The Effect of Screen Quota System on the Self-Sufficiency Ratio in Domestic
Film Markets


Byoungkwan Lee*
Michigan State University
Hyuhn-Suhck Bae
Yeungnam University


Paper submitted to the annual convention of the Association for Education
in Journalism and Mass Communication, July 30-Aug. 2, 2003, in Kansas City, MO


Author note: Byoungkwan Lee (M.A., 2001, Michigan State University) is a
doctoral student in the Mass Media Ph.D. program at Michigan State
University. E-mail: [log in to unmask] Hyuhn-Suhck Bae (Ph.D., 1998,
Michigan State University) is an associate professor in the department of
Media and Communication at Yeungnam University in Korea. E-mail:
[log in to unmask] The authors thank Dr. Stephen Lacy for his
invaluable comments on this paper.


* Please send all correspondence to the first author at The Department of
Advertising, Michigan State University, East Lansing, MI
48824-1212.  E-mail: [log in to unmask]; Phone: (517) 355- 7581.

 The Effect of Screen Quota System on the Self-Sufficiency Ratio in
Domestic Film Markets

Abstract
This study examined the impact of the screen quota system and other
determinants on the self-sufficiency ratio. The regression models show that
the quota system is a weak predictor of the self-sufficiency ratio,
suggesting the system may not be an effective mechanism to limit foreign
cinemas screened in its own territory. GDP, box office revenue, and
production investment were found to be strong predictors; cultural discount
is a moderate; and English-speaking is a weak determinant.


 The Effect of Screen Quota System on the Self-Sufficiency Ratio in
Domestic Film Markets

The powerful influence of the U.S. film distribution on the international
film market makes many other countries regard it as a serious threat to
their cultural sovereignty and domestic film industry. The dominance of the
U.S. film in the international film market has been maintained by several
factors such as the largest domestic market, production in English,
characteristics of its industry and market, and the Hollywood system .
Various economic and institutional barriers have been used by different
countries as mechanisms to protect domestic markets from U.S. films. These
means include "funding state corporations, direct subsidy to film
production, content regulation, quotas, tax concessions, entry barriers,
licensing conditions, and international co-productions treaties" .
The "screen quota" system is a means of governmental support and protection
for the domestic film industry, obliging exhibitors to set aside a minimum
number of screenings of domestic movies. The General Agreement on Tariffs
and Trade (GATT) and the Deregulation Agreement of Organization for
Economic Cooperation and Development (OECD) allow the screen quota system
in international trade (CDMI, 2000). Despite U.S. pressure for removal of
the system, it still operates in dozens of countries around the world
including France, Hungary, Italy, Korea, Spain, etc. France requires
cinemas to screen domestic films five weeks per quarter, while Italy
requires 100 days a year and Korea at least 106 days a year to screen
domestic films. Spain sets the screen quota requiring exhibitors to show
one day of EU films for every three days of non-EU films (Terra Media, 2002).
The purpose of this study is to explore the economic impact of the screen
quota system on the self-sufficiency ratio in domestic film markets.
Identifying other determinants of the self-sufficiency ratio in
international markets, this study investigated whether the screen quota
system functions as an institutional barrier to protect the domestic film
industry. The "self-sufficiency ratio", defined as the proportion of
domestic film's share in gross box office revenues, is affected by various
factors . While empirical studies on the antecedents that can influence the
film market are abundant (e.g., ; ; ; ; ; , very few have studied the
impact of regulatory factors such as the screen quota system on the film
industry.[1] As such, it is expected that this study might provide
empirical evidence on whether the screen quota system has functioned well
as an institutional mechanism to protect domestic film markets from the
dominance of the U.S. films.

Screen Quota System and Self-Sufficiency Ratio
The screen quota system is a governmental regulation that makes it
compulsory for movie theaters to screen the feature films of national
origin for a specified period of time. In general, the screen quota system
is regarded as a "cultural exception" in various international agreements
even though it goes against the principle of national treatment, which
prohibits discrimination between locally produced and imported goods . This
reasoning is based on the assumption that films are both cultural products
and content products that differ from many other goods or products.
In the economic aspect of media products, it is also considered that films
are different from general goods and traditional services because films are
content products. While the first-copy production cost is very high, costs
after production are relatively low as they can be easily copied (Litman,
2000). A combination of the nature of motion pictures as a cultural product
and the dominance of the U.S. films in the international film trade may
threaten other countries' cultural identity and domestic film industry. The
reason why U.S. motion picture distributors have been so powerful depends
primarily on the relative size and strength of the U.S. market compared to
those of other countries. U.S. motion picture and television producers can
largely make up for their production costs from the domestic market alone,
and given the public goods nature of the mass media and the fact that the
greatest expense is the first-copy production cost, distribution prices to
foreign countries only need cover the incremental expenses (Litman, 2000).
As such, it is inevitable that a film's profitability depends on an
exclusive distribution. The Windows Model for films also explains the
worldwide domination of the U.S. films. The Windows Model, according to
Litman (2000), refers to the exhibition sequence of films to maximize the
present value of profits across the many new exhibition windows such as
theater, VCR, pay-per-view, pay cable, and broadcasting television. As
Litman (1998) addressed, the prevalence of an elaborate and dynamic
windowing process in the U.S. allows additional funds for up-front
investment, and, in turn, this added budgetary flexibility is a key factor
in intensifying the strong market power of the U.S. motion picture
distributors in the international film trade.
The screen quota system is recently the hottest issue in the Korean film
industry. The current screen quota system rules that all theaters in Korea
are required to show domestic products at least 106 days a year. While the
U.S. government has criticized the system as being counter to free trade,
the Korean industry groups called for a "cultural exception" to allow for
protection of local cinema (Paquet, 2002). The Coalition for Cultural
Diversity in Moving Images (CDMI), a group seeking to enforce the screen
quota system, has argued that the screen quota system has functioned as an
appropriate institutional mechanism to protect the Korean film industry
from the dominance of U.S. film distribution since 1993 .[2]
As figure 1 shows, it seems that domestic films' market share in Korea has
increased since 1993 even though the amount of imported U.S. films has
continuously increased. Two years ago, the government decided to defer
initiating any talks about abolishing the screen quota system until
domestic films' market share accounts for more than 40 percent of the film
market inside Korea . Last year, however, domestic market share accounted
for nearly 50 percent of the market (Korean Film Commission, 2002), and in
January, 2002, the Korean government announced that the bilateral
investment treaty (BIT) between the U.S. and Korea will be settled in the
first half of this year. According to a Korean government officer, "the
government has not yet reviewed the abolishment of the screen quota system.
But they are leaning towards reducing the number of days that Korean films
play at theaters" . Thus, the current success of Korean films convinced the
Korean government to liberalize the screen quota system, and the government
is proposing again to sign the BIT with the U.S. However, Korean filmmakers
and distributors argued that even though their market share has exceeded 45
percent, this is simply an example of statistical prosperity because most
movie attendance concentrated on only two or three movies. Finally, it
seems that conflict between the Korean government looking to abolish or
change the screen quota system and CDMI seeking to defend the system is
unavoidable.
Figure 1 about here
Mexico illustrates support for the argument that the screen quota system is
necessary to protect the domestic film industry. The screen quota system of
Mexico has been discontinued since January of 1993 as a reaction to the
dramatic decrease of the number of domestic film productions, from 100
products a year up until 1990 to 30-40 products a year after that . Some
defenders of the screen quota system argue that since the abolition of the
screen quota system in 1993 when the Mexican government signed the BIT with
the U.S. following the peso crisis, the Mexican film industry collapsed
during the following three years .
Accordingly, at the heart of the screen quota system controversy is the
question of whether the screen quota system affects the domestic film
industry. The first hypothesis examines the relationship between the screen
quota system and the self-sufficiency ratio. The self-sufficiency ratio is
calculated from the formula, D/(D + F), where D is domestic films' revenues
in a given country and F is foreign films' revenues in the same country.
This study predicted that the screen quota system would increase the
self-sufficiency ratio:
Hypothesis 1: The presence of the screen quota system will be positively
associated with
the self-sufficiency ratio.

Determinants of the Self-Sufficiency Ratio
Market Size and Production Investment. Economic factors that affect the
self-sufficiency ratio can be explained by the Industrial Organization (IO)
model. The IO model posits that market structure affects conduct that
subsequently influences media performance. Market structure is determined
by market size that is positively correlated with production budgets
(Wildman, 1995). As the growth of market size makes the market highly
competitive, increased production investment would be required. Market
competition is positively related to the diversity of film content (Litman,
1992). Thus, as market competition leads to the increase of production
investment, it also affects both diversity and production quality. Both
diversity and production quality in the form of films by the increase of
production investment shift the demand curve for domestic films upward (Oh,
2001). This shift consequently leads to an increase in the self-sufficiency
ratio since it increases overall box office revenue and substitutes
domestic films for foreign films . Distinguishing the market size as the
level of GDP as a potential market size and the box office revenue as a
realized market size, Oh (2001) tested and supported a hypothesis that the
market size is positively related to the self-sufficiency ratio.
This study, however, separated "direct" production investment from
"potential" production investment (market size). Even though the production
investment is a direct antecedent variable of the self-sufficiency ratio,
which is likely to mediate the effect of market size on it, the
relationship between the production investment and the self-sufficiency
ratio was not tested in Oh's study (2001). As such, the following
hypotheses were suggested to examine the relationships between the market
size and the self-sufficiency ratio and production investment and the
self-sufficiency ratio:
Hypothesis 2-1: The level of GDP will be positively associated with the
self-sufficiency
ratio.
Hypothesis 2-2: The level of box office revenue will be positively
associated with the
self-sufficiency ratio.
Hypothesis 2-3: The level of each market size, thus, GDP and box office
revenue, will be
positively associated with the film production investment.
Hypothesis 2-4: The level of the investment in film productions will be
positively
associated with the self-sufficiency ratio.

Cultural Distance and Language. Cultural/linguistic factors also affect the
self-sufficiency ratio. The market size itself, as Hoskins et al. (1997)
argues, does not explain the dominance of the U.S. in the international
film trade. Unlike "culturally neutral" products such as automobiles or
camcorders, media products, especially, movie or television programs,
inevitably involve a "cultural discount" in the international trade. As the
distance between the cultures of two countries is greater, viewers will
watch less foreign programs or films than domestic ones of the same type
and quality, and the value will be less to the foreign broadcaster or
distributor. Cultural discount refers to the percentage reduction in value
of the foreign television program or film (Hoskins & Mirus, 1988). The
interaction of the cultural discount and market size can give an advantage
to the U.S. film distribution on the international film market (Hoskins et
al., 1997). For example, the less the cultural distance between two
countries such as the U.S. and Canada, the greater the competitive
advantage of the U.S. film. In contrast, if the cultural distance between
two countries such as the U.S. and Japan is great, the advantage is
relatively low.
Oh (2001) examined the effect of cultural distance on the self-sufficiency
ratio based on four cultural dimensions identified by Hofstede (1991),
which are power distance, individualism, masculinity, and uncertainty
avoidance. The results of Oh's study (2001) partially supported the
cultural distance hypotheses. He found that while the cultural distance of
power distance between a country and the U.S. was a significant predictor,
other dimensions were mostly not significant.
This study also examines the effect of cultural distance on the
self-sufficiency ratio. In the initial analysis, however, this study
identified a possibility of multicollinearity between the screen quota
variable and each dimension of cultural distance. The value of tolerance
was only less than .15, and the value of Variance Inflation Factor (VIF)
was relatively high (VIF = 6.57). Therefore, in order to investigate the
impact of cultural distance, this study employed a composite index that was
formed based on "the deviation along each of the four cultural dimensions
of each country from the U.S. ranking" . The third hypothesis was suggested
as follows:
Hypothesis 3: Cultural distance between a country and the U.S. will be
positively related to the self-sufficiency ratio.

Language is also an important component of the cultural discount . Several
studies have explored the role of language in international films and
television programs ;;. Hoskins and his colleagues (1997) found that
English-speaking countries were likely to pay more for U.S. television
programs than non-English-speaking countries. On the contrary, Dupagne and
Waterman (1998) unexpectedly identified English fluency as a significant
negative predictor of U.S. television fiction imports. In the context of
the film industry, the English-speaking variable was not found to be a
significant negative predictor of the self-sufficiency ratio . The final
hypothesis of the present study expected that the linguistic factor would
negatively affect the self-sufficiency ratio:
Hypothesis 4: An English-speaking country, excluding the U.S., will be
negatively related to the self-sufficiency ratio.

METHOD
Model Building and Predictions
Due to the expected multicollinearity among two market size variables and
product investment, three different multiple regression models were
employed in order to test the hypotheses. While model 1 includes GDP as a
"potential" market size variable in the equation, model 2 and 3 enter box
office revenue (BOX) as a "realized" market size variable and film
production investment (PI) into the equation, respectively.
All three models commonly use the presence of the screen quota system (SQ),
cultural distance (CD), and English-speaking variable (ES) as the
predictors of the self-sufficiency ratio (SSR). SQ and ES were coded as
dummy variables (yes = 1 and no = 0). All beta coefficients of predictors
but ES are expected to be positive. The models are as follows:

Model 1: SSR = _0 + _1SQ + _2GDP + _3CD + _4ES + e; _1>0, _2>0, _3>0, _4<0
Model 2: SSR = _0 + _1SQ + _2BOX + _3CD + _4ES + e; _1>0, _2>0, _3>0, _4<0
Model 3: SSR = _0 + _1SQ + _2PI + _3CD + _4ES + e; _1>0, _2>0, _3>0, _4<0

Data Collection
Cross-sectional data from 1997 was collected from 20 countries by using a
convenience sampling method.[3] 1997 data was the most recent ones that
were available in all the 20 countries at the time of study. The list of
countries that utilize the screen quota system was obtained from CDMI and
numbered eight--Brazil, France, Greece, Hong Kong, Italy, Japan, South
Korea and Spain.
The data for the self-sufficiency ratio, the investment in film
productions, and box office revenues were obtained from various issues of
Screen Digest (September 1996, June 1998, September 1998, and June 2000)
and European Cinema Yearbooks (1999, 2000, and 2001). The GDP data were
collected from the International Financial Statistics Yearbook (2000).
One-year previous data for box office revenue and GDP were collected. Using
previous data for these variables is more relevant than using current data
because they function as a signal for determining the investment in film
production (Oh, 2001). The unit of currency for all measures was U.S. dollars.
Data for calculating the cultural distance index between the U.S. and a
given country was collected from Hofstede (1991) and calculated using the
following equation by Kogut and Singh (1988):
     4

i=1
CDj = _{(Iij – Iiu)2 / Vi} / 4

where Iij is the index for cultural dimension i and country j, Iiu stands
for the index for cultural dimension i and the U.S., Vi  is the variance of
the index of dimension i, and CDj indicates the cultural distance between
country j and the U.S. Among the sample, four English-speaking countries
were identified: Australia, Canada, Hong Kong, and the United Kingdom.

RESULTS
This study used least squares multiple regression to test hypotheses. All
independent variables were entered simultaneously into three regression
equations. Overall, as shown in Table 1, each model was significant
accounting for 49 percent to 73 percent of the variance in the
self-sufficiency ratio. As the results of diagnostic tests and the
examination of residuals, no outlier or multicollinearity problems were
identified in the models.
Table 1 about here

Hypothesis 1 predicted a positive relationship between the screen quota
system and the self-sufficiency ratio. All three models, although not
statistically significant, show a positive relationship between these two
variables (Table 1).[4] In model 1 and 3, the regression coefficients of
screen quota are relatively high, so the screen quota system was associated
with an increase of about 4.0 percent for model 1 and about 3.9 percent for
model 3 in the self-sufficiency scale. In model 2, an increase in screen
quota of one point was associated with an increase of only .84 points in
the self-sufficiency ratio. However, the beta coefficients of all three
models (.04 to .17) show that the screen quota system is a weak or very
weak predictor compared to other determinants.  Thus, hypothesis 1 is
weakly supported.
Hypothesis 2-1 and 2-2 predicted that market size affects the
self-sufficiency ratio. Table 2 shows the results of bivariate correlation
analysis that Ln GDP was highly related to the self-sufficiency ratio (r =
.62). Model 1 and 2 show that Ln GDP and Ln box office revenue were related
to an increase of almost 8.5 percent and 9.5 percent in the
self-sufficiency scale, respectively. The beta coefficients of Ln box
office revenue and Ln GDP were very high, and the coefficient of the former
(b = .80) was greater than that of the latter (b = .72). These results were
consistent with the findings of Oh (2001). As such, the results suggest
that Ln GDP and Ln box office revenue are strong predictors of the
self-sufficiency ratio. Thus, hypothesis 2-1 and 2-2 are supported.
Table 2 about here

The relationship between market size and product investment was also
examined. As shown in Table 2, a very strong positive association between
Ln GDP and Ln investment (r = .85) and between Ln box office and Ln
investment (r = .90) was found. Thus, hypothesis 2-3 is supported.
The results also showed that Ln investment was strongly correlated with the
self-sufficiency ratio (r = .73). The third regression model shows that Ln
investment predicted the increase of the self-sufficiency ratio by around
7.0 percent. In model 3, the beta coefficient of Ln investment indicates
that it is the strongest predictor of the self-sufficiency ratio.
Hypothesis 2-4 that the level of the investment in film productions will be
positively associated with the self-sufficiency ratio was supported.
Therefore, these findings can support the assumption about a causal
relationship among three variables, which suggests the increase of market
size causes more investment that leads to the self-sufficiency ratio.
Hypothesis 3 was proposed to examine the impact of cultural distance on the
self-sufficiency ratio. Even though the results of a bivariate correlation
analysis shows a weak relationship between them (r = .20), regression
models supported the hypothesis that cultural distance would be a moderate
positive predictor of the self-sufficiency ratio (b  = .41 for model 1, .45
for model 2, and .33 for model 3). Model 3 shows that cultural distance was
associated with an increase of almost 3.8 percent in the self-sufficiency
scale, and it barely missed statistical significance even with the small
sample size (b = .33,  p = .067). As such, this suggests that cultural
distance remained a moderate predictor for the self-sufficiency ratio even
when controlling for other independent variables.  Hypothesis 3 is supported.
Finally, hypothesis 4 predicted that an English-speaking country, excluding
the U.S., would be negatively related to the self-sufficiency ratio. Both
bivariate correlation (r = -.12) and the standardized in model 2 and 3 (b =
-.08 and -.17) shows that English-speaking is a weak, negative predictor of
the self-sufficiency ratio. That is, the self-sufficiency ratio is likely
to be higher in the non-English speaking countries. But model 1 shows that
the language factor is a positive but very weak predictor of the
self-sufficiency ratio. Therefore, hypothesis 4 is partially supported.
Overall, the three models show that among the variables in the equation,
market size (GDP and box office revenue) and product investment are strong
predictors; cultural distance is a moderate predictor; and screen quota and
English-speaking are weak predictors of the self-sufficiency ratio. Among
the three regression models, model 3 emerged as the best model, accounting
for more than 70 percent of the variance (adjusted R2 = .73).

DISCUSSION
This study attempted to examine the effect of the screen quota system on
the self-sufficiency ratio and identify other determinants that explain the
variance of the self-sufficiency ratio in a country. This study included
the screen quota system and production investment in the regression
equation, not included in the Oh study (2001).
First of all, this study shows that the screen quota system is not a strong
predictor of the self-sufficiency ratio. This suggests that the system
should not be an effective mechanism to limit the number of foreign cinemas
screened in its own territory, although the countries have employed the
system due to low self-sufficiency ratio. Actually, the screen quota
requirement is often ignored by local exhibitors who seek to make more
profits in some countries. For example, in Italy, theater owners did not
comply with a 1965 law mandating a 25-day-per-quarter screen quota (U.S.
Department of State, 1994). Throughout the 1990s such requirement was
regularly ignored by Korean exhibitors, screening local cinemas for up to
50 days/year less than the level stipulated by the law (Paquet, 2002).
Thus, both defenders and opponents of the screen quota system need to pay
attention to the findings of the present study: the screen quota system is
not a protective measure for domestic film markets, as well failing to
provide a barrier to free international trade of films.
Consistent with previous studies, this study found that market size was
found as the strongest predictor of the self-sufficiency ratio. Both GDP as
a potential market size and box office revenue as a realized market size
strongly determine the self-sufficiency ratio. That is, countries whose GDP
is lower and whose market size is smaller tend to depend on imported
cinemas more than countries with higher GDP and a larger market. Also, it
seems valid that as an indigent economic condition brings about the lack of
investment for domestic film, the self-sufficiency ratio gets lower. The
findings on the relationship between product investment and the
self-sufficiency ratio strongly support this reasoning. Considering the
very strong positive correlation between box office revenue and product
investment, this study found an empirical cue to be able to explain a
causal relationship in which market size affects the product investment
that leads to the self-sufficiency ratio. As shown Figure 2, this leads to
a simple path model:
Figure 2 about here

In order to test this model, a simple path analysis was conducted using
PACKAGE (Hunter & Gerbing, 1982). The results of analysis for the missing
links between Ln box office revenue and the self-sufficiency ratio (error =
-.04,  z = -.20,  p > .05) indicated that the data were consistent with the
model suggested above. The overall fit test also showed that the data were
consistent with the model (c2 (1) = .04,  p > .05). Although this was a
preliminary test with few single variables and a small sample size, the
results imply a causal order among these three variables.
The impact of cultural discount on the self-sufficiency ratio was also
found. The findings show that the increase of cultural distance decreases
the demand for foreign films, especially, U.S. films. As Oh (2001)
addressed, this finding could explain why some countries show a relatively
high self-sufficiency ratio even though they have a small domestic market
size. Thus, it seems that the high cultural discount between two nations
might diminish the market advantage (or market power) of the country that
has a great market size in the international film market.
Finally, the three models of this study indicate that the effect of the
language factor could be weak-negative, very weak-negative or very
weak-positive, respectively. Thus, this suggests that English-speaking
countries with a market power such as the U.S. should have little or no
linguistic handicap in the international film trade. In relation to this
result, Oh (2001) addressed two potential effects. First, big-budget U.S.
films have little linguistic handicap in other English speaking countries.
Second, English speaking countries face less of a handicap in exporting
their films to the wealthiest English linguistic markets including the U.S.
All the findings of this study considered, since market size, cultural
distance and language are virtually unable to be controlled, a given
country has only two options to improve the self-sufficiency ratio in its
own motion picture market: one is to increase product investments; and the
other is to operate the screen quota system effectively with other
protective mechanisms. While the former is an active incentive, the latter
acts as passive protection. Korea's case shows that production investment
and good domestic movies attract local cinema fans. Four elements--talented
filmmakers, remarkable technological innovation, effective marketing
strategies, and virtual abolition of government censorship--have worked for
a renaissance of the Korean movie industry since 2000. The market share of
local movies skyrocketed to nearly 50 percent and the number of Korean
films exported swelled to 38 in 2000 from 15 in 1995 (Park, 2002).
However, because production investment is also affected by market size, it
would be required to keep the balance between the two options. To the
countries with small market sizes and low GDP, the screen quota system will
be not an ultimate method of preserving their film markets, but the least
way to keep their markets from foreign major film producers and distributors.
A reality check supports the above argument. In Asia, Taiwan has an import
quota, which limits the amount of foreign cinemas entering the country. But
the quota has been reduced every year due to Taiwan's annual trade
negotiation with the U.S. Taiwanese film directors and scholars have been
concerned that the glut of Hollywood films on Taiwan's screens threatens
the survival of its local film industry if Taiwan joins the World Trade
Organization (WTO) (Yu, 1999). The case of Mexico is more dramatic. Due to
1994's North American Free Trade Agreement (NAFTA), the Mexican government
reduced a screen quota from 30 percent to 10 percent dedicated to domestic
films. In 1998, Mexico only produced 3 films (CDMI, 2000).
For the previous 3-5 years, only a few countries have been successful in
capturing a market share of more than 25 percent for their domestic films.
France with its broadcasting quotas and financial subsidies, Japan with its
massive subsidies and support in the sixties to form a network of theaters
that screen only Japanese motion pictures and its competitiveness in
animation, and Korea with its screen quotas (Calling for the Establishment,
2002).
This study has several limitations. Using cross-sectional data might limit
the validity of the findings of this study. A longitudinal study is
suggested as an avenue for future research. Also, the sampling limitation
weakens the findings. For example, most countries sampled in this study
were European countries. As more samples from other regions such as Asia,
the Middle East, South America, and Africa are included, the findings about
the relationship between the cultural discount and the self-sufficiency
ratio will be more meaningful. In addition, using a small sample of English
speaking countries and dummying the variable might make it difficult for
this study to identify the impact of the linguistic factor on the
self-sufficiency ratio. As an alternative measurement, it is suggested that
an English fluency index (e.g., Dupagne & Waterman, 1998) be used. Finally,
future research might address not only the screen quota system but also
other regulatory variables (i.e., import quotas, tariffs, and licensing
procedures), which could be incorporated into an institutional protection
factor.
 REFERENCES


World cinema and market shares: Clarity in the confusion. (1996,
September). Screen Digest, 201-208.
Worldwide cinema: Weakness hidden by record growth. (1998, September).
Screen Digest, 201-208.
World film production/distribution: Economic re-structuring. (1998, June).
Screen Digest, 129-136
Yu, S. (1999, November 23). Taiwan's film industry threatened by WTO entry.
Taipei Times. Retrieved November 11, 2002 from
http://www.taipeitimes.com/News/archives/1999/11/23/0000011875.

Figure 1. Market Share of Korean Films and the Amount of Imported U.S. Films










 +
+
Product Investment

Self-Sufficiency Ratio
Ln Box Office Revenue



Figure 2. Path Model among Market Size, Product Investment, and
Self-Sufficiency Ratio

 Table 1. Multiple Regression for Predicting the Self-Sufficiency Ratio
Ba
Beta
T
F
Adj. R2
Model 1
5.63**
.49
  Constant
-45.34

-2.65*
  Ln GDP
8.50
.72
3.46**
  Screen Quotab
4.04
  .17
   .78
  Cultural Distance
4.75
.41
1.63
  English Speakingb
2.51
.09
.45
Model 2
5.86**
.51
  Constant
-44.82
-2.71*
  Ln Box Office
  9.50
.80
3.55**
  Screen Quota
   .84
  .04
   .15
  Cultural Distance
5.21
.45
1.77
  English Speaking
-2.26
-.08
-.42
Model 3
13.87***
.73
  Constant
-22.63
-3.42**
  Ln Investment
7.04
.85
5.97***
  Screen Quota
3.88
  .16
1.08
  Cultural Distance
3.78
.33
1.97
  English Speaking
-5.05
-.17
-1.26

* p < .05, ** p < .01, *** p < .001.
a unstandardized regression coefficients
b Dummy variable as yes = 1 and no = 0.
 TABLE 2. Pearson Correlation Matrix among Variables
SSR
SQ
LN GDP
LN BOX
LN PI
CD
SQa
  .523**
LN GDP
  .615**
  .248
LN BOX
  .618**
  .335
  .936**
LN PI
  .726**
  .216
  .848**
  .903**
CDb
  .195
  .459*
-.355
-.389*
-.341
ESa
-.120
-.153
  .022
  .222
  .274
-.473*

Note. SSR = self-sufficiency ratio; SQ = screen quota; LN GDP = logarithm
of GDP; LN BOX = logarithm of box office revenue; LN PI = logarithm of
product investment; CD = cultural distance; ES = English-speaking.
* p < .05, ** p < .01, *** p < .001, one-tailed.
a Dummy variable as yes = 1 and no = 0.
b Higher value means greater distance between a country and the U.S.

[1] Notes
  Even though literature shows several studies on the effect of television
program quotas in the international broadcasting market (e.g., , there has
been very few studies on the effect of screen quotas in the international
film market. The findings of failed to support a hypothesis that the
presence of television program quota in a country would be negatively
related to the percentage of U.S. fiction imports.
[2] 2 According to CDMI, although the screen quota system in Korea was
adopted in 1966, the system has become more effective since 1993 (CDMI, 2000).
[3]  Countries sampled in this study were Argentina, Australia, Belgium,
Brazil, Canada, Denmark, Finland, France, Germany, Greece, Hong Kong,
Italy, Japan, Netherlands, Norway, South Korea, Spain, Sweden, Switzerland,
and the United Kingdom.
[4]  Inferential statistics show the coefficients of screen quota are not
significant. Non-significance may have resulted from the small sample size
of this study. If this study had had 30 or more countries, this would have
been significant. However, the sampling of countries with the quota system
consists of almost the population of the countries employing the system,
and this study also used a non-probability sample. Thus, inferential
statistics are not necessary. But this study reported the inferential
statistics in Table 1 for the readers who are interested in further research.