consumer behaviors in respect of shop attributes—depending on whether or not support was given—showed that the product factor (5% significant level), convenience, and the atmosphere factor (1% significant level) showed differences (

Table 3. Analysis of the difference in visit behavior depending on support.

Variable	Support Status	Mean (Std.D)	t-Value	p-Value
# of visit	No Support	3.37(3.40)	1.158	0.247
	No support	2.96(2.76)
Expenditure	No Support	37,953(20,480)	1.980	0.048 *
	No support	33,108(26,441)
Convenience	Supported	2.55(0.67)	5.543	0.000 **
	No support	2.14(0.66)
Product	Supported	3.15(0.59)	2.202	0.028 *
	No support	2.99(0.78)
Atmosphere	Supported	3.03(0.60)	5.622	0.000 **
	No support	2.63(0.71)
Location	Supported	3.35(1.01)	1.402	0.162
	No support	3.18(1.20)
Satisfaction	Supported	3.49(0.62)	3.908	0.000 **
	No support	3.19(0.86)

* p < 0.05, ** p < 0.01.

). The expenditure amount and the overall satisfaction levels were different at the 5% and 1% significance levels, respectively, and thus, the average value in the districts where received support was made was high. We may infer that, whatever the prior motivation for intervention, the political support of the government for the self-employed retailers has had an effect.

4.3.2. Characteristics Analysis per Support Type (MNL)

For the theoretical examination of the MNL model in order to understand government support type-specific characteristics, this study comprehends the signs of the parameter, relationship with conventional theory, and statistical examination using the t-value of the individual independent variable. The Likelihood test was used to examine the whole model and the model Χ² shows a significant value of 280.7241 at the 1% level (

Table 4. Results of the model analysis.

Variables		Facility			Management			Facility + Management			No Support
	Coef.	S/E	t-Value	Coef.	S/E	t-Value	Coef.	S/E	t-Value	Coef.	S/E	t-Value
Constant	−0.960	0.17	−4.447 ***	−0.934	0.17	−3.905 ***	0.388	0.24	1.304	1.50	0.24	4.951 ***
Gender	−0.071	0.06	−0.985	0.089	0.07	1.197	−0.146	0.10	−1.395	0.128	0.09	1.364
Age	−0.058	0.02	−2.643 ***	−0.008	0.02	−0.353	0.037	0.03	1.220	0.029	0.03	0.913
Education	−0.102	0.03	−2.497 **	−0.070	0.04	−1.441	0.224	0.06	3.552 ***	−0.051	0.06	−0.804
Marital status	0.188	0.09	2.009 **	0.133	0.08	1.476	−0.101	0.10	−0.974	−0.219	0.10	−2.069 **
Income	0.000	0.00	5.476 ***	0.000	0.00	3.149 ***	−0.001	0.00	−6.484 ***	0.000	0.00	0.467
Visit	−0.004	0.00	−0.705	0.034	0.00	5.226 ***	−0.028	0.01	−2.832 ***	−0.001	0.00	−0.133
Expend	0.000	0.00	2.254 **	0.000	0.00	0.762	−0.000	0.00	−1.413	0.000	0.00	−0.351
Convenience	0.120	0.04	2.908 ***	−0.128	0.04	−3.009 ***	0.167	0.05	2.930 ***	−0.158	0.05	−2.691 ***
Product	0.116	0.03	3.199 ***	−0.105	0.04	−2.593 ***	−0.048	0.04	−0.983	0.037	0.05	0.745
Atmosphere	−0.061	0.04	−1.548	0.206	0.05	4.156 ***	−0.010	0.06	−0.178	−0.135	0.06	−2.260 **
Location	0.040	0.02	1.902 *	0.082	0.02	3.555 **	−0.103	0.02	−3.777 ***	−0.019	0.02	−0.703
Satisfaction	0.012	0.03	0.381	0.061	0.03	1.594	0.078	0.04	1.710 *	−0.152	0.04	−3.268 **
	Log likelihood function						−403.6818
	Restricted log likelihood						−544.0438
	Model X2						280.7241 *
	Number of Observation						398

*, **, *** indicates values at the 10%, 5%, and 1% significance levels, respectively. * p < 0.10, ** p < 0.05., *** p < 0.01.

). In this research, the type of government support—facility, management, facility + management, and non-support—is applied as a dependent variable. Independent variables that are used in this analysis are classified into the demographic, visit, and retail attribute factors. The results of the analysis of the government support type areas follows.

First, the visitors in the commercial districts that were facility supported comprised mostly married respondents [26], low age groups [17,64], and low education levels (not interested in complex marketing techniques). It is assumed that the high-income level and expenditure size are because of high accessibility as key shop attributes [37,39] and shopping convenience [64,65] accompany consumer awareness about product factors [57,64,66].

Consumers in the commercial districts with management support did not show significantly higher expenditure levels, regardless of high income level and frequency of visits [37]. In addition, among the shop selection attributes, the satisfaction level for shopping convenience and product quality was low whereas the satisfaction level for location was high [37,39,67,68]. Therefore, the consumer is more likely to be satisfied with visit attributes and the management improvements, etc., if they are adjacent to the customer’s residence [30,57]. The consumers in commercial districts receiving management support have a higher possibility of visitors focusing on experience, such as cultural events and space, etc., not the shopping itself.

For visitors in districts with facility and management support, convenience facilities in particular—such as large marts; and infrastructure facilities—should be enhanced. This is because visitors to these areas tend to have high education levels [57] and seem to be satisfied with the various marketing methods, such as events, and provision of coupons. In addition, the number of visits is low possibly because the location is not good [69]. Often only established commercial districts in these urban regions remain because the trading complexes moved to the outskirts owing to the development of new cities, etc. The satisfaction level for infrastructure facility and payment convenience for shopping was higher despite low visit frequency, and therefore, it is assumed that there is some overall satisfaction [65].

Typical visitors in districts with no support are unmarried and cannot spend much time purchasing life necessities. In addition, since they do not shop as a family, they do not consider the pleasure in shopping. Therefore, to these consumers, the convenience and pleasure in shopping are not big problems and they prefer the traditional retail districts, which have no government support and the overall satisfaction level is assumed to be low. This, of course, also illustrates which types of districts tend to attract official support (or not).

4.3.3. Cause and Effect Analysis between Satisfaction Level and Behavior Intention

The results show (

Table 5. Effect relationship between satisfaction level and behavior intention.

Dependent Variable	Independent Variable	Unstandardized		Unstandardized	t-Value
		B	Standard Error	Β
Satisfaction	Constant	1.237	0.125		9.888 *
	Revisit	0.273	0.042	0.336	6.564 *
	recommendation	0.341	0.045	0.389	7.606 *
R2 = 0.442, Adjusted R2 = 0439, F = 157.386

* p < 0.01.

) that the relationship between the satisfaction level and behavior intention had a positive effect on revisit intention and recommendation intention [42,43,47].

5. Conclusions

5.1. Results

This study has empirically examined the effect of government support on the prospects for self-employed retailers in traditional districts. The results show that political support for the self-employed who conduct business inside the traditional retail districts had a partial effect. Especially, for the policy beneficiary, the customer’s satisfaction level in specific political support has a very important meaning. The study analyzes the factors necessary for the continuous growth of traditional retail districts, considering consumption behavior and shop selection attributes. The analysis results are as follows.

First, typical the shop selection attributes were convenience, products, atmosphere, and location—as based on prior studies [20,27,51,52,54,59,60,62]. In addition, for thoroughness, analysis was made in four types of districts: some that received only facility support, some only management support, some with both management and facility support, and finally those that did not receive any support. The consumers who visited these locations were typically young and married with high-income levels, and thought that the shopping-related facilities (parking lots, recreational facilities, and cultural center, etc.) and payment convenience, etc., were very important. In addition, product (price, quality, and variety, etc.) and locational factors were also perceived to be important. It seems that the young married people who visited the locations generally make large expenditures. We are aware of the fact that customers search for a wide variety of high-quality products, irrespective of whether they shop in new retail formats or at a more traditional retail format. Thus, it is necessary for self-employed individual retailers in traditional retail districts to offer products that meet customers’ needs. The commercial districts that received only facility support attracted individuals in their 20s through to their 30s who preferred one-stop shops where infrastructure and convenience facilities are perfect, such as marts and shopping malls. In addition, consumers with low education levels are not interested in customer attraction through various marketing mix techniques and only visit if the facility is clean and pleasant. In addition, married people had a high visit ratio in traditional retail districts which received facility support and these consumers have a high possibility to be visiting with their family for shopping; therefore, they naturally prefer places with pleasant facilities. It can be said that the parking lots, buildings, elevators, and toilets, among others, which also comprise the focus of the governmental support contribute toward enhancing the shopping satisfaction of the consumers. Additionally, improving the store interiors, cleanliness of the toilets, and convenience of the parking lot, among others, also enhances the levels of consumer satisfaction and the overall shopping experience.

Consumers who visited traditional districts that received only management support have relatively high-income levels and make frequent visits for necessities; however, their spending size was not significant statistically. These consumers prioritize convenience as they typically live a short distance away. Thus, they perceive the importance of atmosphere and the shop map as the visit attributes. It shows that the traditional retail districts that received management support, such as discount events and cultural experiences seem to have a positive effect on the number of visits by providing things for the consumers to see and enjoy.

Consumers who earlier shopped at large discount stores or malls, now express a willingness to shop in traditional retail districts that have a more pleasant environment. To benefit, self-employed individuals operating in traditional retail districts must improve their facilities to make consumers feel comfortable while they shop.

Visitors to commercial districts that received both management and facility support showed low satisfaction level about their location since the local urban form had changed with the transfer of major complexes to the outskirts, which seems to have negatively affected visit frequency. However, these visitors have a high level of education, and thus, consider the satisfaction of convenience stores and infrastructure facilities for shopping as important, and prefer various marketing techniques, such as provision of coupons. Thus, despite low visit frequencies, there was a positive effect on overall satisfaction level.

Visitors to commercial districts with no support showed negative coefficient values overall. The lack of infrastructure for shopping convenience, facilities of the commercial district and individual shops, and the satisfaction level of the partial atmosphere was also negative; therefore, the overall satisfaction seemed to be low. Therefore, government support can help sustain the regional commercial districts and the individual self-employed through consumer behavior and the magnification of satisfaction levels.

5.2. Implication and Limitations

Governments naturally support places where accessibility is not good regarding facilities and management support. Thus, the visit frequency of consumers was low, but the satisfaction level for the facility and marketing for shopping conveniences was high, and thus, there was a positive effect on overall satisfaction level. Although accessibility was not good, the results showed the enhancement in overall satisfaction level through improvement in shopping convenience facilities, such as expansion of parking facilities. Thus, government support also enhanced customer satisfaction levels. Although regions with management support showed mediocre effects, the result is important for regions with facility and management support and only owing to the increase in customer satisfaction led by the government support, is it possible to sustain the traditional retail district. On the other hand, overall the value of the coefficient was low in traditional retail districts that do not receive government support.

It is virtually impossible for the self-employed working in non-supported traditional retail districts to implement marketing strategies according to consumer needs owing to their lack of funds and know-how. This implies that it is impossible to confront and fight the aggressive management strategy of new business types, such as large shopping malls, and eventually this will be a very important point in the destruction of self-employed businesses. This is why many feel it is necessary for the government to support the self-employed. However, the government should customize its policy in accordance with the characteristics and situations of the corresponding commercial districts.

This study is different from typical studies that have focused on customer satisfaction for food in new format retail shops and large marts. This study attempted, conversely, to study consumer behavior depending on government support for the traditional retail districts in Korea. The study did not use time series data for quantitative analysis that conventional studies often do, but measured customer satisfaction in line with the effect of policy support over a short period of time. This enabled an empirical survey of customer satisfaction by each government support type. It also empirically verified that these policies have a short-term positive effect on the sustainability of the self-employed.

The theoretical contribution of this study is as follows. This study empirically analyzed the effect of conventional shop selection attributes on customer behavior intentions: in other words, the variables in the government’s policy that had a positive effect on consumers and the effect on revitalization of traditional retail districts. According to the results [51,52,54,60], facility, product, and location variables play an important role in enhancing the satisfaction level of consumers [51,52,54,60]. Although consumers felt satisfied when they visited traditional retail districts, they preferred districts equipped with parking lots, recreational facilities, and cultural facilities: such as they would experience in a large mart or large shopping mall.

In addition, they preferred traditional retail districts equipped with all prices, qualities, and varieties of products, and considered the location factor to be very important. Thus, this study re-verified the findings of prior studies that consumers consider facility, product, and location as important during shopping. The practical distribution of this study is as follows. The problems in the traditional retail districts affect the commercial activities of the self-employed and the regional economy. This has become a major global issue in the USA, Europe, and Japan. England and Japan have even proposed various schemes to revitalize the traditional retail districts through alternatives such as the Business Improvement District (BID) and Town Centre Management (TCM) etc. [70,71]. Similarly, the problems of the traditional retail districts during the city development process are becoming a social issue in Korea. Thus, the Korean government supports the self-employed and the traditional retail districts in the nation by setting up a plan at the ministry level.

Such support enforces the ordinary-person-comes-first policy of the Korean government, which appears as the core social issue, and becomes the political object that the members of congress consider first. Thus, the Korean government enacts and practices related laws, such as the Distribution Industry Development Act and the Special Act for Nurturing Traditional Markets, etc. Further political support will be needed [6] in less developed countries where the self-employed experience difficulties due to the advance of trans-national corporations (TNCs) into the domestic distribution industry. Nations that face degraded traditional retail districts can enact policy development similar to that of Korea. The result of this study can also be utilized in (re)vitalization policies of traditional retail districts in regions such as China and Southeast Asia.

This study nevertheless has the following limitations, regardless of its theoretical and empirical outcome. There is a lack of wider representativeness in the empirical analysis. There are 1439 traditional retail districts across the nation that the Korean government manages and supports; however, among them, only 400 customers in five cities (nine traditional retail districts) could realistically be surveyed. Further studies could expand the target regions for a more even distribution of questionnaires, to enhance the objectivity and reliability of the survey results. In addition, this study focuses on the short term outcome effect of the policy. It is helpful to know how much consumers have spent for at least 5–10 years in the times series in the corresponding traditional retail district. This can be the most accurate method to measure the policy effect.

However, as it is not possible to acquire the time series data, the authors used the measurement of consumer satisfaction of government policy for a short period of time. To enhance the accuracy of results, future studies must examine consumer spending in time series in the corresponding regions. Lastly, the survey included customers who visit the traditional retail districts; however, it was not possible to measure their levels of satisfaction when visiting individual shops regarding the marketing and facility factors. Further studies will be necessary to measure the satisfaction level of consumers who visit the shops of self-employed individuals through a refined study design, as the business type-specific consumer behavior.