In our present society, the Internet, and the popularization of social media in particular, has made it possible for people to write and share various kinds of information at a low cost. One type of information that can be shared in such a way is “word-of-mouth.”
“Word-of-mouth” refers to information on products, services, or companies that is exchanged between consumers. Traditionally, this was information passed on orally from person to person. The extent that information was transferred was limited by the confines of social boundaries, and the effectiveness of word-of-mouth for a particular piece of information rapidly diminished as time passed [1, 2]. However, the recent popularization of social media has enabled information sharing with an unspecified large number of people. As one piece of information can persistently remain as a string of text on the Internet, word-of-mouth can now reach much further than the limits of the conventional word-of-mouth . This new Internet-based word-of-mouth is referred to as online reviews or e-WOM (electronic word-of-mouth).
By having knowledge of a product or service beforehand through word-of-mouth, the consumer can purchase that product or service with reduced risk and have access to the information he or she wanted to know. Thus, economists point out that people pay attention to word-of-mouth information when they are buying something, to eliminate any information asymmetry. In fact, Litvin et al. , Kim et al. , and Gao et al.  all show that word-of-mouth contributes to eliminating the information asymmetry that consumers face. Moreover, especially on the Internet, a variety of word-of-mouth is shared, such as reviews on EC websites (hereinafter “EC”), and information sharing on e-WOM websitesFootnote 1 and via social networking services (hereinafter “SNS”). For example, it is commonplace behavior nowadays to look at reviews on e-WOM and EC websites and to take notice of various kinds of information before deciding to purchase.
For these reasons, economists conduct various kinds of empirical research to study the effect of e-WOM on consumer behavior and to provide insight for marketing, in particular. For example, Chevalier and Mayzlin  showed that reviews of books on EC websites had a significantly positive effect on sales. Duan et al.  conducted a study of movie box-office revenues by considering the endogeneity of reviews. Although the rating of the reviews did not affect box-office revenues, the quantity of review postings strongly affected these revenues. Results similar to those for the movie industry were obtained by Liu , as well. Similar studies showing the quantity of word-of-mouth postings positively affecting expenditures were reported by Park and Lee , Öğüt and Onur Taş , and Babić Rosario et al. . Furthermore, Amblee and Bui  focused on the attributes of review posters and analyzed the effects of both professional reviews and general user reviews on the number of downloads of free software. This study obtained roughly the same results in terms of both significance and importance for both types of reviews. Vermeulen and Seegers  demonstrated that word-of-mouth had a positive effect on people’s recognition and impression of hotels. Also looking at hotels, Ye et al.  conducted an empirical analysis on the effect of online reviews on the number of reservations in Chinese hotels and showed that the number of reservations increased by more than 5% when the review ratings of travelers rose by 10%.
However, although the empirical studies accumulated to date are so numerous, most of the studies took a micro-viewpoint in their analysis, focusing on the effect of reviews on the sales of individual services and products. Such an analytical viewpoint is most important in discovering what effect word-of-mouth has on the sales of specific products and services from the short-term perspective of marketing. However, these studies do not identify whether the customers have been snatched from the same market, or whether the reviews have served to stimulate consumer appetite and boost the level of consumption in the entire market. For example, in the case of word-of-mouth for restaurants, if people already thinking of finding a restaurant meal for 1000 yen on that day looked at word-of-mouth and selected a different restaurant for their 1000-yen meal, at a micro-level, the sales of a specific service rose as a result of a review. On the other hand, the amount of expenditures by the individual does not change, and in terms of the overall market, there has been no change in consumption.
The question arises, therefore, as to whether e-WOM affects the total expenditures of the market in the same way that e-WOM results in a positive effect on sales. There are two conceivable ways in which the effect of e-WOM could influence expenditures. The first is the potential effect of reducing expenditures by enabling consumers to find cheaper prices. For example, a person looking for a 1000-yen meal at a restaurant may, through word-of-mouth, find a meal of the same quality priced at just 800 yen, thereby reducing expenditures. The second is the potential effect of increasing expenditures by way of a person who was not planning to consume, or who was planning to eat more cheaply, looking at word-of-mouth and having his or her appetite for consumption stimulated resulting in a higher priced purchase. For example, a person searching for a 1000-yen meal in a restaurant may, through word-of-mouth, discover a restaurant offering exactly what he or she likes, leading to the purchase of a meal of 1500 yen or an increase in the frequency of eating out, thereby causing a rise in expenditures.
From a short-term marketing perspective, it is not important to identify whether “the customer was snatched from a competitor” or whether there was “a boost in the consumption in the overall market.” However, the following two points provide valid reasons to conduct an empirical analysis to test for the effect of the latter. The first point is that although winning a customer from the competition will lead to a profit for one’s own company in the short term, in the long term, it is important from even a marketing perspective to expand the overall market (broaden the base). The second point is that, from an economic perspective, although the former does not reflect GDP, the latter does affect GDP and, therefore, influences the macro economy. Many countries set a single target for GDP growth as part of their economic policy. In Japan’s case, the aim is to achieve 600-trillion yen GDP for 2020, and to achieve this goal, it will be necessary to boost consumption.
Furthermore, it is also important to analyze the boost in consumption on a macro-level from the perspective of Internet policy. That is because the Internet has a large number of users, and although it can be agreed that it is a useful tool, many online services are used for free, and it has, therefore, been pointed out that the economic contributions of the Internet are small. For example, Brynjolfsson and McAfee  estimated the value of Internet services in the United States to have a monetary base not exceeding 4.2 billion dollars. However, as mentioned earlier, Internet services are not confined only to the Internet. If Internet services were to stimulate consumption in other markets, they could be said to have a larger economic contribution. Presently, many countries around the world are considering a large range of Internet policies to address such issues as intellectual property infringement, protection of personal information, cyber security, innovation, protection of public safety, and freedom of expression. Among the developed countries, although there are voices asserting that freedoms must continue to be fundamentally protected in cyberspace, recently, there has been a strengthening of voices wishing for regulative policies, particularly among EU countries. When such policies are considered, there is a risk that if the value of the Internet is evaluated as diminishing, the effect of these policies on society will also be evaluated as diminishing.
In this study, taking the above into account, we carried out an empirical analysis of the effect of people’s e-WOM on e-WOM and EC websites on boosting consumption on a macro-level. The originality of this paper is twofold: First, we conduct a cross-industry, quantitative analysis using data from the same time period and quantitatively test for differences in the effect of e-WOM by industry. Second, in addition to analyzing consumer behavior using the survey data from more than 30,000 completed questionnaires, we calculate the elasticity coefficient of the boost to consumption on a macro-level using a precise two-step generalized method of moments (GMM), giving adequate consideration to endogeneity.
This paper is organized as follows: Sect. 2 describes the details of the analysis data and shows the correlation between rate of use of e-WOM and expenditures. Section 3 shows the model used in the analysis. Section 4 shows the result of conducting estimation that takes into account the endogeneity using the two-step GMM and clarifies the elasticity coefficient of e-WOM in each field. In Sect. 5, we analyze the boost to consumption provided by the subjective evaluation of people and supplement the results of Sect. 4. In Sect. 6, we state our observations based on the results obtained thus far.