Linking Export Competitiveness and Import Market Targeting.

INTRODUCTION

Studies on international competitiveness tend to address the subject at either the macro (nation) level or the micro (firm) level (Waheeduzzaman & Ryans, 1996). There are few studies which explicitly or implicitly address both levels. This is unfortunate because one

This paper is intended to generate micro (firm) level managerial implications for U. S. exporters from the analysis of macro (nation) level international trade data. The objective of this paper is to present the results of a competitiveness-related analysis of global and bilateral export and import trade flow data to show how U. S. exporting companies can better deploy their resources in the domain of the targeting of import markets. Currently, no works that deal with the targeting of import markets (e.g., Green & Allaway, 1985; Kumar, et. al., 1994) explicitly deal with the issue of competitiveness of the exporter in the targeting of import markets.

The thesis of this paper is that the global export competitiveness of U. S. firms operating in a given product-defined industry tends to shape the firms' selection of import markets to target. Particularly, it is posited that U. S. firms that are relatively competitive in serving a given product-defined global export market tend to concentrate their resources in serving countries that have high global import market potential. Conversely, U. S. firms that are relatively uncompetitive in serving a given product-defined global export market tend to target less intensively countries that have high global import market potential. If we are successful in proving this thesis to be true, then the process of targeting import markets by U. S. exporters can be made much more efficient. The efficiency is improved when countries that are otherwise identified as being targetable by U. S. exporters by virtue of their high import market potential are removed from further consideration due to the lack of ability of U. S. exporters to successfully serve these markets from a competitiveness perspective.

THE RELEVANT MEASURE OF "COMPETITIVENESS"

As Heidensohn and Hibbert (1997) acknowledge, there is no single measure of competitiveness that is universally acceptable to all creators and consumers of competitiveness-related studies. One reason behind this has already been alluded to: namely, that competitiveness studies tend to be directed at either the macro (nation) level or micro (firm) level--but not both. Another reason has to do with the role that the "competitiveness" variable plays in a given study (Waheeduzzaman & Ryans, 1996). The competitiveness variable can be a dependent variable, an independent variable or an intermediary variable.

The measure of competitiveness that we use directly reflects the conception of competitiveness that was used in the Report of the President on U. S. Competitiveness (Bowen, 1985), which was sponsored by the President's Commission on Industrial Competitiveness. This conception is presented below in a passage taken from Waheeduzzaman and Ryans (1996).

   "Competitiveness is the degree to which a nation can, under free and fair
   market conditions, produce goods and services that meet the test of
   international markets while simultaneously maintaining or expanding the
   real incomes of its citizens." [Italics added]

Particularly, our measure of "competitiveness" is product-specific and concerns the magnitude of share gained (lost) by U. S. exporters in the global export market defined by a 3 digit SITC-based product category over a specified time period. The greater the share of the 3 digit SITC-based global product market that U. S. exporters gain over the specified time period, the more "competitive" they are deemed to have become in that particular product category over that time period.

Several points of explication and justification regarding our choice of this conceptualization are in order. First, this measure clearly falls under the "market share coefficient" category which was identified by Heidensohn and Hibbert (1997) as one of five viable conceptualizations of "competitiveness". Second, our measure of competitiveness has potential for solving problems which are associated with more generally used measures of competitiveness. To illustrate this, we address the three measures of competitiveness used by Heidensohn and Hibbert (1997) in their sectoral analyses of Europe's competitiveness. In that study, the authors used multiple measures of competitiveness: growth rates of 2 digit and 3 digit SITC based industries, foreign trade intensifies of production in these industries (i.e., share of production exported), and revealed competitive advantage.

Each of these individual measures has a significant shortcoming. For instance, Breuer (1996) notes that the growth rate of an industry is biased by virtue of the fact that an industry that has a low absolute level of production output in one country will tend to grow at a higher percentage growth rate than will the same industry in another country which has a higher absolute level of production output. Therefore, Heidensohn and Hibbert's (1997) use of percentage growth rate to identify competitive industries tends to lead to the identification of industries that are relatively small in terms of absolute level of output, everything else being equal.

The concept of foreign trade intensity (percentage of production output exported) is also flawed. Particularly, the smaller the size of the domestic market's consumption of domestic production of the focal product, the more competitive domestic producers of the product will appear to be--independent of how well domestically produced output actually fares in global markets. Similarly, the concept of "revealed competitive advantage" has a major flaw. The evaluation of the revealed competitive advantage that a country has in comparison with other countries in a trading group with respect to a given product type concerns the computation of the value of the country's exports of the focal product as a percentage of the value of the country's total exports, in comparison with a similar computation for the group of trading countries as a whole. If the percentage for the focal country exceeds the percentage for the group of countries, then the focal country has a "revealed competitive advantage" in exporting the focal product. The bias of this measure is seen when one notes that a drop in the value of the focal country's exports of other unrelated products will suggest an increase in the focal country's competitiveness with respect to the product group in question.

In this paper, we use shift-share analyses in establishing export competitiveness to overcome the problems with other measures. In particular, we use shift-share analyses to evaluate the extent to which U. S. exporters gained (or lost) share in the global export market over a specified time period. The greater the share of the global product-defined export market that U. S. exporters gain, the more "competitive" U. S. exporters are seen as being with respect to the focal product group over the specified time period. Shift-share analyses have not been used to establish export competitiveness before. However, the same benefits that Green and Allaway (1985) describe in using shift-share analyses to analyze import trade flows to establish global import market potential are also available to the analyst in evaluating export trade flows to establish export competitiveness.

Export competitiveness is a "supply" side issue in that it is a characteristic of firms that supply product. Import market potential is the "demand" side counterpart to export competitiveness and deals with the ability and propensity of buyers in the foreign market to purchase the product from sellers not domestic to that market. Import market potential is addressed in the next part.

IMPORT MARKET TARGETING AND THE EVALUATION OF IMPORT MARKET POTENTIAL

Import market targeting concerns the selection of one or more foreign markets to target by an exporting firm. Kumar, et al. (1994) suggest that there are three major issues that should be evaluated by an exporting firm in the selection of one or more import markets to target:

1. import market potential

2. barriers to imported goods

3. competitiveness of the import market

Kumar, et al., (1994) indicate that the import market potential evaluation will likely be the most important item of the three to evaluate. In the very most basic sense, the evaluation of import market potential will tell the exporter whether a given foreign market could ever supply a sufficient volume of business to justify the exporter's entry into the market. The latter two items, barriers to imported goods and competitiveness of the import market, can be viewed as constraints which shape the ultimate profitability of the exporter's sales to the target market. These constraints have little meaning to the exporter if import market potential is deemed to be too low.

The empirical evaluation of import market potential has been exhaustively dealt with by Green and Allaway (1985). In that work, the authors deal with the strengths and weakness of single item measures of import market potential, ones such as (1) percentage growth rate of imports coming into a market and (2) the absolute value of imports coming into a market. Green and Allaway (1985) conclude that shift-share analysis provides the most valid and reliable single item measure for evaluating the potential of an import market.

The work of Green and Allaway (1985) deals almost solely with the "demand" side in the import market targeting process. The main "supply" side recommendation that they make is that the exporter have access to a competitive product offering. The careful reader may find within this recommendation the seeds for a theory which links the competitiveness of an exporter with the potential of the import market that the exporter targets. This theory can be summarized as follows: The import market which an exporter should target is dependent upon the global export market competitiveness of that firm. If a firm is very competitive in the product-defined global export market, then the firm will tend to target import markets which have globally the largest import market potential for that product. On the other hand, if the exporting firm is relatively uncompetitive in the product-defined global export market, then the firm will tend to target import markets other than those with the globally largest import market potential.

This discussion has centered on the (exporting) firm. However, as previously indicated, the empirical analysis will be conducted using data at the macro (nation) level. In order to be able to link empirical results generated using macro (nation) level data to the micro (firm) level, one must make two assumptions. The first assumption is that the variables tested not differ qualitatively at the two measurement levels and in such a fashion that would preclude one from applying nation level results at the firm level.

The validity of this first assumption can be clearly seen when one notes the character of the data that will be analyzed, data which are available from use of the National Trade Data Bank (supported by the U. S. Government) and the PC/TAS CD-ROM, the latter offering data supplied by the United Nations through UNCTAD. In each of these two cases, the data which are analyzed are merely direct aggregations of data supplied by individual firms in the process of their transacting their export and import business activities. Clearly, there is no factor which is involved in the simple aggregation of firm level data that would preclude the application of results using nation level data at the firm level.

The second assumption is that there is an essential homogeneity of competitiveness of U. S. firms that export products which fall in any given 3 digit SITC-based product category. A key justification for this assertion comes from the extension of the concept of "market evolution", a domestic marketing concept (Kotler, 1991) into the export domain. According to Kotler (1991), domestic markets tend to swing back and forth over time between fragmentation and consolidation. In the fragmentation stage, different companies offer products with different features to different customers in different market segments. In the consolidation stage, one set of product features will come to stand out as being most appealing to customers in all domestic groups, and all competing companies will be forced to offer essentially the same specifications in their products in order to survive. This tends to cause the surviving companies to become more operationally similar in their domestic activities.

The extension of this similarity into the export domain stems from the predominant tendency of U. S. firms to compete "multidomestically" (and not globally) in export markets (Makhija, Kim & Williamson, 1997). When a multidomestic company engages in export competition, it tends to treat export markets as essentially extensions of the domestic market. Consequently, the operational similarity of U. S. firms in the domestic market is extended to the export market, and the essential homogeneity of the competitiveness of U. S. exporters follows as a consequence.

Prior to the statement of the hypothesis, we must also demonstrate how we can address the topic of import market targeting at the micro (firm) level by evaluating data at the macro (nation) level. To do this, we must again state that foreign country import data are generated by a straightforward aggregation of individual firm import transactions. The import market targeting activities of U.S. exporting firms as a group are reflected in the specific pattern of their total exports spread across identified foreign importing countries. Expressed at the nation level, the theory proposed in this paper would suggest that the more globally competitive U. S. exporters of a given type of product are, the greater is the concentration of their exports to countries that are empirically established as having globally largest import market potential for the specific 3 digit SITC-based product. The hypothesis which is to be empirically tested can now be stated.

   HYPOTHESIS: There is a positive correlation between (a) the level of global
   export competitiveness of U. S. exporters of a given product and (b) the
   percentage of total U. S. exports of the product in question that are made
   to countries that are highest in global import market potential regarding
   that product.

A description of the procedure which will be used to test this hypothesis follows.

A PROCEDURE FOR TESTING THE HYPOTHESIS

The procedure which is used for testing the hypothesis involves the gathering of secondary data and analysis of this data using both shift-share analyses and a simple correlation procedure. The steps which are involved in each shift-share analysis are defined precisely in this procedure. Steps One and Two involve choice of a time period and relevant product categories, respectively, for the gathering of data. Steps Three through Five generate the competitiveness figures for U. S. exporters of the 3 digit SITC-based products that are addressed. Steps Six through Eight enable us to identify the three foreign countries that represent the highest import market potential for each 3 digit SITC-based product that is addressed. Following Steps Nine through Eleven will show the percentage of total U. S. exports of each 3 digit SITC-based product that is shipped to the three countries that are identified in Steps Six through Eight. Step Twelve is the actual test of the hypothesis and involves the correlation of the export competitiveness figures derived in Steps Three through Five with the import percentages derived in Steps Nine through Eleven. (Please note that while the steps themselves are presented in bold face type, the results of implementation of the steps are presented in regular face type.)

Step One: Choose a time period over which export competitiveness and import market potential will be established. (Since the testing of the hypothesis is of a predictive nature, data which concern the targeting of import markets will be one year later.)

The 1991-1993 time period is chosen as the relevant time period for calculating U. S. firms' export competitiveness and import market potential evaluations. Data concerning the import markets that are actually targeted by U. S. exporters will therefore be 1994 data.

Step Two: Using the reference text which is entitled International Trade Statistics Yearbook--Trade by Commodity, select a randomly chosen sample of 3 digit SITC-based product categories to evaluate.

Starting by choosing at random one of the first five 3 digit SITC numbers in the reference text, we subsequently took every fifth 3-digit SITC-based product category to constitute the sample. Accordingly, we had a sample with 40 SITC-based product categories and therefore 40 observations.

Step Three: For each of the product categories in the sample, look to the reference text which is entitled International Trade Statistics Yearbook--Trade by Commodity and select the 20 countries that have the largest level of exports of the 3 digit SITC-based product in question.

(Note 1: Invariably, the U. S. will be one of the 20 countries.)

(Note 2: Only the top 20 countries were selected, since these countries will account for 98%-99% of global exports in each case. Including all exporting countries would tend to be unwieldy from a computation point of view.)

It is easy to identify the 20 countries with the largest exports of the 3 digit SITC based product in question because countries in the list that is presented are given in descending order of exports. Thus just take the top 20 countries.

Step Four: For each 3 digit SITC-based product, get the beginning and end year exports of the focal product for these 20 countries.

We took the exports for 1991 and 1993 for each of the 20 countries and put them into an electronic disk data set, for further analysis.

Step Five: For each 3 digit SITC-based product, perform a shift-share analysis on the data alluded to in Step Four. This will provide directly a quantitative measure of the export "competitiveness" of U. S. exporters of the focal 3 digit SITC-based product.

We present the "Sub-steps" that constitute the shift-share analysis. (The reader is advised to look at Green and Allaway (1985) for a more detailed assessment of the structure of the shift-share procedure.)

   Sub-step A: Add the exports for the 20 countries for each of the beginning
   and end years. The sums are designated as GLOBEXPBEGIN AND GLOBEXPEND,
   respectively.

The sums are taken for years 1991 and 1993 for the 3 digit SITC-based product in question.

   Sub-step B: Compute the percentage growth (contraction) Which took place
   over the relevant time period. To do this, compute the following:

      RATIO = (GLOBEXPEND/GLOBEXPBEGIN)

   If RATIO is equal to or greater than one, then the percentage growth is
   equal to (RATIO - 1.00).

   If RATIO is less than one, the percentage contraction is equal to (1.00 --
   RATIO).

   Sub-step C: For each of the 20 countries in the array that relates to the
   focal 3 digit SITC-based product, multiply the beginning year export figure
   by RATIO. This product is equal to the "expected" level of exports for that
   country for the end year IF THAT COUNTRY'S INCREASE (DECREASE) IN EXPORTS
   TRACKED PRECISELY THE INCREASE (DECREASE) IN EXPORTS FOR THE 20 COUNTRIES
   AS A WHOLE.

   Sub-step D: For each country, subtract the "expected" level of exports
   (calculated in Sub-step C) from the "actual" level of exports for the end
   year. This difference is called the "net shift" for the country. If the
   "net shift" is positive, then the country gained share in the global export
   market for the 3 digit SITC-based product in question. If the "net shift"
   is negative, then the country lost share over the relevant time period.

   Sub-step E: Arrange the 20 countries in descending order regarding their
   net shift figures. The country with the highest positive net shift figure
   is deemed to have the most competitive exporters over the measurement time
   period. The country with the negative net shift figure having the largest
   absolute value is deemed to be the country which is least competitive in
   exporting the focal product over the measurement time period.

   In our study, the U. S. was fairly evenly balanced regarding the
   competitiveness evaluations for each of the 40 X 3 digit SITC-based product
   categories in question.

When Sub-step 5E is completed for each of the 40 X 3 digit SITC-based product categories, then all of the "competitiveness" calculations will have been completed. The next series of steps (Step Six through Step Eight) are structured to enable one to identify the three countries with the globally largest import market potential for each of the 3 digit SITC-based product categories being evaluated. (We identify three countries rather than just identifying the one country with the globally largest import market potential to add stability to our calculations regarding where the globally largest import market potential exists.)

Step Six: For each of the product categories in the sample, look to the reference text which is entitled International Trade Statistics Yearbook--Trade by Commodity and select the 20 countries that have the largest level of imports of the 3 digit SITC-based product in question.

(Note 1: Invariably, the U. S. will be one of the 20 biggest importing countries. One may wonder how the U. S. can be at the same time one of the 20 biggest exporters and 20 biggest importers of a given 3 digit SITC-based product. The answer concerns the pervasiveness of "intra-industry" trade throughout the study. One should note, however, that the U. S. is excluded in this analysis, since targeting the U. S. by U. S. exporters is not a permissible alternative in this study.)

(Note 2: Similar to the treatment in Steps Three through Five, we look at the top 20 importing countries (excluding the U. S.), since they tend to account for 98%-99% of the total global imports.)

Step Seven: For each of the 3 digit SITC-based product categories, get the beginning and end years imports for each of these 20 countries.

We took the imports for 1991 and 1993 for each of the 20 countries and put them into an electronic disk data set, for further analysis.

Step Eight: For each 3 digit SITC-based product category, perform a shift-share analysis on the data alluded to in Step Seven. This will enable the analyst to rank order these countries regarding their global import market potential for the product category in question.

To perform this set of shift-share analyses, follow precisely the same series of five "Sub-steps" as was followed in Step Five. The only difference here is that one is using import data, not export data. (The interpretation of the results is dramatically different, however, since the results relate to import market potential and not export competitiveness.) For each 3-digit SITC-based product category in the sample, there will be 20 shift-share figures generated--one for each of the 20 countries identified as the largest importers of the focal product.

The sole objective of the completion of Steps Six through Eight is to enable one to identify for each of the 3 digit SITC-based product categories in question the 3 countries that are deemed to have the best prospects regarding global import market potential for the focal product. It is important to note that we do not make any further use of figures generated in Steps Six through Eight in the testing of the hypothesis. The performance of the next series of steps (Step Nine through Step Eleven) will enable one to determine, for each 3 digit SITC-based product, the percentage of total U. S. exports of the 3 digit SITC-based product in question that are exported to the 3 countries identified as having the largest global import market potential for that product. In testing the hypothesis, we will correlate this percentage figure with the "net shift" figure relating to U. S. export competitiveness, the latter which was determined in Steps Three through Five.

Step Nine: Determine total U. S. exports of each 3 digit SITC-based product category for the relevant year. This figure can be gotten from either the "U. S. Foreign Trade Highlights" program in the National Trade Data Bank or from PC/TAS.

The relevant year in this case is 1994.

(Note 1: Recently, persons who support the NTDB have decided to not publish the SITC-based figures that are in this program. Apparently, they want to foster the use of 10 digit Harmonized Code (HS) figures. The figures that were used in this research came from an older CD-ROM version of the NTDB.)

(Note 2: In order to get the 3 digit SITC-based exports from PC/TAS, one will have to add together all of the 4 digit SITC-based data so as to derive the 3 digit figure. Only 4 and 5 digit data are presented directly in PC/TAS. The SITC data are structured in a "nested" fashion. All 5 digit SITC numbers that are subsumed under a given 4-digit SITC number will sum to that 4-digit number.)

In the case of the research under consideration, the relevant year for this data was 1994, not 1993. We are attempting to demonstrate that we can predict the percentage of 1994 U. S. exports shipped to the 3 high importing countries (out of total 1994 U. S. exports of the 3 digit SITC based product in question) by looking at the export competitiveness of the 3 digit SITC based product, established using 1991-1993 export data.

Step Ten: For each 3 digit SITC-based product category in question, determine U. S. exports of the product to each of the 3 countries that were identified in Steps Six, Seven and Eight as having the largest global import market potential for the focal product. The data can be gotten either from the "U. S. Foreign Trade Highlights" program in the NTDB (as described in Step Nine) or from PC/TAS (also described in Step Nine).

(Note: These data are bilateral data. They represent U. S. exports of the given product to a specific country--not total U. S. exports of the product in question.)

The bulk of this category of data was taken from the "U. S. Foreign Trade Highlights" program in the older CD-ROM version of the NTDB. Whenever such data were not available (for whatever reason) from the NTDB, we used data from PC/TAS, summing 4 digit SITC data into 3 digit data (as previously described in Note 2 of Step Nine).

Step Eleven: For each 3 digit SITC-based product category, calculate the sum of the exports to the 3 countries (gotten in Step Ten) and divide this sum by total U. S. exports of the 3 digit SITC-based product category in question (gotten in Step Nine). This percentage quantitatively represents the extent to which U. S. exporters of the product in question tended to target the three countries with the globally largest import market potential.

Step Twelve: Test the hypothesis by correlating the U. S. "export competitiveness" figures (gotten in Step Five) with the percentage figures representing the extent to which U. S. exporters tended to target countries with globally the largest import market potential (gotten in Step Eleven).

RESULTS OF THE EMPIRICAL TEST OF THE HYPOTHESIS AND DISCUSSION

Presented in the Table are the results of the empirical test of the hypothesis. The correlation is seen as significant at the .0486 level, thus confirming the hypothesis. The results suggest that the more "competitive" U. S. firms are in the export of a given type of product, the more U. S. exporters tend to concentrate their export activity in the service of markets that have globally high import market potential. When U. S. exporters of a product are not relatively competitive in a global export market context, then they tend to concentrate their export activities on countries that have lesser import market potential. In doing this, U. S. exporters are implicitly acknowledging that they are unable to compete in the global markets where the investment "stakes" and risk are high, and the potential profit returns commensurately large.

TABLE
TEST OF THE CORRELATION BETWEEN U. S. EXPORT
COMPETITIVENESS AND PERCENTAGE OF TOTAL U. S. EXPORTS
GOING TO COUNTRIES WITH HIGH IMPORT MARKET POTENTIAL

Statistical method: simple correlation analysis

Model: B1 = C + error

                   Analysis of Variance

                  Sums of
Source      DF    Squares    Mean Square   F Value   Prob > F

Model        1     1478.43     1478.43      4.152     0.0486
Error       38    13532.20      356.11
Total       39    15010.63

                    Parameter Estimates

                 Parameter                 T for
Variable    DF     Est.      Std. Error    Para=0    Prob > T

Intercept    1     20.19        3.06        6.596     0.0001
C            1      0.26        0.13        2.038     0.0486

B1 = % of total U. S. exports of a given 3 digit SITC-based
product going to the three countries with the globally highest
import market potential for that product, for the year 1994.
(The identity of the 3 countries was established using
import data covering the 1991-1993 time period.)

C = "export competitiveness" of the U. S. with respect to the 3
digit SITC-based product in question, measured over the 1991-1993
time period.

MANAGERIAL IMPLICATIONS

The managerial implications for individual exporting firms are significant. When management of a U. S. exporting firm goes about the process of selecting one or more import markets to target with a given product, this selection should be tempered by management's evaluation of the global competitiveness of all U. S. exporters of the product in question. If U. S. exporters of the focal product tend as a group to be very competitive in the global export market, then (given the assumption of essential homogeneity of competitiveness of U. S. exporters of this product) management of the focal firm can think seriously about targeting one or more countries that have globally the largest import market potential. If on the other hand U. S. exporters of the focal product tend to be uncompetitive, then management should perhaps consider steering clear of these markets and concentrate efforts on countries where U. S. firms have already achieved some degree of success--even if the market potential of these countries is not as large. The rationale is that getting a piece of a small pie is better than getting no piece of a larger pie.

REFERENCES

Bowen, H. P. (1985). U. S. comparative advantage: a review of theoretical and empirical literature. In Bowen, H. P., Learner, E. E. and Sveikauskas, L (1987) Report of the President on U. S. Competitiveness. Washington, DC: U. S. Government Priming Office.

Breuer, J. B. (1996). The growing export market. Business & Economic Review. July-September, 11-16.

Green, R. T. and Allaway, A. W. (1985). Identification of export opportunities: a shift-share approach. Journal of Marketing. 49 (Winter), 83-88.

Heidensohn, K. and Hibbert, E. P. (1997). A sectoral analysis of Europe's international Competitiveness. Competitiveness Review. 7,2, 25-37.

Kotler, P. (1991). Marketing Management: Analysis, Planning, Implementation and Control. Englewood Cliffs, NJ: Prentice-Hall, Inc.

Kumar, V., Antonie, S and Joachimsthaler, E. A. (1994). An interactive multi-criteria approach to identifying potential foreign markets. Journal of International Marketing. 1, 29-52.

Makhija, M. V., Kim, K. and Williamson, S. D. (1997). Measuring globalization of industries using a national industry approach: empirical evidence across five countries and over time. Journal of International Business Studies. Fourth Quarter, 679-710.

Waheeduzzaman, A. N. M. and Ryans, J. K., Jr. (1996). Definition, perspectives and understanding of international competitiveness: a quest for common ground. Competitiveness Review. 6, 2, 7-26.