See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/286510475

Does the Rise of the Korean Wave Lead to Cosmetics Export?

Article  in  Journal of Asian Finance Economics and Business · September 2015

DOI: 10.13106/jafeb.2015.vol2.no4.13.

CITATIONS READS

3 4,864

1 author:

Young Seaon Park

Chulalongkorn University
3 PUBLICATIONS   14 CITATIONS   

SEE PROFILE

All content following this page was uploaded by Young Seaon Park on 14 April 2016.

The user has requested enhancement of the downloaded file.

 Young-Seaon Park / Journal of Asian Finance, Economics and Business Vol. 2 No. 4 (2015) 13-20 13

Print ISSN: 2288-4637 / Online ISSN 2288-4645

doi: 10.13106/jafeb.2015.vol2.no4.13.

Does the Rise of the Korean Wave Lead to Cosmetics Export?

2)

Young-Seaon Park*

[Received: June 15, 2015 Revised: September 30, 2015 Accepted: October 15, 2015]

9%. Transportation costs for long distance, tariffs, and other

Abstract non-tariff barriers can explain part of the reasons why the na-
tional border still matters.
As Grossman (1998) speculates, however, the reason why
The purpose of this research is to identify the relation be-
distance matters so much might also be due to the lack of fa-
tween the Korean wave and Korean cosmetics export. Instead
miliarity or cultural difference. Many researchers adopting the
of using UN COMTRADE data as with other researches on the
gravity model include common languages (Boisso & Ferrantino,
similar studies, this paper adopts Google Trends query index
1997; Melitz, 2008), colonial experience (Rose, 2000;
with keyword ‘Korean drama’as a proxy variable for cultural
Eichengreen & Irwin, 1998), and ethnic ties (Rauch & Trindade,
trade. With controlling export determining factors such as GDPs
2002; Combes, Lafourcade, & Mayer, 2005; Wagner, Head, &
of import and export countries, distance, R&D, and FTA, this
Ries, 2002) as proxies for cultural ties between the trading
paper examines whether the Korean wave represented by
partners. The border effect is not restricted to trade in manufac-
Google Trends contributes to the explosive increase of Korean
tured goods. Miroudot, Sauvage and Shepherd (2013) find that
cosmetics export in the recent years. Moreover, this study also
trade costs in services are much higher than those in a goods
investigates the possible effects of the Korean wave on export
sector. Their research data also suggest that trade costs in
that could vary according to the different trade groups by classi-
manufactured goods have fallen considerably over the last dec-
fying import countries into two groups: 74 countries worldwide
ade, but they have remained essentially stable in services
and 9 ASEAN member countries. The results reveal that the
markets. Blum and Goldfarb (2006) show that physical distance
Korean wave indeed leads to cosmetics export to ASEAN coun-
reduces trade even in online products and services that should
tries but show weak relation with cosmetics export to worldwide.
be free of trade costs.
In line with the above reasons, a few of recent studies focus
Keywords: Korean wave, Cosmetics, Cultural proximity,
on the trade diffusion effects of the popularity of Korean cultural
Google Trends, Gravity model.
contents such as TV dramas and music (Park & Choe, 2009;
Kim & Ahn, 2012). The remarkable success of Korean cultural
JEL Classification Codes: F14, F10, D10.
industry is especially evident in Asian countries. According to
the data from Korea Creative Content Agency (KOCCA), during
the 8-year period between 2005 and 2013, the value of Korean
1. Introduction music export increased more than 10 times from 22 million dol-
lars to 227 million dollars. Among the total music export value
Many of research questions raised in the last two decades in 2013, 97.4% went to Asian countries such as Japan, China
on international trade flows have centered on the trade costs and South East Asian countries. The situation of the Korean ex-
derived from the territorial borders. The famous research by port of broadcasting contents is not much different. During the
McCallum (1995) on Canada-US regional trade patterns trig- same 8-year period the value of Korean export of TV contents
gered a flood of studies in search of the true magnitude of the increased from 121 million dollars to 309 million dollars and
border effect. Obstfeld and Rogoff (2000) include the home bias 73% of the total export of TV contents in 2013 went to Asian
in trade as one of the six puzzles in international countries.
macroeconomics. Disdier and Head (2008) examined 1467 dis- The advantage of adopting trade in cultural goods as a proxy
tance effects estimated in 103 papers and found that, on aver- for cultural links between trading partners is that, while other
age, 10% increase in distance lowers bilateral trade by about cultural factors such as common language, colonial experience,
and ethnic ties are time invariant, cultural trade changes within
* Deputy Director. [KOTRA New Delhi #2, 12th Floor, DLF Cyber a short period of time. Thus, the dynamic nature of trade pat-
Terraces, Building 5-A, DLF Cyber City, Phase-lll, Gurgaon-122002, terns can be captured with the use of the cultural trade as a
Haryana, India. Tel: (+91 124) 462-8500, Fax: (+91 124) 462-8501 proxy variable for cultural ties between trading partners. George
E-mail: [email protected]]
14 Young-Seaon Park / Journal of Asian Finance, Economics and Business Vol. 2 No. 4 (2015) 13-20

and Becker (1977) challenge the traditional assumption of the 2. Korean Cosmetics Industry
stability of tastes and preferences through case studies of ad-
diction, habitual behavior, advertising and fashions. In accord- Korean cosmetics industry enjoys fast growth thanks to
ance with the above reasoning it is also possible to assume heightened health interests among the population and increasing
that frequent exposure to foreign cultural contents can shift do- investment in the cosmetics production. According to the data
mestic consumers’ buying behavior (Rauch & Trindade, 2009; from Korea Health Industry Development Institute (KHIDI) shown
Bala & Van Long, 2005; Park, 2014). There are ample cases of in Table 1, Korean domestic production of cosmetics reached
media reports and publications in Korea claiming that the 7280 million dollars in 2013 from 4049 million dollars in 2009
Korean wave greatly contributed to the Korean export in partic- with an average annual growth of 16.0%, far outpacing the
ular and to the Korean economy in general. However, most of average Korean GDP growth rate of 3.2% in the same period.
such reports are based either on pure speculations or on the The remarkable aspect of the industry is that whereas most of
anecdotal researches which lacks rigorous analysis. other industries were severely affected by the financial crisis
This paper attempts to study the trade creation effects of the and recorded no growth at all in 2009, Korean cosmetics mar-
Korean wave, with special consideration of the Korean cos- ket grew by as high as 11.8% in the same year.
metics export. The Korean cosmetics industry is chosen be-
cause of the remarkable performance of the industry in the
midst of the economic recession in recent years. Also, cos-
metics can be categorized as partially substituted products and <Table 1> The Market Size of the Korean Cosmetics Industry
so their consumption is closely connected not only to the prod- (Unit: US$ million, %)
uct price and quality but also to the images the products impart. CAGR
2009 2010 2011 2012 2013 YoY
One of the possible reasons why Korean cultural contents have (2009~2013)
become so popular in Asian countries may be due to the at-
Market
tractive faces of Korean entertainers. Since the concept of beau- 4,336 5,456 5,947 6,231 6,962 11.7 12.8
Size
ty can be shared in the specific regions, the rise of the Korean
wave might explain the explosive export increase of Korean
Production 4,049 5,202 5,763 6,321 7,280 15.1 16.0
cosmetics to Asian countries in the recent years.
Whereas most of the researches on trade creation effects of
the Korea wave (i.e., Park & Choe, 2009; Kim & Ahn, 2012) Export 416 597 805 1,067 1,290 20.8 32.9
use United Nations COMTRADE data for both manufactured and
cultural goods, this paper adopts data extracted from Google Import 702 851 989 978 972 -0.1 8.9
Trends search queries as a proxy measure for trade in cultural
goods. As the Internet is widely used for the consumption of Trade
-286 -254 -184 89 318
audio-visual products, trade of manufactured cultural goods such Balance
as CDs, magnetic tapes, and cinematographic film may fail to Source: Korea Health Industry Development Institute
represent the true scale of cultural trade. * YoY stands for year over year and CAGR stands for compound
Choi and Varian (2012) demonstrate that Google Trends can annual growth rate.
help in predicting the present. Google Trends search queries
have recently been used to estimate the current level of dis-
ease activities such as ‘influenza’, ‘bird flu’ and ‘swine flu’ in the
countries (Ginsberg et al., 2009; Eysenbach, 2006). Askitas and
Zimmermann (2009) demonstrate strong correlations between
keyword searches and unemployment rates using monthly
German data. As the Internet searching activities become part
of daily lives around the world, analyzing search key word quer-
ies may also reveal each person’s cultural affinity to a certain
country.
The paper proceeds as follows: The next section describes
the dynamic growth of the Korean cosmetics industry. The third
section proposes a gravity model approach for empirical analysis Source: Korea Health Industry Development Institute
with data description. The fourth section provides the estimation
results. The fifth section considers the implications of the results <Figure 1> Korean Export of Cosmetics (Unit: US$ thousand)
and concludes.
The export performance of the Korean cosmetics industry in
the recent years is also noteworthy. According to the data from
Korea Health Industry Development Institute (KHIDI), the value
of Korean cosmetics export during 11 year periods (2002 –
 Young-Seaon Park / Journal of Asian Finance, Economics and Business Vol. 2 No. 4 (2015) 13-20 15

2013) increased more than 10 times from 123 million dollars to Secondly, Asian women also share a similar concept of
1,289 million dollars and, as shown in Figure 1, the export val- beauty and skin color. In Korea ‘whitening’ skin product is popu-
ue of Korean cosmetics in 2012 outpaced the import value for lar because white skin is synonymous with beauty. The descrip-
the first time. tion of a woman’s skin as ‘Jade white’has been a common ex-
The major export destinations of Korean cosmetics are Asian pression for praising her beauty in both Korea and China. There
countries such as China, Hong Kong, Japan, Taiwan and South are poems from Tang Dynasty in China that described women
East Asian countries. Figure 2 shows that the top 10 export as ‘Jade white’ and ‘creamy tinted’ (Alexander, 2011). According
destination countries occupy 90% of the total Korean cosmetics to the study by Li et al. (2008), contemporary meanings of
export and among them only The United States and The whiteness in Asia are influenced by Western ideologies as well
Russian Federation are not Asian countries. Among the South as traditional Asian values and beliefs. They concluded that skin
East Asian countries Thailand, Singapore, and Malaysia are in- whitening and lightening products not only promise to fulfill the
cluded in the top 10 export destinations. desire for white and fair skin as a route to higher status, but al-
There might be several explanations why Korean cosmetics so empower women to control their own bodies and alter
are highly preferred among Asian people rather than people in nature.
the other part of the world. Firstly, Korean companies’ efforts on Thirdly, since the advertisement and product brand image are
satisfying domestic consumers paid off not only in the domestic critical aspects of the cosmetics industry, star marketing method
market but also in the overseas market. Korean women are of the Korean companies is thought to have greatly contributed
very conscious of the skin care and so the majority of the prod- to the success of Korean cosmetics in the Asian countries. For
uct development of cosmetics companies thereby focuses on example, Laneige, one of the renowned cosmetics brands of
skin care products. Biotechnology takes an increasingly im- Amore Pacific Corp, made a contract for product advertisement
portant role in protecting skin from aging. Korea’s number one with an actress Song Hye-Kyo in 2008. The actress achieved
cosmetics manufacturer, Amore Pacific Corp, established a bio- stardom in Asia for her appearances in successful TV dramas
science R&D center in 2011 and LG Household and Health such as ‘Autumn in My Heart’ and ‘Full House’ and also fre-
Care, the number two cosmetics company, formed a technology quently appeared in Chinese movies. Also, the number four
partnership with a renowned women’s hospital, Cha Medical player in the Korean cosmetics market, The Face Shop, be-
Center, on stem cell research (Kang, 2012). Taking into account came highly successful partly due to the signing of the actor
the Korean women’s special interest in ingredients of the cos- Bae Yong-joon of the popular TV drama ‘Winter Sonata’ as an
metics, Korean companies have also developed skin care prod- endorser.
ucts made of such natural, organic and herbal medical in-
gredients as green tea, ginseng, sacred lotus root, and bamboo
extract. Sulhwasoo, an exceedingly successful brand by Amore 3. Empirical Strategy and Data
Pacific Corp, is noted for its fragrance of ginseng and the use
of many traditional medicinal herbs and roots. The kin interests
3.1 Empirical Strategy
in clean and fair skin are not an exclusive property of Korean
women but are shared among Asian women. For example, the
This paper adopts the gravity model for the empirical analysis
Japanese brand Shiseido also created a herb-infused face
of the impact of the Korean wave on the export of the Korean
cream developed especially for Chinese skin (Alexander, 2011).
cosmetics utilizing the export panel data from 2005 to 2013.
The gravity model has been widely used for analyzing the pat-
tern of international trade because it has produced some of the
clearest and most robust findings in economics (Leamer &
Levinsohn, 1995). Even though the gravity model has been
used for bilateral trade data since Tinbergen (1962), it became
truly popularized in empirical work after the concept of multi-
lateral resistance (Anderson & Van Wincoop, 2001) was in-
troduced and fixed effects revolution (Feenstra, 2004; Redding &
Venables, 2004) started to capture the multilateral resistance
terms. The combination of its consistency with theory and easy
implementation leads to rapid adoption of the model in empirical
work (Head & Mayer, 2014).
With the inclusion of multilateral resistance, Anderson and
Van Wincoop (2001) derive the following theoretically founded
Source: Korea Health Industry Development Institute gravity equation of international trade:
1−σ
yiyj ⎛ τ ij ⎞
xij = ⎜ ⎟
<Figure 2> Korean Export of Cosmetics per Country as of 2013 yw ⎝ Π iΡj ⎠
where xij denotes exports from country i to j, yi and yj are
16 Young-Seaon Park / Journal of Asian Finance, Economics and Business Vol. 2 No. 4 (2015) 13-20

GDPs of each country, τijis the gross bilateral trade cost, σ > 1 23 times increase, making Thailand the 6th largest importer of
is the elasticity of substitution, and Πi and Ρj represent country Korean cosmetics. To a lesser degree yet still remarkable are
i’s outward and country j’s inward multilateral resistance varia- other South East Asian countries’ import increases. During the
bles, respectively. same period the export of Korean cosmetics to Philippines in-
As shown by Shepherd (2013), taking the natural logarithms creased 11.3 times, to Indonesia 10.7 times and to Malaysia 5
of all variables, the above equation can be transformed as fol- times.
lows:
log Xij = C + Fi + Fj + (1 − σ )[log τ ij ] <Table 2> Korean Export of Cosmetics in the Selected Years (HS
C = − log Y Code 33)
Fi = log Yi − log Π i (Unit: US$ thousand)
Fj = log Yj − log Ρj Country 2005 2007 2009 2011 2013
log τ ij = log Distij + log Cultureij + FTAij
China 71,837 101,354 127,057 217,027 315,788
The first term, C, is equal to world GDP but for estimation
purposes it can be a coefficient because it is constant across Hong Kong 29,804 31,373 45,484 97,328 214,668
all exporters and importers. The next term, Fi, is shorthand for
a full set of exporter fixed effects. Taking the same approach, Japan 39,288 42,092 86,560 139,560 159,032
Fj is a full set of importer fixed effects. logτij is the sum of cost U.S. 34,954 35,795 43,078 66,847 107,117
variables such as distance, cultural proximity, and FTAs be- Taiwan 33,686 30,381 34,840 68,566 94,809
tween country i and country j. Thailand 2,739 4,148 13,974 61,945 77,889
To analyze the impact of the Korean wave on the export of
Vietnam 9,812 11,448 15,944 32,273 42,452
the Korean cosmetics, adopting and modifying the above equa-
tions, this paper estimates the following econometric model: Malaysia 4,783 8,631 13,186 23,895 41,459
lnX kjt = β 0 + β1lnGDPkt + β 2lnGDPjt + β 3lnDistkj + β 4 ln Trends jt + β 5 FTAkjt Singapore 15,311 15,788 16,583 33,107 41,144
+ β 6 ln R & Dkt + γ j + δ t + ε kjt Russia 4,632 5,230 4,765 9,966 25,918
where k indicates the export country (Korea), j is the import Australia 5,084 4,375 5,967 8,908 15,509
country and t is time; ln denotes natural logarithms; Xkjt is the Mongolia 5,485 7,162 6,391 10,484 13,021
flow of Korean cosmetics export to j country in t period; GDPkt
Indonesia 842 837 1,527 9,044 11,664
and GDPjt are GDPs of Korea and import countries respectively;
Distkj is the geographical distance between Korea and import Philippines 612 2,200 2,156 6,945 11,145
countries; Trendsjt is Google Trends search queries with the UK 3,284 3,953 4,359 5,819 10,658
keyword "Korean drama" in the import countries as a proxy vari- Iran 6,661 8,000 8,925 15,093 9,177
able for cultural ties between the two parties. FTAkjt is a dum- UAE 3,007 4,165 5,036 4,968 8,931
my variable denoting the presence of Free Trade Agreements
Canada 4,081 2,839 2,438 3667 7,560
between Korea and import countries. R&Dkt is Korean cos-
metics industry’s investment in research and development. γj Myanmar 274 204 2,167 2447 5,100
and δt are import country and year fixed effects respectively France 2,297 2,131 1,983 5997 5,034
and εkjt is a disturbance term. sub total 278,473 322,106 442,420 823,886 1,218,075
others 16,431 19,001 22,063 36,989 58,902
3.2 Data
total 294,904 341,107 464,483 860,875 1,276,977

The study analyzes Korean export data of cosmetics to 74 Source: Korea Customs Service
countries during the period from 2005 to 2013 collected by
Korea Customs Service (http://www.customs.go.kr). According to This paper adopts Google Trends query data as a proxy variable
Harmonized System Codes (HS Code) cosmetics are in the cat- for cultural affinity between Korea and its trading partners. The
egory of 33 described as ‘Essential oils and resinoids; perfum- trend information of Google searches over time in a particular coun-
ery, cosmetic or toilet preparations’. Even though the two digit try is available by Google Trends (http://www.google.com/trends).
category can be further divided into more specific product lists, Google Trends does not report the raw level of queries for a given
this paper chooses the data of Korean cosmetics export based search term, but instead it reports a query index, meaning the rela-
on HS Code 33. tive search volume in that region at a point in time. The query in-
As is shown in Table 2, during the eight year period Korean dex starts at 0 in January 1, 2004 and the numbers at later dates
export of cosmetics grew more than 4times and the biggest im- indicate the percentage deviation from the query share on January
porters are China, Hong Kong and Japan. The explosive export 1, 2004 (Choi & Varian, 2012).
growth to Thailand is especially noteworthy. In 2005 Thailand’s The search keyword for this paper is ‘Korean drama’ during
import value of Korean cosmetics was mere 2.7 million dollars 2005–2013 periods in Google Trends because this term may
and this figure grew to 61.9 million dollars in 2013, which is a capture the most important aspects of the Korean wave in the
world. TV dramas embody rich cultural traits such as family re-
 Young-Seaon Park / Journal of Asian Finance, Economics and Business Vol. 2 No. 4 (2015) 13-20 17

lationships, social interaction, history, latest trends, food and both public and private companies measured in Korean billion
fashion (Park, 2014). Figure 3 shows the Google Trends popu- won, is from Korean Health Industry Statistics System
larity comparison among Korean, Japanese and Chinese dramas (http://khiss.go.kr).
during 2004 –2014 periods and it is evident that the popularity
of Korean drama relative to those of Japanese and Chinese in- <Table 3> Data Description
creased considerably over the time. Variable Obs. Mean Std. Dev Min. Max.
Export 666 8685.8 30093.9 0 336795
Trends 278 1446.0 1092.2 0 4236
Distance 666 7718.4 3824.8 956 19435
GDPk 666 1084.5 134.8 898 1305
GDPj 666 798.8 1968.4 3 16768
FTA 666 0.2432 0.4293 0 1
R&D 666 168.2 66.7 66 274

4. Estimation Results
Source: Google Trends

<Figure 3> Google Trends by Keywords Comparison: 2004 - 2014

4.1 Baseline Results

The adoption of Google Trends data as a proxy variable for The results of pooled OLS, random effects, fixed effects and
cultural affinity in this paper has several advantages. Firstly, as Poisson pseudo-maximum likelihood estimator (PPML) for the
Google has become a mainstream general search engine since panel data analysis are reported in Table 4. OLS (1) does not
last decade, the accumulated query information may reveal each include fixed effects of import country and year dummies while
individual’s level of tastes and preferences over a particular pe- OLS (2) includes all of them. The results show that OLS (2)
riod in a specific region. As information and communications with adjusted R2 0.9818 is a theory consistent and much im-
technology (ICT) infrastructure has been improved and the num- proved estimator than OLS (1) with adjusted R2 0.6675. The
ber of the Internet users has also increased all over the world overall results from several estimators confirm that, in line with
during the last decade, the term ‘Big data’recently becomes a the gravity model’s assumption, the dependent variable is pos-
household name. Secondly, whereas the available trade data of itively correlated with GDPs of Korea and its trading partners
cultural goods are limited to relatively small number of countries, and negatively correlated with the geographical distance. Korean
the data in Google Trends can cover extensive regions. Thirdly, R&D investment in the cosmetics industry does not show any
to the best knowledge of the author of this paper, there has not significance in general and, in most cases, it shows even neg-
been any attempt to use Google Trends query data as a proxy ative relationships. The result can be interpreted that it takes a
variable for cultural affinity in the field of international trade. relatively long period of time to reap the fruit of R&D
Google Trends query index data is available on the weekly ba- investment. The FTA variable clearly indicates statistical sig-
sis with the exception of a few countries. For the purpose of nificance in most of the estimators except for OLS estimator
simplicity and convenience in matching with other yearly basis without fixed effects.
data, this paper combines the entire weekly index data and Google Trends, as a proxy variable for cultural affinity, shows
transforms them to yearly data. significant correlation in OLS (1), random effects and fixed
Table 3 shows the descriptive statistic in which the total ob- effects. However, the theory consistent fixed effects treated OLS
served number of variables is 666, except for Trends with 278 (2) and PPML do not show any signs of correlation. The result
observed variables. Since each country’s condition of adopting implies that the Korean wave is not much relevant when it
and using the Google search engine is different, the Trends in- comes to Korean export of such partially substituted products as
dex data of early years in some countries are simply not cosmetics to countries throughout the world. The Korean wave,
available. Distance information measured in kilometers between even if more people in the world have recently been recogniz-
Seoul, the Korean capital city, and the capital cities of import ing the phenomenon, is still primarily in the domain of Asian
countries is from timeanddate.com (http://www.timeanddate.com). episodes.
The data of GDPs measured in US billion dollars are from
World Development Indicators. The information of FTA, a dum-
my variable indicating the presence of FTA between Korea and
trading partners, is from Korea Customs Service. The data of
R&D, the total R&D expenditure on cosmetics industry from
18 Young-Seaon Park / Journal of Asian Finance, Economics and Business Vol. 2 No. 4 (2015) 13-20

<Table 4> Dependent Variable: Log Export (OLS, RE, FE), Export <Table 5> Dependent Variable: Log ASEAN Export (OLS, RE, FE),
(PPML) ASEAN Export (PPML)

Variable OLS(1) OLS(2) RE FE PPML Variable OLS(1) OLS(2) RE FE PPML

-1.2602 1.7711** 1.5540*** 1.2976*** 2.8882*** 1.7112 0.6541 1.7126*** 1.4464**

Ln GDPk Ln GDPk
(0.9193) (0.7465) (0.3871) (0.4200) (0.8762) (1.4261) (1.6366) (0.5771) (0.6831)

0.7351*** 1.3908*** 0.8703*** 1.3840*** 0.0885 0.6821*** 1.7977** 1.0305*** 1.3457*** 1.0801
Ln GDPj Ln GDPj
(0.0545) (0.2481) (0.0924) (0.2532) (0.3236) (0.1233) (0.7022) (0.3057) (0.5014) (1.2553)

-2.2835*** -5.5608*** -2.4351*** -3.3946*** -0.1943 -8.8183 -1.1820

Ln Distance Ln Distance
(0.1269) (0.5610) (0.2456) (0.5297) (0.5858) (10.9982) (2.3443)

0.4953*** 0.0013 0.1089*** 0.0790** 0.0393 0.5728*** 0.1423* 0.3123*** 0.3027*** 0.1759***
Ln Trends Ln Trends
(0.0629) (0.0418) (0.0372) (0.0369) (0.0527) (0.1297) (0.0775) (0.0646) (0.0679) (0.0622)

-0.0388 0.3326*** 0.2142* 0.2551** 0.3113*** -0.6208 0.5339** -0.1327 -0.1458 0.7283***
FTA FTA
(0.1726) (0.1251) (0.1142) (0.1174) (0.1136) (0.5333) (0.2094) (0.1939) (0.2028) (0.1290)

-0.8395** -0.2374 0.0450 -0.0542 -0.1472 0.5424 -0.3699 0.3242 0.2043 -0.4149
Ln R&D Ln R&D
(0.3497) (0.2175) (0.1206) (0.1318) (0.2574) (0.6129) (0.4306) (0.2340) (0.2844) (0.8946)

33.0793 36.1557 11.3130 -10.2506 15.1225 -11.3957 70.3290 -2.3021 -11.2496 4.3147
Constants Constants
(5.7707) (7.1118) (3.2624) (2.4797) (6.6898) (9.7555) (99.6936) (19.8651) (3.6903) (3.7664)

Importer Importer
No Yes No No Yes No Yes No No Yes
fixed effects fixed effects

Year Year
No Yes No No Yes No Yes No No Yes
fixed effects fixed effects

Observations 274 274 274 274 277 Observations 57 57 57 57 57

2 2
Adjusted R 0.6675 0.9818 0.6011 0.2194 0.9515 Adjusted R 0.5557 0.9708 0.4805 0.4261 0.9824

*, **, *** denote statistical significance at the 90, 95, and 99 percent *, **, *** denote statistical significance at the 90, 95, and 99 percent
levels, respectively. levels, respectively.

GDPs of ASEAN countries clearly indicate positive and strong

4.2 Regional Differentiation relationship with Korean cosmetics export whereas GDP of
Korea shows mixed significances depending on each estimator.
Whereas above data analysis covers Korean export of cos- The results can be interpreted that when it comes to Korean
metics to 74 countries during 2005-2013 periods, this section export of cosmetics to ASEAN countries, the demand condition
concentrates the analysis on export to ASEAN 9 countries in of the importing countries play more important role than export
the same periods. Among 10 ASEAN member countries Lao country’s factor conditions. The distance variable shows a neg-
DPR is dropped because Google Trends data for the country in ative sign as expected but does not indicate any statistical
the specified period is not available. The primary reason why importance. The reason of this result can be easily drawn; all
ASEAN countries are chosen for the analysis is that the South the analyzed ASEAN countries are geographically concentrated
East Asian region is where the Korean wave has been strikingly in the same region and so distance differentiation is rather a
visible and the usage of the Internet including the social net- difficult task.
work has grown rapidly over the years. The FTA variable shows an inconsistent pattern of signs and
Table 5 shows the estimation results for Korean export of statistical significances depending on each estimator considered.
cosmetics to 9 ASEAN member countries. As in the case of the There might be several explanations why the effects of FTAs do
baseline analysis, Pooled OLS, Pooled OLS with fixed effects, not materialize in Korean export of cosmetics to the ASEAN
random effects, fixed effects, and PPML estimators are market. Firstly, the usage of dummy variable is a rather crude
compared. In the case of PPML estimation, the two variables of way of measurement in this case. When trading partners make
distance and Korean GDP are dropped to ensure that the esti- free trade agreements the tariff rates of some products drop im-
mates exist. mediately but in other product lines the tariff rates decrease
gradually over the years according to the agreed schedule.
Thus, the dummy variable cannot reflect the effects of gradual
decrease of tariff rates over the years. Secondly, each FTA
partner has different trade policies and market conditions. It is
possible that some countries, even before making an FTA, may
 Young-Seaon Park / Journal of Asian Finance, Economics and Business Vol. 2 No. 4 (2015) 13-20 19

have already maintained zero tariff rates in the cosmetics prod- metics companies can put more emphasis on advertising
uct lines. In this case initiating a new FTA does not affect the through social networks such as Facebook, Instagram and Line
trade volume of cosmetics. Therefore, the FTA dummy variable as cultural contents have increasingly been shared this way by
cannot fully represent the market opening status of the specific people around the world.
industry in the import countries and so the result could be in- Further researches could extend the method and findings of
terpreted taking this caution into consideration. the study. In particular, in addition to cosmetics industry covered
The most striking feature of the analysis is the consistently in this paper, other industries can be examined using Google
positive and significant results of the Google Trends variable Trends proxies. Also, besides ‘Korean drama’, other Google
throughout all the estimators. It is evident from the results that Trends query keywords can be tested to find the most appro-
South East Asian people’s interests in Korean cultural contents priate term according to the research themes and purposes.
do affect their buying behavior of Korean cosmetics, as different
from people in other parts of the world.
References
5. Conclusion Alexander, J. (2011). Beauty Products in Asian Market Culture.
BADM 395 Consumer Behavior. West Virginia: Shepherd
This paper uses a new variable to present some of the first University.
evidence on the trade creation effects of the Korean wave on Anderson, J.E., & Van Wincoop, E. (2001). Gravity with gravitas:
the cosmetics industry. The adoption of the Google Trends a solution to the border puzzle. NBER Working Paper
search query as a proxy variable for cultural affinity is a noble No. 8079.
approach. As increasing number of people around the world Askitas, N., & Zimmermann, K.F. (2009). Google econometrics
uses the Internet for searching information or entertainment pur- and unemployment forecasting. German Council for
poses, the analysis of search queries of the most popular web- Social and Economic Data (RatSWD) Research Notes,
site, Google, in connection with cultural affinity seems an appro- 41.
priate and timely task in the field of international trade. The Bala, V., & Van Long, N. (2005). International trade and cultural
great advantage of using Google trends as a proxy variable is diversity with preference selection. European Journal of
that the data reflect the changing patterns of users’ searching Political Economy, 21(1), 143-162.
behavior while other proxy variables in general are time Blum, B.S., & Goldfarb, A. (2006). Does the internet defy the
invariant. law of gravity?. Journal of International Economics,
Using the gravity model, this paper firstly analyzed the data 70(2), 384-405.
of the Korean cosmetics export to 74 countries around the Boisso, D., & Ferrantino, M. (1997). Economic distance, cultural
world. Whereas economic sizes of Korea and import countries, distance, and openness in international trade: Empirical
geographical distance, and the presence of free trade agree- puzzles. Journal of Economic Integration, 12(4), 456-484.
ments show statistical significance, Google Trends search query Choi, H., & Varian, H. (2012). Predicting the present with goo-
of ‘Korean drama’does not reveal a clear correlation with Korean gle trends. Economic Record, 88(s1), 2-9.
cosmetics export. When the export destination countries were Combes, P-P., Lafourcade, M., & Mayer, T. (2005). The
narrowed to 9 ASEAN member countries, the results show trade-creating effects of business and social networks:
strong evidence that the Korean wave indeed leads to the ex- evidence from France. Journal of International Economics,
port of Korean cosmetics. 66(1), 1-29.
There can be several implications drawn from the study. Disdier, A-C., & Head, K. (2008). The puzzling persistence of
Firstly, besides the traditional proxy variables used in the gravity the distance effect on bilateral trade. The Review of
model for cultural affinity such as common languages, colonial Economics and Statistics, 90(1), 37-48.
experience, and ethnic ties, Google trends query data can be- Eichengreen, B.J., & Irwin, D.A. (1998). The role of history in
come another important proxy variable. Secondly, the Korean bilateral trade flows: The regionalization of the world
wave’s contribution to the export of Korean cosmetics is evident economy. Chicago: University of Chicago Press, 33-62.
in Asian countries but not clear in other part of the world in Eysenbach, G. (2006). Infodemiology: tracking flu-related search-
general. Since the Korea wave has risen high mainly in Asian es on the web for syndromic surveillance. AMIA Annual
countries and cosmetics export to the region also increased dra- Symposium Proceedings Volume 2006. American Medical
matically, the finding is consistent with the general assumption. Informatics Association, 244-248.
Thirdly, Korean free trade agreements with many countries paid Feenstra, R.C. (2004). Advanced International Trade: Theory and
off in the cosmetics industry. Whereas the Korean wave greatly Evidence. New Jersey: Princeton University Press.
contributed to increasing consumption in the Asian market, FTA George, J.S., & Becker, G.S. (1977). De gustibus non est
is a more important factor in exporting cosmetics to the world disputandum. The American Economic Review, 67(2),
market. 76-90.
Lastly, in addition to the star marketing method, Korean cos- Ginsberg, J., Mohebbi, M.H., Patel, R.S., Brammer, L.,
20 Young-Seaon Park / Journal of Asian Finance, Economics and Business Vol. 2 No. 4 (2015) 13-20

Smolinski, M.S., Brilliant, L. (2009). Detecting influenza Rauch, J.E., & Trindade, V. (2009). Neckties in the tropics: a
epidemics using search engine query data. Nature, 457, model of international trade and cultural diversity.
1012-1014. doi: 10.1038/nature07634. Canadian Journal of Economics, 42(3), 809-843.
Grossman, G.M. (1998). Comment on Deardoff, in Jeffrey A. F. Redding, S., & Venables, A.J. (2004). Economic Geography and
The Regionalization of the World Economy, NBER International Inequality. Journal of International
Project Report. Chicago: The University of Chicago Economics, 62(2004), 53-82.
Press. pp. 29-31. Rose, A.K. (2000). One money, one market: the effect of com-
Head, K., & Mayer, T. (2014). Gravity Equations: Workhorse, mon currencies on trade. Economic Policy, 15(30), 8-45.
Toolkit, and Cookbook, Chapter for the Handbook of Shepherd, B. (2013). The gravity model of international trade: A
International Economics, Vol. 4, 131-195. user guide. Bangkok: United Nations publication.
Kang, C-K. (2012). The Cosmetics Industry Adopts a High-Tech Tinbergen, J. (1962). Shaping the world economy; suggestions
Makeover. Seri Quarterly, 5(3), 99-103. for an international economic policy. New York:
Kim, J-G., & Ahn, S.Y. (2012). An empirical study on effects of Twentieth Century Fund.
Korea’s cultural exports. Journal of Korea Trade, 16(2), Wagner, D., Head, K., & Ries, J. (2002). Immigration and the
25-48. Trade of Provinces. Scottish Journal of Political
Korea Creative Content Agency (2015). Report on content in- Economy, 49(5), 507-525.
dustry statistics. Retrieved March 5, 2015, from
http://www.kocca.kr
Korea Customs Service (2015). Korea cosmetics export Appendix: List of Countries
statistics. Retrieved April 16, 2015 from
http://www.customs.go.kr/kcsweb/user.tdf?a=user.newTrade Argentina Hong Kong Pakistan
statistics.NewTradestatisticsApp&c=1001&mc=STATS_INQ Australia Hungary Philippines
U_TRADE_010
Korea Health Industry Development Institute (2015). Korea Austria India Poland
cosmetics export statistics. Retrieved April 15, 2015 from Azerbaijan Indonesia Portugal
http://khiss.go.kr/treeManager.do?task=treeFrm&vw_cd=MT
Bahrain Iran Puerto Rico
_ ITITLE&list_id=358_C &h
txt_code=12536971932186930964991651009680 Bangladesh Ireland Qatar
Leamer, E. E., & Levinsohn, J. (1995). International trade theo-
Belgium Israel Romania
ry: the evidence. Handbook of International Economics,
3, 1339-1394. Brazil Italy Russia
Li, E.P.H., Min, H.J., Belk, R.W., Kimura, J., Bahl, S. (2008). Brunei Japan Saudi Arabia
Skin lightening and beauty in four Asian cultures.
Advances in Consumer Research, 35, 444-449. Bulgaria Jordan Singapore
McCallum, J. (1995). National Borders Matter: Canada-U.S. Cambodia Kazakhstan Slovak Republic
Regional Trade Patterns. American Economic Review,
Canada Kuwait South Africa
85(3), 615-623.
Melitz, J. (2008). Language and foreign trade. European Chile Lebanon Spain
Economic Review, 52(4), 667-699.
China Lithuania Sri Lanka
Miroudot, S., Sauvage, J., Shepherd, B. (2013). Measuring the
cost of international trade in services. World Trade Colombia Macao Sweden
Review, 12(4), 719-735. Cyprus Malaysia Switzerland
Obstfeld, M. & Rogoff, K. (2000). The Six Major Puzzles in
International Macroeconomics. Is There a Common Czech Republic Mexico Taiwan
Cause?. In B.S. Bernanke and K. Rogoff (Eds.), NBER: Denmark Mongolia Thailand
Macroeconomics Annual 2000 (pp. 339-390). Cambridge,
MA: MIT Press. Ecuador Morocco Turkey
Park, S., & Choe, J. I. (2009). The Trade Creation Effects of Egypt Myanmar UAE
Hallyu. Bank of Korea Economic Research Institute:
Estonia Nepal UK
Economics Analysis, 15(1), 73-96.
Park, Y.S. (2014). Trade in Cultural Goods: A Case of the Finland Netherlands Ukraine
Korean Wave in Asia. Journal of East Asian Economic France New Zealand United States
Integration, 18(1), 83-107.
Rauch, J.E., & Trindade, V. (2002). Ethnic Chinese networks in Germany Nigeria Vietnam
international trade. Review of Economics and Statistics, Greece Norway
84(1), 116-130.

View publication stats