UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Japanese manufacturing greenfields : the provincial location decision Brown, Iain A. 1993

You don't seem to have a PDF reader installed, try download the pdf

Item Metadata

Download

Media
[if-you-see-this-DO-NOT-CLICK]
[if-you-see-this-DO-NOT-CLICK]
ubc_1993_spring_brown_iain_andrew.pdf [ 4.31MB ]
Metadata
JSON: 1.0086082.json
JSON-LD: 1.0086082+ld.json
RDF/XML (Pretty): 1.0086082.xml
RDF/JSON: 1.0086082+rdf.json
Turtle: 1.0086082+rdf-turtle.txt
N-Triples: 1.0086082+rdf-ntriples.txt
Original Record: 1.0086082 +original-record.json
Full Text
1.0086082.txt
Citation
1.0086082.ris

Full Text

JAPANESE MANUFACTURING GREENFIELDS: The Provincial Location Decision by IAIN ANDREW BROWN B.Comm., The University of British Columbia, 1982 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE IN BUSINESS ADMINISTRATION in THE FACULTY OF COMMERCE AND BUSINESS ADMINISTRATION (Department of International Business) We accept this thesis as confirming to the required standard  THE UNIVERSITY OF BRITISH COLUMBIA April 1993 ©Iain Andrew Brown  In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission.  (Signature)  Department of  International Business  The University of British Columbia Vancouver, Canada  Date  DE-6 (2/88)  April 21, 1993  ABSTRACT This paper examines why Japanese manufacturing greenfields locate in a particular Canadian province. We find that the location preference is based primarily on the present distribution of Japanese and Canadian firms. A secondary factor is market access, which includes the transportation costs of exporting the Canadian manufactured product to Japan. Other important factors are energy and labour costs. Having utilized quantitative methods to determine that the presence of Japanese and Canadian firms are the main reasons why new greenfields select the province they will locate in, we question the value of using tax dollars to attract investments to locations lacking substantial industry activity.  TABLE OF CONTENTS ABSTRACT ^  ii  TABLE OF CONTENTS ^  iii  LIST OF TABLES ^ LIST OF FIGURES ^  vi  ACKNOWLEDGEMENT ^  vii  1. INTRODUCTION ^  1  2. JAPANESE MANUFACTURING INVESTMENT IN CANADA . . . 3 2.1. The Greenfield Component of Foreign Investment ^ 3 2.2. Japanese FDI ^ 8 3. RELATED RESEARCH ^  10  4.  COMPILING OUR DATA ^ 15 4.1. Dependent Variable - Japanese Manufacturing Greenfields^15 4.2. Independent Variables - Provincial Characteristics ^ 16 4.3. Ship Rank, Energy and MetGDP ^ 19 4.4. Industry Agglomeration as a measure of Manufacturing Activity ^ 22 4.5. Data Collection Summary ^ 24  5.  OUR CONDITION LOGIT ECONOMETRIC MODEL ^ 26  6. RESULTS ^ 7.  CONCLUSION ^ 7.1. Factors Influencing Japanese Manufacturing Greenfield's Location Decision ^ 7.2. Limitations of Our Results ^ 7.3. Policy Ramifications for Provinces Seeking FDI ^ 7.4. Further Study and Extensions ^  8. BIBLIOGRAPHY ^  30 36 36 38 40 41 43  APPENDIX 1: FIFTY-FIVE JAPANESE MANUFACTURING GREENFIELDS IN CANADA ^  46  APPENDIX 2: PROVINCIAL CHARACTERISTICS ^ 49 APPENDIX 3: NOTES TO JAPANESE MANUFACTURING GREENFIELDS AND PROVINCIAL DATA ^ 52 APPENDIX 4: JAPANESE MANUFACTURING GREENFIELDS PROVINCIAL INDUSTRIAL LEVEL FOR 1987 (WAGES, FUEL, and REVENUE) 53 APPENDIX 5: JAPANESE MANUFACTURING GREENFIELDS NATIONAL INDUSTRIAL LEVEL FOR 1987 (WAGES, FUEL, and SHIPMENTS) 60 APPENDIX 6: CONDITIONAL LOGIT REGRESSION RESULTS VARIABLES & THEIR SOURCE ^  62  LIST OF TABLES TABLE 4.1: Number of Japanese Manufacturing Greenfields in Canada ^ 16  TABLE 4.2: 1989 (1980) Provincial Ranking ^  25  TABLE 5.1: Energy Intensity ^  29  TABLE 6.1: Final Results ^  30  TABLE 6.2: Substituting Low Energy Price ^  31  TABLE 6.3: Testing Unionization, Tax & Crime Rates ^ 32 TABLE 6.4: Testing Factor Intensities ^  33  TABLE 6.5: Replacing MetGDP with Quebec Dummy ^ 34 TABLE 6.6: Adding a Quebec Dummy Variable ^ 35  LIST OF FIGURES FIGURE 2.1: Canada's Liabilities to Non-Residents ^ 5 FIGURE 2.2: Acquisition and Greenfield FDI Inflows ^ 7 FIGURE 2.3: 1989 Worldwide Gross Outflow of FDI ^ 8 FIGURE 4.1: Provincial Lowest Energy Costs ^ 21 FIGURE 4.2: Provincial Average Energy Costs ^ 21 FIGURE 6.1: MetGDP Variable ^  34  ACKNOWLEDGEMENT I thank all those who have been inspirational during my work on this thesis. I am particularly indebted to John Ries for his assistance throughout the preparation of this paper. My gratitude for John's contribution is only dwarfed by my respect for his unselfish dedication to the project. Keith Head's assistance in the development of our model was most appreciated. I thank Jim Brander for helping to polish and structure this paper for final presentation. Anna Kwan deserves recognition for compiling this manuscript.  lain Brown JAPANESE MANUFACTURING GREENFIELDS: The Provincial Location Decision ^page 1  1. INTRODUCTION  In recent years Canada has sought new Japanese manufacturing investments. (New investments are often referred to as greenfield investments). Public policy interest in greenfield investment arises in part because of the political benefits of job creation that are associated with them. Another reason that provinces are interested in attracting Japanese manufacturing greenfields is technology transfer. However, job creation tends to be the host province's main motivation in soliciting Japanese manufacturing greenfields. Regardless of the reason(s) why provinces may wish to attract Japanese manufacturing greenfields, in order to have better success in attracting such investments, it would be useful for the provinces to know what attracts Japanese manufacturing greenfields to a particular province. This thesis focuses on this issue by asking the question what provincial characteristics influenced the greenfields to locate in the particular province they selected?  It turns out that between 1980 and 1991, 84% of Japanese manufacturing greenfield investment locating in Canada established their new facilities in either Ontario (64 %) or British Columbia (20 %). Both these provinces attracted more greenfield investment than their national share of manufacturing. By using condition logit regression we discover that Japanese manufacturing greenfields prefer to locate in provinces with a concentration of Canadian firms in their industries, and in provinces with other Japanese manufacturing greenfields. Other important provincial characteristics in determining location preference include wage and energy  lain Brown JAPANESE MANUFACTURING GREENFIELDS: The Provincial Location Decision ^page 2  costs as well as geographic proximity to Japan. Past research on American state location decisions, which used similar statistical methodology as we employ, supports our findings in that state characteristics, similar to our provincial characteristics, are identified as reasons for manufacturing firms establishing new facilities in a particular location. However, our results may suggest evidence of pure agglomeration effects. That is, if the combination of our variable designed to measure geographic proximity to Japan and our variable that counts the number of already existing Canadian firms in the establishing greenfield's industry, capture the endowment effect, then the significance of our variable measuring the size of existing Japanese investment may be evidence of pure agglomeration effects.  Section 2 provides some background on Japanese investment. Section 3 reviews related research. Section 4 explains how the data for our variables was compiled. Section 5 describes our econometric model. Section 6 summarizes our results. Section 7 is our conclusion that includes a discussion on the limitation of our results, policy implications and research questions that follow naturally from this study.  lain Brown JAPANESE MANUFACTURING GREENFIELDS: The Provincial Location Decision ^page 3  2. JAPANESE MANUFACTURING INVESTMENT IN CANADA  The purpose of this section is to familiarize the reader with foreign investment so that an appreciation is gained for what we mean by a Japanese manufacturing greenfield. We point out some of the possible benefits that are associated with foreign direct investment. However, we are not saying that other forms of investment are any less interesting or beneficial, nor that our methodology would only work for Japanese investments. Our methodology could have been used to study, for example, the provincial location decision of German greenfields, but we elected to study Japanese greenfields and hence this why we provide an overview of Japanese manufacturing investment in Canada.  2.1. The Greenfield Component of Foreign Investment  Before addressing the issue of why we chose to measure the greenfield component of Japanese foreign direct investment (as opposed to, say, American direct investment) let us first distinguish between foreign portfolio investment (FPI) and foreign direct investment (FDI).  FDI is ownership (with control) of real domestic assets by a foreigner. Statistics Canada considers foreign ownership to exist when a foreigner owns more that 10% of the equity of an investment'. Usually, FDI is undertaken by corporations to take advantage of the  1  ^As Appendix 1 shows the lowest level of Japanese ownership was 17%, with over half the greenfield investments being wholly owned subsidiaries.  lain Brown JAPANESE MANUFACTURING GREENFIELDS: The Provincial Location Decision ^page 4  comparative advantage that foreign production offers or to preserve access to markets in an environment where true global free trade does not exist (i.e., Japan exporting to Canada or the U.S., and for that matter, vice versa).  FDI typically involves the transfer of capital as well as technology, marketing and organizational skills. Management practises such as just-in-time (kanban) inventory control procedures lessen inventory carrying costs, quality circles (kaizen) reduce the number of defects, and consensus decision making (nemawashi) increase productivity, all originated in Japan. However, through technology transfer these management practises are no longer solely the property of Japanese transplants operating in Canada. They have also spread to Canadian owned manufacturing facilities, which now benefit from them. These examples of technology transfer, along with job creation, may motivate provinces to attract FDI, even though an increase in foreign ownership of this domestic industry will occur as ownership of these assets remaining with the foreign firm.  This ownership can take the form of acquisitions or the construction of new facilities (greenfields). Other forms of FDI involve the establishment of a wholly-owned subsidiary or the formation of joint ventures and strategic alliances between firms. (Although strategic alliances may not involve the transfer of capital they do tend to have the other benefits that are associated with FDI). The alternative type of business structure to be used in the host country is important as the choice of organizational form often has significant implications for the transfer of knowledge and other firm-specific skills.  lain Brown JAPANESE MANUFACTURING GREENFIELDS: The Provincial Location Decision ^page 5  Finally, as FDI involves the transfer relatively intangible resources, the corporations involved typically operate with a fairly long time horizon, making flows of FDI stable in comparison to flows of FPI. However, as we see from Figure 2.1, FDI is about one quarter of the stock of foreign investment in Canada as of 1990. FDI is represented by the black wedge in Figure 2.1, and the remaining pieces of the pie are  Canada's Liabilities to Non-res (Cdn BI)  collectively known as foreign STOCKS 20.8  portfolio investment. FPI occurs when foreigners own financial  FDI 126.6  Other. 108.7  assets from another country. Money Market Sec. 25.4  Figure 2.1 2 illustrates that in  1990 foreigners owned $329.6  Cdn. Gov . Bonds 63.2  Prov. Gov . Bonds 65  billion worth of Canadian financial assets.  Corp. Bonds 46.5  Figure 2.1  (Source: Statistics Canada, 1990).  FPI is undertaken mainly by individuals, institutional investors and governments (i.e. central banks). FPI involves transactions in the capital markets (e.g., stocks, bonds, loans or money market securities). Those involved in FPI have different objectives, risk preferences, and investment time horizons than those involved in FDI. The risk-adjusted real rates of return  2  The 108.7 of Other Liabilities to Non-residents includes SDRs (Special Drawing Rights) and other official government flows. Also theses stocks are at historical costs and if the FDI was made some time ago its market value may exceed its book value by a considerable margin. Thus, its is possible that if market figures were available that FDI would represent a bigger slice of the pie.  lain Brown JAPANESE MANUFACTURING GREENFIELDS: The Provincial Location Decision ^page 6  between assets in different countries is a major determinant of FPI flows. Thus, FPI flows are sensitive to changes in inflation, interest and exchange rates as well as portfolio diversification requirements. Accordingly, FPI tends to be in financial assets that are relatively liquid and mobile with fixed maturity dates.  Just as FDI and FPI are less than perfect substitutes, acquisitions and greenfields are different types of FDI, if only for the reason that the former does not involve the same location decision. The location of a new plant is clearly a choice; whereas, it could be argued that an acquisition did not involve a location analysis because the site already exists. When a company is acquired, it may be acquired despite of the location of its manufacturing facilities. That is, in an ideal situation the purchaser would prefer the manufacturing facilities to be elsewhere (perhaps in another province). Nevertheless, if the purchase takes place, because the overall investment makes sense, then this non-ideal location was not actually chosen. However, in a greenfield investment, the ideal location would be selected. Thus, we have included only Japanese greenfields in our study.  A secondary reason for only considering greenfields is unique to Canada: what we call "the FIRA (Foreign Investment Review Agency) effect." Prior to 1985 FIRA (a government organization designed to discourage foreign investment) may have inhibited acquisitions. Greenfields apparently did not offset the reduced level of acquisitions during the FIRA years (1974-1985). Presumably a Canadian greenfield was not the foreign firm's second choice after a Canadian acquisition and the direct investment simply went elsewhere.  lain Brown JAPANESE MANUFACTURING GREENFIELDS: The Provincial Location Decision ^page 7  In 1985 FIRA was replaced by Investment Canada, which is a government body with a mandate to promote foreign investment. The mid 80's were also a period of heightened merger and acquisition activity. We suspect that a combination of these factors explains the increase of acquisitions in 1985 (refer to Figure 2.2).  Acquisition and Greenfield FDI Inflows  Regardless of the reason,  Cdn. Billions  acquisitions dramatically 14  increased in 1985. This imbalance of acquisition activity from the beginning of the decade to the end of the decade is another reason we have elected to study the  12 10 8 6 4 2 0 1980 1981 1982 1983 1984 1985 1986 1987 198819891990  Investment Canada  •  Greenfield* Acquisitions  Figure 2.2  greenfield component of FDI.  In summary, FDI and FPI cannot be treated as perfect or even close substitutes because they are motivated by different factors. We have elected to measure FDI rather than FPI. We selected the greenfield component of FDI to avoid the FIRA effect as well as any controversy over an acquisition being a "non-location choice."  lain Brown JAPANESE MANUFACTURING GREENFIELDS: The Provincial Location Decision ^page 8  2.2. Japanese FDI  Another reason that we are interested in Japanese greenfield FDI is Japan's recent dominance as an exporter of FDI. For example, Figure 2.3 shows Japan was the leading exporter of FDI in 1989. With Japan being a prominent world exporter of FDI this is one reason to study Japanese FDI. Secondly, although American FDI into Canada is also significant, in future work we wish to objectively measure the effect  1989 Worldwide Gross Outflow of FDI (%)  of the Canada/U.S. Free Trade  France Developing Countries 9.9^4.5  Agreement on FDI into this  • France A  Germany  trade area. To do this we will  ■ Japan  need to measure the location  •:::  United Kingdom  decision of an outside party  • Other Developed  and by definition this excludes the U.S.  United States  ^ Canada  United Kingdom 16.3  United States 13.5  E Developing Countries  United Nation, World Investment Report  Figure 2.3  For these reasons we have chosen to analyze Japanese greenfield FDI. We have elected to look at manufacturing greenfields (as opposed to say distribution warehouses) for reasons such as the technology transfer and job creation benefits associated with these investments. The time frame of our study is 1980 to 1991. This time frame was selected because it has really only been the last decade that Japanese FDI has become such a noticeable factor in industrial host countries.  lain Brown JAPANESE MANUFACTURING GREENFIELDS: The Provincial Location Decision ^page 9  However, host countries are generally made up of regional political territories and in Canada these are called provinces. Often these provinces try to out bid each other (i.e., with tax holidays) for greenfield investments, possibly because of the job creation or technology transfer associated with these new manufacturing facilities. Regardless of the reason(s) why provinces seek to host FDI, to gain insight into how host provincial governments may be more successful in attracting Japanese manufacturing greenfields we will study the provincial location decision of these greenfields.  lain Brown JAPANESE MANUFACTURING GREENFIELDS: The Provincial Location Decision ^page 10  3. RELATED RESEARCH  Before we describe how the data was collected for this study, we will first discuss related research to this paper. These studies have aided us in selecting the variables that we will present in the next section and in designing the conditional logit econometric model that we will describe in Section 5. In this regard we note that logit regression was used by Bartik (1985), Carlton (1983), Coughlin, et. al. (1991), Head, Ries, and Swenson (1993), Luger and Shetty (1985), Schmenner, et. al. (1987), and Woodward (1992).  Bartik (1985) looked at the influence of state characteristics on new manufacturing plant location decisions, emphasizing unionization and taxation. The location data included the Fortune 500 companies' manufacturing plants from 1972-1978. He found a significantly large negative effect of a state's unionized rate, while wages had only a marginal negative significance. The corporate tax rate also had a negative effect on plant location. Bartik estimated a powerful effect of existing manufacturing activity on new business location, i.e., "a state with 10% greater existing manufacturing activity will have an 8% or 9% greater number of new plants" (Bartik, 1985).  Schmenner, et. al. (1987) also examined new plant openings by Fortune 500 firms between 1970-1980. Their innovations were to include plant-specific characteristics (data was derived from surveying managers of the firms) in magnifying or tempering the state-specific  lain Brown JAPANESE MANUFACTURING GREENFIELDS: The Provincial Location Decision ^page 11  effects. They argued that state characteristics such as wage rates would be more important to particular plants (i.e., wages will matter more to labour intensive plants). By interacting state and plant characteristics, Schmenner found that "Labor unionism, climate, population density, and building costs are prominent influences on location. . . Tax variables, on the other hand, make a weak showing" (Schmenner, et. al., 1987). Also the interaction terms were usually significant.  Carlton (1983) concentrated on a few industries that were selected to ensure equivalence (i.e., location choice is not hindered by local supply and demand factors). His results included: [the] "wage effect cannot be measured very precisely; energy [prices] have a large effect; taxes and state incentive programs do not seem to have major effects; [and] existing concentrations of employment matter a great deal with the effect being stronger for industries with smaller average plant size" (Carlton, 1983).  Insignificance of taxes was also evident in Luger & Shetty (1985), who measured the manufacturing activities of three specific industries. Their objective was to measure the elasticity of FDI in relation to a state's promotional programs, as well as the effect of agglomeration and urbanization economies, and labour market conditions. They concluded "that agglomeration economies and wage rates are the most important determinants of new plant location. . . [and] public policies do not have a uniform effect on industries" (Luger & Shetty, 1985). To represent agglomeration attraction a measure of total manufacturing activity was used; thus they did not provide evidence on industry-level agglomeration. Moreover, unmeasured state  lain Brown JAPANESE MANUFACTURING GREENFIELDS: The Provincial Location Decision ^page 12  endowments (forests, research universities, or other omitted state characteristics) might be reflected in the coefficient on manufacturing activity.  Like Luger & Shetty before them, Smith & Florida (1992) chose to explore a select industry (automobile-related) and the industry-specific relationships associated with it. They concluded that Japanese auto-related parts suppliers tend to locate near their assembly plant and have a preference for areas with greater aggregate manufacturing activity. Taxes had a marginal showing and unionization was not a factor, while wages had a positive relationship. This positive relationship could be due to the need for technical skills that might be reflected in wages. For similar reasons, education level also had a significant positive effect.  Coughlin's (1991) research data included all FDI transactions (acquisition, equity increase, joint venture, merger, new plant or plant extension). His results indicated that FDI located in states with higher per capita incomes and higher levels of manufacturing activity. Coughlin also found that FDI is attracted to areas with lower wages and higher unemployment rates, while higher taxes deterred FDI. Unionization had a marginal positive significance, which could be "because of the increased productive efficiency in manufacturing stemming from unionization" (Coughlin, et. al., 1991).  Woodward's (1992) study involved a measure of total manufacturing activity, rather than selecting certain industries. He analyzed Japanese-affiliated manufacturing investments in the U.S. between 1980-1989. The model worked on the assumption that  lain Brown JAPANESE MANUFACTURING GREENFIELDS: The Provincial Location Decision ^page 13  "Japanese firms, like all firms, seek branch locations with the highest expected profits" (Woodward, 1992). This relationship revealed a preference for strong markets and low unionization rates. In addition, "Japanese manufacturing plants are most likely to select counties characterized by manufacturing agglomeration, low unemployment and poverty rates, and concentrations of educated, productive workers" (Woodward, 1992).  Head, Ries, and Swenson (1993) found that Japanese firms tend to locate near both other Japanese firms and U.S. firms that were in the same specific industry. They were able to use industry-specific variables for both Japanese and American companies and found both to be significant. They also found that the attractiveness of a state is increased by the level of industrial activity on bordering states. They showed that the positive relationship between industry activity and location is partly due to agglomeration externalities, not simply an endowment effect.  As not all the research measured the same variables, it is difficult to generalize the findings of these studies. However, we are able to draw some conclusions from these papers. Intuitively, unionization, wages, and taxes should pay significant roles in the location decision. Bartik (1985), Schmenner (1987), and Woodward (1992) suggested unionization is negatively correlated with FDI, while Coughlin (1991) showed a marginal positive relationship. Luger & Shetty (1985) and Coughlin (1991) concluded a negative significance for wages, while Smith & Florida (1992) took the opposite view. The differences in these results may stem from industryspecific differences, in that capital intensive industry may be willing to "put up" with higher  lain Brown JAPANESE MANUFACTURING GREENFIELDS: The Provincial Location Decision ^page 14  wages and unionization in exchange for better trained and more productive workers. While labour intensive industries would place more importance on wages. Hence factor intensities could be the reason for the different findings.  Taxes are generally found to be insignificant or inconclusive, with the exception of Bartik (1985) who found taxes had a negative effect. While never quite defined the same, when a variable is used to measure the level of manufacturing activity the results are usually that "existing manufacturing" attracts new manufacturing. In our next section we build on this research to explain why Japanese manufacturing greenfields locate in the particular province that they do, once they have elected to enter the Canadian market.  lain Brown JAPANESE MANUFACTURING GREENFIELDS: The Provincial Location Decision ^page 15  4. COMPILING OUR DATA  4.1. Dependent Variable - Japanese Manufacturing Greenfields  Data collection for this study was itself a major task as this data is not readily available. Furthermore, data available on Japanese-owned manufacturing facilities in Canada do not always clearly distinguish greenfield operations from acquisitions.  To construct our detailed list we obtained much of our information from Japanese sources (Toyo, Jetro, Dodwell). At times these directories of Japanese companies in Canada provided conflicting information. Not one of them included all the companies that appear in Appendix 1. Furthermore, some information had to be verified by contacting the firms directly. This was often the case when Jetro claimed that a firm was a distribution manufacturing firm, but this firm was not listed in other manufacturing directories. It usually turned out that these firms were solely distribution companies. We found Toyo to be most reliable. We supplement the data from these three directories with information on Japanese companies in Ontario compiled by the Government of Ontario.  We compiled a comprehensive list of fifty-five Japanese manufacturing greenfields that had located in Canada since 1980 (see Appendix 1). These investments were spread over six provinces, but were mainly concentrated in British Columbia (BC) and Ontario. Table 4.1  lain Brown JAPANESE MANUFACTURING GREENFIELDS: The Provincial Location Decision ^page 16  portrays the distribution and growth of these firms in Canada during the period from 1980 to 1990. For example, we note that British Columbia received one investment in 1980 and two more Japanese greenfields located in the province in 1983; thus, from 1980 to 1983 three greenfields had located in this province.  TABLE 4.1  NUMBER OF JAPANESE MANUFACTURING GREENFIELDS IN CANADA (Accumulated Totals Since 1980) Province  1980  1981  1982  1983  1984 I^1985  1986  1987  1988  1989  1990  Alberta  0  0  0  0  0  0  0  0  0  0  1  BC  1  1  1  3  3  3  3  3  5  8  11  NB  1  1  1  1  1  1  1  1  1  2  2  Ontario  4  5  6  7  10  13  20  27  30  33  35  Quebec  0  0  0  0  0  0  2  2  2  4  4  Sask.  0  0  0  0  0  0  0  1  2  2  2  Canada  6  7  8  17  26  34  40  49  55  11^I  14^I  BC = British Columbia^NB = New Brunswick^Sask. = Saskatchewan  4.2. Independent Variables - Provincial Characteristics  With our dependent variable data collected we turned our attention to establishing and gathering factors (our independent variables) that may influence the selection of the province to locate in. To help determine what these factors may be we relied on recent work [Bartik (1985), Carlton (1983), Coughlin (1991), Head, Ries, and Swenson (1993), Luger and Shetty  lain Brown JAPANESE MANUFACTURING GREENFIELDS: The Provincial Location Decision ^page 17  (1985), Schmenner (1987), Smith & Florida (1992), and Woodward (1992)] that has looked at similar location decisions in the United States. These papers focused on state choices and thus were very applicable for our study on provincial location choices.  The following list represents independent variables for which we collected Canadian provincial data. A brief description of each is contained in this list. (For more detail and actual time series data see Appendix 2, with notes to this data in Appendix 3). With the exception of our Ship Rank, Average Energy and MetGDP variables (which are described in further detail in the next section) and our Canadian Industries and Japanese Greenfields variables (described in Section 4.4) the following variables are relatively self explanatory.  1. Provincial Population (Cansim'). This is a measure designed to capture the relative size of the provincial economy. 2. Percentage of population living in Metropolitan areas (Cansim). Although the population of the province may be larger than another province, if it is very spread out this tends to hamper economic activity. Thus, a measure of concentration of the province's population is useful. In Canada metropolitan areas include cities with over 100,000 people. 3. Crime rate, a probabilistic variable is a measure of reported offenses per 100,000  people (Cansim). This variable is included as a measure of quality of life, which is increasingly receiving more attention. Thus, we thought that it was appropriate to see if the quality of life decision effected the provincial location process of Japanese manufacturing greenfields.  4. Unemployment rate, which has both positive and negative attributes associated  with it. For instances, high unemployment is positive in terms of available work force, but is a negative in terms purchasing power (Cansim). Nevertheless,  Computerized on-line database of Statistics Canada.  lain Brown JAPANESE MANUFACTURING GREENFIELDS: The Provincial Location Decision ^page 18  despite the mixed effects of this variable, it is a general economic indicator; thus, its inclusion in our study. 5. Unionization rate is for total provincial labour force; however, for the years we  checked, the manufacturing unionization rate did not significantly differ (Statistics Canada, Advisory Services). Unionization is include in our model as manufactures often are said to avoid it. However, unionization may also be a measure of quality of the work force and as such capital intensive industries could see it as a desirable quality.  6. Area in square kilometres not including fresh water (Canada Year Book). We  include this variable as a measure of room to grow and natural resources.  7. All production managers tend to be concerned with the cost of average manufacturing Wages (Manufacturing Industries of Canada). However, high Wages may also be associated with skilled productive workers and as such high Wages in themselves may not be a bad thing. 8. Provincial GDP (Cansim). This variable is included as a measure of market size. 9. Highway Kilometres per square kilometre (Roads and Transportation Association  of Canada) is designed to measure the level of existing infrastructure, which is thought to be a factor in attracting new industry.  10. Average Energy is a weighted average cost of fuel in the province (Energy  Statistics Hand Book). The weights are based on existing industrial use for 1992, of each fuel, in each province. The cost of energy is generally of concern to manufacturers and as such is included in our study.  11. Low Energy is a measure of the lowest priced form of energy in the provinces  at the time the Japanese manufacturing greenfield was established (Energy Statistics Hand Book). The cost of energy is important to manufacturers and as such is included in our study.  12. Airports with control towers (Transport Canada) are thought to represent a  measure of transportation infrastructure as well as the preferred choice of executive transportation. Hence its inclusion.  13. Provincial corporate Tax rate (Canadian Tax Reporter). As businesses are usually established to make profits, the tax rate that they will face is an important factor and possibly may determine the manufacturing facility's provincial location. However, high taxes in themselves are not necessarily bad. It depends on what the tax dollars are spent on, and how efficiently this process is carried out. For  lain Brown JAPANESE MANUFACTURING GREENFIELDS: The Provincial Location Decision ^page 19  example, if tax dollars are spent effectively on improving needed infrastructure, then this may be seen as a positive. 14. Percentage of Provincial Labour Force in Manufacturing is a measure of provincial manufacturing activity (Cansim). It has been included in our study as a possible reason for plant location because Japanese manufacturing greenfields may only be attract to the relatively more industrial provinces. 15. The percentage of Canadian Manufacturing Labour Force that resides in the province is a measure of manufacturing activity (Cansim). It has been included in our study as a possible reason for plant location because Japanese manufacturing greenfields may only be attracted to the relatively more industrial provinces. 16. Agglomeration of specific Canadian Industries in each province is a measure of industry-specific manufacturing activity (Manufacturing Industries of Canada) in each province. This variable is included as firms in the same industry may be attracted to each other. 17. Agglomeration of Japanese Greenfields in each province is a measure of Japanese manufacturing activity (Toyo, Jetro, Dodwell). We have included the data in Table 4.1, as we wish to test if Japanese manufacturing greenfields are attracted to provinces that already have a base of such firms. 18. Ship Rank is a measure of ease of water access to Japan and is important when the Japanese-owned firm is shipping the manufactured product back to Japan. This variable could also be of some important when parts are being shipped to Canada to be assembled (although closeness to the market may be more important).  4.3. Ship Rank, Energy and MetGDP  Our Ship Rank variable is a combination of geographic location and port facilities and is a measures of ease of water access to Japan. This is important not only for shipping over parts to be assembled in Canada, but also to ship back finished goods to Japan, such as lumber from Japanese-owned sawmills in British Columbia.  lain Brown JAPANESE MANUFACTURING GREENFIELDS: The Provincial Location Decision ^page 20  British Columbia is ranked highest because it has deep water ports and the closest access to Japan. New Brunswick, with a deep water port, but requiring ships bound for Japan to utilize the Panama Canal, is second. Quebec, which requires a trip up the St. Lawrence Seaway, is third. Ontario fairs worse than Quebec, due to the canal that must be travelled to enter the Great Lakes from Montreal. The remaining two provinces hosting Japanese manufacturing, are the only two land locked provinces in the country; thus, Alberta and Saskatchewan score worse in this category (Appendix 2 presents a Ship Rank for all provinces).  Apart from water transportation, manufacturing goods are commonly moved by both rail and truck. However, comparing the length of rail tracks in each province is less favourable than comparing the total paved highway kilometres, because of the history of Canadian rail development. Rail development was more prominent in the east during the earlier years of Canadian industrialization; whereas, road transportation developed later and as such lacks the eastern bias associated with rail development. Thus, we have elected to concentrate on highway, rather than rail kilometres.  Along with our shipping variable we also devised our own method for measuring energy costs. Our Low Energy variable is a measure of the lowest priced fuel (among crude oil, natural gas or electricity) that prevailed at the time the Japanese decided to establish the greenfield in Canada. The second variable we constructed (which was also calculated by converting fuel prices to dollars per million Btu's) was a weighted average cost variable. Our Average Energy variable was calculated utilizing weights based on existing industrial  lain Brown JAPANESE MANUFACTURING GREENFIELDS: The Provincial Location Decision ^page 21  consumption of fuel in each province. These two variables  Provincial Lowest Energy Costs 7  are displayed graphically in Figure 4.1 and Figure 4.2.  6—  5B4—  Finally, let us explain our MetGDP variable, which is designed to capture the characteristics of the  0  1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990  —Alberta^+British Columbia *New Brunswick + Ontario^Quebec^+ Saskatchewan Statistics Canada, Energy Statistics Hand Book  Figure 4.1  provincial market, which is also important to foreign investors. In general, a strong  Provincial Average Energy Costs 10  market is found where there is a concentration of people with purchasing power. Thus, we would expect the combination of urban concentrations and gross domestic provincial  2— 0 ^ 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 --Alberta^+ British Columbia *New Brunswick + Ontario^* Quebec^+ Saskatchewan Statistics Canada, Energy Statistics Hand Book  product to be an influential Figure 4.2 factor. Urban concentration is represented by the percentage of the provincial population living in a city with a population greater than 100,000 people. To arrive at our single measure of market size (MetGDP) we multiplied GDP by Metropolitan.  lain Brown JAPANESE MANUFACTURING GREENFIELDS: The Provincial Location Decision ^page 22  4.4. Industry Agglomeration as a measure of Manufacturing Activity  Industry agglomeration in itself may attract additional industry. That is industry attracts industry. Krugman (1991) suggests that users and suppliers of intermediate inputs will locate around one another in an effort to minimize transportation costs and encourage economies of scale. Also, technological spillovers may cause firms in the same industry to locate in proximity to each other. Smith & Florida (1992) point out that Japanese auto part suppliers tend to locate near Japanese auto manufacturers. Bartik (1985) attributes "existing manufacturing activity" as the reason for other manufactures locating there. Our study addresses this issue of whether the existence of manufacturing attracts new different kind of manufacturing to the area. If so, we would expect our variables measuring manufacturing intensity (i.e., variables 14 and 15 in Section 4.2) to be reasons for Japanese greenfields locating in the province that they do.  If however, the level of provincial overall manufacturing activity is not as good a measure as industry-specific clusters are of provincial location choice, then we would expect our Canadian Industries and Japanese Greenfields variables to be more significant. Building on the work of Head, Ries, and Swenson (1993), our Canadian Industries and Japanese Greenfields variables are counts of establishments in specific industries, accumulated at the provincial level. To determine Canadian Industries we utilised SIC classifications (based on 4, or in some cases 3, digits). That is, on a provincial basis we counted all Canadian industries that had the same SIC classification as the entering Japanese greenfield. Appendix 4 contains the industry-specific counts on a provincial basis for the greenfields, with Appendix 5 providing the national counts.  lain Brown JAPANESE MANUFACTURING GREENFIELDS: The Provincial Location Decision ^page 23  We also used the notion that Japanese firms are attracted to locate in proximity to other Japanese manufacturing facilities. Unfortunately, the industry-specific data (based on SIC classifications) that was collected for Japanese manufacturing facilities already in Canada, was of little use, because of the small number of industry-specific Japanese firms present in each province at time of entry. Therefore, rather than use a Japanese count of existing firms in the same industry we utilized our data presented in Table 4. I', which provided a count of Japanese manufacturing greenfields that existed in the province prior to the new Japanese entrant.  While there is a possibility that the general level of manufacturing will in itself attract additional manufacturing facilities to locate there, we suspect that location choice is driven by industry-specific establishment counts. However, as Head, Ries, and Swenson (1993) point out, the location of like industry, an agglomeration effect, should not be confused with endowment effects. For example, if a new pulp mill locates in a less industrial province (say, British Columbia rather than, say, Ontario) then it may have done so for either industry-specific reasons (most Canadian pulp mills are in B.C.) or endowment effects (most trees close to the Pacific Ocean are in B.C.). In general, it is difficult to distinguish between agglomeration and endowment effects.  4  For a true measure of Japanese greenfields we also included greenfields established prior to 1980. Accordingly, this increased the accumulated totals in Table 1. Alberta had one; British Columbia had two; Ontario had two; Quebec had four; and Saskatchewan and New Brunswick had no Japanese manufacturing greenfields locate in the province prior to 1980.  Lain Brown JAPANESE MANUFACTURING GREENFIELDS: The Provincial Location Decision^page 24  4.5. Data Collection Summary  We were able to collect Canadian data for all the variables listed in Section 4.2. We summarize our variables in Table 4.2. In this table each provincial variable is ranked for the years 1989 and 1980. The 1980 ranking only appears in brackets if the ranking actually changed. We observe that approximately one third of the rankings do change. Generally, these changes are minor rearrangements. Only 3% of the rankings actually changed more than three places and apart from the Ontario's unemployment rate becoming the lowest in Canada in 1989, and New Brunswick becoming more unionized in 1989, all other major changes in provincial rankings occurred in the area of provincial corporate tax rates.  TABLE 4.2 1989 (1980) PROVINCIAL RANKING Prov  Rank  Pop  Met  Crime  UI  Union  Area  Wage  GDP  Hw  Engy  Air  Tax  P-m  C-m  Ship  ON  1  1  1  6  1(4)  3(5)  3  9(8)  1  6  5(4)  2  7(6)  1  1  6  PQ  2  2  2  1(2)  6(7)  9(8)  1  7(6)  2  8  6  3  1(4)  2  2  5  BC  3  3  5  10(8)  5  7(9)  2  10  3(4)  9  3  1  2(10)  5(3)  3  1  AB  4  4  3  8(10)  2(1)  1(2)  4  8(9)  4(3)  5  1  4  4(2)  8  4  9  MN  5  5  4  7  4(3)  8(7)  6  3(2)  5(6)  7  4(5)  7  10(8)  4  5  7  NS  6  7  7  5  7(6)  2(4)  9  4  7  2  9(8)  9  6(5)  6  6  3  SK  7  6  6  9  3(2)  5(1)  5  6(7)  6(5)  3  2  6  5(7)  10  8  10  NB  8  8  9  3  8(9)  6(3)  8  5  8  4  8(7)  8  8(3)  3(5)  7  2  PE  9  10  10  4  9(8)  4(6)  10  1  10  1  7(9)  10  3(1)  9  10  8  NF  10  9  8  2(1)  10  10  7  2(3)  9  10  10  5  7  9  4  Population % of Population living in Metropolitan Area Crime Rate Unemployment Rate Unionization Rate Area in Square Km Average Manufacturing Wages Provincial GDP Highway Km per Square Km Average Energy Airports with Control Towers Provincial Corporate Tax Rate % of Provincial Labour Force in Manufacturing % of Cdn Manufact. Labour Force in Province Shipping - ease of water access to Japan  1 1 1 1 1 1 1 1 1 1 1 1 1 1 1  = Highest = Highest = Lowest = Lowest (+ economic indicator) = Lowest = Biggest = Lowest = Highest = Highest = Lowest = Most = Lowest = Highest = Highest = Easiest  - 25 -  9  PROVINCE^TOTAL SCORES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10.  Ontario^53 Quebec^57 British Columbia^69 Alberta^69 Manitoba^86 Nova Scotia^89 Saskatchewan^90 New Brunswick^95 Prince Edward Island^106 Newfoundland^111  NOTES: 31% of Rankings changed. 3% of Rankings changed 3 or more places.  lain Brown JAPANESE MANUFACTURING GREENFIELD& The Provincial Location Decision ^page 26  5. OUR CONDITION LOGIT ECONOMETRIC MODEL  Building on related research (Section 3) our model also seeks to explain the location selection process, by determining which factors (from our list in Section 4.2) were significant. That is, why did a Japanese manufacturing greenfield select one province over another, after having made the decision to locate in Canada? The structure of our model enables us to analyze the fifty-five Japanese manufacturing greenfields that have located in Canada during 1980 to 1991.  While in theory these firms could have selected any one of ten provinces and two territories, as we know from Table 4.1 and Appendix 1, these investments only involved a total of six provinces. Recall that almost two thirds of the new investment went to Ontario (35 investments), with British Columbia (11 investments) accounting for a fifth. This leaves Alberta with one investment, Saskatchewan with two investments, Quebec with four investments and New Brunswick with two investments.  Based on this it is not realistic to say that the whole of Canada represented a choice set, accordingly, we limited our choice set to only the provinces that had been picked in the last decade. (However, for comparative purposes, Table 4.2 and Appendix 2 contain provincial data on all ten provinces). Thus, there are 330 (6 X 55) dichotomous choices that stem from our 55 dependent variables. While there is no hard and fast rule on sample size, our  lain Brown JAPANESE MANUFACTURING GREENFIELDS: The Provincial Location Decision ^page 27  small sample size may be problematic; however, we have relied on McFadden (1974), who suggests that logit regression is suitable for sample sizes over 50.  We assume that a Japanese manufacturing greenfield will choose to invest in a particular province if and only if it will maximize profit. Formally, the jth province is chosen by the i th firm, if and only if  lIy = max {IL; m =^  (1)  where Hy denotes the profit of the i th firm given that it locates in the jth province (for j = 1,...,6). Following Carlton (1983) and Coughlin (1991), we assume that  Ilw = c + Xift + E y^(2)  where c is the constant term; X, is a vector of observable characteristics for the jth province; ft is a vector of unknown coefficients to be estimated; and E y is the random term. If the error term is independent and has a Weibull distribution, McFadden (1974) shows that  P = exp {X,J6} / E exp {X,fi}^(for k = 1,...,6)^(3)  whereft is obtained by maximum likelihood estimation; and Pi denotes the probability a Japanese manufacturing greenfield locating in province j. This decision depends on the level of the  lain Brown JAPANESE MANUFACTURING GREENFIELDS: The Provincial Location Decision ^page 28  province's characteristics that affect profits relative to the other provinces. In general, the explanatory variables are provincial characteristics which are independent of the investment.  Canadian Industries, Japanese Greenfields, Labor and Energy, however, contain  provincial characteristics that are specific to each investment, and the variation is due to industry differences.The first two variables were introduced in Section 4. Labor and Energy are two additional variables we use to measure factor intensities. We personalize the data for each Japanese establishment (based on work by Schmenner, 1987) by interacting industry labour (energy) intensity and wages (energy) costs in each province. Table 5.1 illustrate the calculation use to arrive at our Energy variable. The Energy Intensity column for firm 10 is pulp industry fuel costs divided by pulp industry sales. To determine Energy, Energy Intensity is multiplied by the Average Energy price in each province prevailing at the time the greenfield is established. The calculation to arrive at our variable to measure labour intensity (Labor) uses the same methodology. To calculate Labor, Wage Intensity is multiplied by Wages prevailing at time of entry in each province.  Table 5.1 also shows the variation in our data. For example, we see that energy  prices vary among provinces (i.e. Alberta's Average Energy cost in 1990 was half of Quebec's 1990 Average Energy cost). Variation also occurs over time (i.e. from 1986 to 1990 Quebec's Average Energy cost increased by $.61 per 1,000,000 Btu's, whereas the Average Energy cost  in Alberta decreased by $.26). In addition to these time and provincial variations that are occurring throughout our model, we also experience industry variation amongst Canadian  lain Brown JAPANESE MANUFACTURING GREENFIELDS: The Provincial Location Decision ^page 29  Industries, Japanese Greenfields, Labor and Energy. In this regard Table 5.1 points out the  number of pulp mills in British Columbia is greater than the number in Ontario. This is an example of industry-specific variation.  TABLE 5.1  ENERGY INTENSITY (In Thousands)  Firm #  SIC #  Energy Intensity  10  2711  355/5817  Mitsubishi 1990  Pulp  =0.06103  28  3231  100/39093  Honda 1986  Auto  =0.00256  Province  Avg. Energy $/MBtu  Energy  Cdn Indust. Count  Alberta B.C. N.B. Ontario Quebec Sask.  3.87 5.74 8.61 6.81 7.51 5.05  0.2362 0.3503 0.5255 0.4156 0.4583 0.3082  2 16 6 5 8 1  Alberta B.C. N.B. Ontario Quebec Sask.  4.13 5.89 8.04 6.32 6.90 5.28  0.0106 0.0151 0.0206 0.0162 0.0177 0.0135  0 3 0 15 5 1  Finally Table 5.1 shows how our model works. Mitsubishi can chose from six locations, and while Honda will also chose from the same six provinces, between the two investments there are a total of twelve alternatives. Honda and Mitsubishi face different location choices because of different investment times and the investments being in different industries. In our next section we reveal which provincial characteristics influenced the provincial location decision of the Japanese manufacturing greenfields.  lain Brown JAPANESE MANUFACTURING GREENFIELDS: The Provincial Location Decision  ^page 30  6. RESULTS  The following final results were obtained by estimating equation (3).  TABLE 6.1 FINAL RESULTS LOG OF LIKELIHOOD FUNCTION NUMBER OF CASES NUMBER OF CHOICES Parameter^Estimate Manufacturing Wages -8.35985 -7.11482 Average Energy -.350917 Ship Rank Canadian Industries .657384 Japanese Greenfields 2.16394 Provincial MetGDP -.880592  -55.0821 55 330 Standard Error 5.35172 2. 83188 .204142 .281352 .612262 .340797  t-statistic -1.56209 -2.51241 * 5 -1.71898 2.33652 * 3.53434 * -2.58392 *  The log of each variable was taken, with exception of the Ship Rank variable. Testing the theoretical relationships presented in the previous section was severely constrained by the limited sample size. The variables presented in Table 6.1 are those that had consistent effects across various specifications. To narrow down our list of twenty independent variables to the model above, the variable's significance was measured in terms of its coefficient and tstatistic (10 % significance level). For example, recall that as a measure of transportation services we had collected provincial data on Highway Kilometres, Airports, and Ship Rank variable. We selected Ship Rank because throughout our testing it remained generally significant, while the other two were not significant.  5  ^Asterisked variables are statistically significant.  lain Brown JAPANESE MANUFACTURING GREENFIELDS: The Provincial Location Decision ^page 31  To a manager both the lowest available energy cost and the average energy costs are important. The choice between the two measures of energy cost was not intuitively obvious because both could be a guiding force for new plant location. To select the most appropriate measure of energy costs it was necessary to independently test both of our energy cost measures. As Table 6.2 illustrates, when Low Energy replaced Average Energy, the model was not improved. Thus, average energy prices are more influential in a location decision than the price of the lowest available energy source.  TABLE 6.2 SUBSTITUTING LOW ENERGY PRICE LOG OF LIKELIHOOD FUNCTION NUMBER OF CASES NUMBER OF CHOICES Parameter^Estimate Manufacturing Wages -3.58545 -3.14202 Low Energy -.068488 Ship Rank -.750461 Provincial MetGDP .710716 Canadian Industries Japanese Greenfields 1.92872  -55.9915 55 330 Standard Error 4.25787 1.48912 .135675 .324385 .272957 .592186  t-statistic -.842076 -2.10999 * -.504793 -2.31349 * 2.60377 * 3.25695 *  Recall Table 4.2 where we found that the variables that changed the most in provincial ranking from 1980 to 1989 were Tax, Unionization, and Unemployment. We did not examine the influence of unemployment, given the previously discussed ambiguity of high unemployment as a location factor (i.e. , being a drawing card for some, yet a deterrent to others). However, we decided to test the provincial crime rate, as safety is considered to be important to the Japanese. Our Crime variable was offenses per 100,000 divided by provincial population.  lain Brown JAPANESE MANUFACTURING GREENFIELDS: The Provincial Location Decision ^page 32  TABLE 6.3  TESTING UNIONIZATION, TAX & CRIME RATES LOG OF LIKELIHOOD FUNCTION NUMBER OF CASES NUMBER OF CHOICES Parameter^Estimate Manufacturing Wages -11.2525 -8.59998 Average Energy -.360553 Ship Rank .651787 Canadian Industries 2.18708 Japanese Greenfields -1.27346 Provincial MetGDP .033940 Unionization .179514 Tax -14.3812 Crime  -54.1813 :^55 330 Standard Error 6.69886 3.25604 .397184 .289458 .806161 1.39549 .089459 .165387 31.7353  t-statistic -1.67977 -2.64124 * -.907774 2.25175 * 2.71296 * -.912557 .379387 1.08542 -.453161  Unlike Bartik (1985), who found tax and unionization to be significant in the United States, as Table 6.3 illustrates, we did not find this result. Nor did Crime have any influence in the location decision of Japanese manufacturing greenfields. Thus, we could not support the stereotype that high rates of unionization, crime and taxes drive business away.  However, our results show that Wages and Average Energy are negative and fairly significant throughout our testing. At first this may appear to contradict our findings in Table 4.2. In Table 4.2 we see that Ontario and British Columbia have the highest wages and neither  has the lowest energy prices, yet they account for 84% of new Japanese manufacturing investment in Canada in the last decade. This is not a contradiction, because while Japanese greenfields prefer to avoid high wages and energy costs, they have an even stronger preference for like industries and presence of other Japanese greenfields. Thus, they are willing to pay the higher price for inputs in order to be in areas of industry specific concentration.  lain Brown JAPANESE MANUFACTURING GREENFIELDS: The Provincial Location Decision ^page 33  Next we tested to see whether labour intensive industries were more concerned about the level of wages, and energy intensive industries were more concerned about energy costs. As portrayed in Table 6.4 inconsistent results were found in that neither the Labor or Energy variable were significant. In our sample, it appears that firms in labour or energy  intensive industries tended to choose Ontario or British Columbia in spite of high wages and mid-priced energy costs.  TABLE 6.4  TESTING FACTOR INTENSITIES LOG OF LIKELIHOOD FUNCTION NUMBER OF CASES NUMBER OF CHOICES Parameter^Estimate Labor 12.47260 Energy -3.207280 Ship Rank .025374 Provincial MetGDP -.500709 Canadian Industries .690563 Japanese Greenfields 1.15395  -59.0074 :^55 330 Standard Error 12.5914 30.5082 .129654 .288151 .270134 .401528  t-statistic .990568 -.105128 .195703 -1.73766 2.55637 * 2.87389 *  In our section on independent variables we explained how we calculated MetGDP. We theorized that Japanese investors prefer large, concentrated markets, so we created this variable by multiplying the percentage of provincial population living in a city with provincial GDP. As Table 6.1 illustrated the coefficient for MetGDP is negative and this counter-intuitive  relationship also occurs for GDP (see Appendix 6).  Re-examining the data suggested that Quebec might be unduly influencing the model since it has the second highest GDP and Metropolitan population (see Table 4.2 and  lain Brown JAPANESE MANUFACTURING GREENFIELDS: The Provincial Location Decision ^page 34  Figure 6.1) yet has only  had four Japanese greenfield investments in the last decade. To explore this, we created a Quebec dummy variable. The results in Table 6.5 were inconclusive in that when  -A=*=/t=# 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990  Quebec replaced the MetGDP variable it too Figure 6.1  was negative and significant; however, when both the Quebec dummy variable and the MetGDP variable were included in the model, as illustrated by Table 6.6, neither was significant.  TABLE 6.5 REPLACING MetGDP WITH QUEBEC DUMMY LOG OF LIKELIHOOD FUNCTION NUMBER OF CASES NUMBER OF CHOICES Parameter^Estimate -5.12543 Average Energy Manufacturing Wages -10.5204 Canadian Industries .523389 1.10012 Japanese Greenfields -.485113 Ship Rank Quebec Dummy -1.42687  -56.1173 55 330 Standard Error 2.6080 5.71940 .271664 .424306 .217339 .677821  t-statistic -1.96527 -1.83942 1.92660 2.59274 * -2.23206 * -2.10509 *  Lain Brown JAPANESE MANUFACTURING GREEIVFIELDS• The Provincial Location Decision  ^  page 35  TABLE 6.6  ADDING A QUEBEC DUMMY VARIABLE LOG OF LIKELIHOOD FUNCTION NUMBER OF CASES NUMBER OF CHOICES Parameter^Estimate Manufacturing Wages -8.71039 -7.00878 Average Energy Ship Rank -.366620 .648374 Canadian Industries Japanese Greenfields 2.08530 Quebec Dummy -.157868 Provincial MetGDP -.816484  -55.0720 55 330 Standard Error 5.89270 2.92989 .232202 .28773 .822973 1.10961 .564654  t-statistic -1.47817 -2.39216 * -1.57889 2.25341 * 2.53387 * -.142273 -1.445990  Intuitively the explanation is that the Japanese are merely following a Canadian trend of avoiding Quebec due to government intervention into areas such as language legislation. Furthermore, one would suspect that a Japanese greenfield may be even more concerned about the separation issue than a Canadian firm. This would be particularly true if the Japanese FDI is made in order to tariff jump. Locating in the "Country of Quebec" may not automatically mean that the industry can benefit from existing free trade agreements within North America. While we can not confirm this, we can confirm that Japanese manufacturing greenfields have located in Ontario, which is close to Quebec's wealthy urban population, yet out of its jurisdiction.  lain Brown JAPANESE MANUFACTURING GREENFIELDS: The Provincial Location Decision ^page 36  7. CONCLUSION  In this section we summarize our results and identify the main weaknesses of our analysis. Then we discuss policy ramifications and research questions that stem naturally from our results.  7.1. Factors Influencing Japanese Manufacturing Greenfield's Location Decision  We found that new Japanese manufacturing facilities locate next to similar Canadian industry-specific firms and other Japanese firms. Ontario and British Columbia attract a greater proportion of Japanese manufacturing than their national share of manufacturing would suggest. In this regard our data shows that Ontario has approximately half of Canadian manufacturing, yet they have attracted 64% (35/55) of Japanese greenfields. The case in British Columbia is even more profound. The number of new Japanese manufacturing greenfields (11/55 or 20%) is double British Columbia's share of national manufacturing. Thus, we conclude that industry-specific establishment counts are a more important determinant of a Japanese manufacturing greenfield's location than the general level of provincial manufacturing.  Industrial distribution of Canadian firms is not the only factor underlying the location choice. Our results show that Japanese firms prefer provinces where there are relatively more Japanese manufacturing greenfields. That is to attract Japanese industry the presence of  lain Brown JAPANESE MANUFACTURING GREENFIELDS: The Provincial Location Decision ^page 37  similar Canadian industry (i.e., with the same SIC classification) is necessary and once Japanese industry is attracted then this in itself is a factor in attracting other industries from Japan. Furthermore, if two provinces share similar levels of industry-specific Canadian firms then geographic proximity to Japan may be important. To summarize, Canadian Industries and Japanese Greenfields were always significant, with Ship Rank generally being significant. Apart  from these three variables, Average Energy, Wages, and MetGDP were also part of the equation; however, their importance is minor in determining where Japanese manufacturing greenfields will locate in Canada.  This brings us to the question of whether the positive coefficient on the Canadian Industries variable reflects an agglomeration effect or an endowment effect. Greenfields in  Ontario tend to be in the automotive sector, which follows a traditional Canadian pattern. However, while both British Columbia and Ontario have high counts of forestry specific industries the Japanese have elected to locate mostly in British Columbia presumably due to its easier access to the Japanese market. In short, it is the only province with trees across the water from Japan. This ease of transportation is captured by the significance of our Ship Rank variable, which measures the province's water access to Japan. Referring back to Table 4.2 we see that British Columbia's geographic position gives it the comparative advantage over Ontario's geographic position. Head, Ries, and Swenson (1993) suggested that the concentration of Japanese wood product firms in Washington State is the result of the state's forests and 6  6^  Washington State is British Columbia's southern neighbour, and both are bounded by the Pacific Ocean on the west.  lain Brown JAPANESE MANUFACTURING GREENFIELDS: The Provincial Location Decision ^page 38  Washington's geographic proximity to Japan. Thus, endowments--British Columbia's forests and proximity to Japan--may explain the concentration of Japanese forestry investment in British Columbia.  It is difficult to distinguish if clustering is due to agglomeration or endowment. That is, are Japanese pulp mills in British Columbia because it is the only province with trees across the water from Japan (an endowment effect) or because the province of British Columbia has more pulp mills than any other Canadian province (which could be due to agglomeration or an endowment effect)? While the answer is not intuitively obvious, our results may suggest evidence of pure agglomeration effects. That is, if all endowment effects are captured by the combination of the Canadian Industries and Ship Rank variables, then the positive coefficient on the significant Japanese Greenfields variable may be evidence of pure agglomeration effects.  7.2. Limitations of Our Results  The small number of Japanese manufacturing greenfields that have entered the Canadian economy in the last decade limits our ability to assess the determinants of investment. The many possible explanatory variables and little information to distinguish among them (especially in the presence of multicollinearity) adds further to the difficulties in determining actual location factors that attract Japanese manufacturing greenfields.  lain Brown JAPANESE MANUFACTURING GREENFIELDS: The Provincial Location Decision ^page 39  The possibility of omitted explanatory variables exists. Utilizing past research we identified and then tested the vast majority of factors that could influence the location decision. However, it is possible that our error term includes an independent variable that we were not able to identify and test for its significance. If we have omitted variables that are correlated with other independent variables, the estimated coefficients for those variables will be biased.  Another possible problem with our model may be lack of variation amongst non industry-specific provincial characteristics. Table 4.2 illustrates that there is variation among these characteristics. However, the range of variation between these characteristics is not always, from a statistical point of view, as great as we would like it to be. This limited variation of non industry-specific provincial characteristics constrains our ability to distinguish the individual effects of each variable.  With the combination of Ontario (64 %) and British Columbia (20 %) receiving 84 % of Japanese greenfield investment our distribution is skewed. This skewed distribution compounds the problem of our limited variation. Furthermore, this distribution also gives rise to the possibility that a unique omitted feature of either Ontario or British Columbia could be driving the results. Nevertheless, by measuring variation in both non industry-specific provincial characteristics and industry-specific characteristics, despite our small skewed sample size, we feel that our study still provides results that provincial policy makers may find useful.  lain Brown JAPANESE MANUFACTURING GREENFIELDS: The Provincial Location Decision ^page 40  7.3. Policy Ramifications for Provinces Seeking FDI  Regardless of whether we think foreign investment is good or bad for Canada it has become somewhat of a necessity given our desire to spend more than we earn. Accordingly, we find ourselves in the position of needing to attract foreign investment. Because of possible job creation and potential spillovers of management expertise, FDI has been highly sought after in North America. In the U.S. large subsidies have been paid by the individual states trying to attract Japanese greenfield investments.  However, our study suggests that unless Japanese manufacturing greenfields and industry-specific Canadian firms are already in the province, and that the province provides market accessability, the province is unlikely to attract much Japanese investment. Hence, increasing schemes directed at attracting FDI will not have a significant effect if the industryspecific base of firms does not already exist.  If the investment is to export products to Japan, then market access to Japan is determined by geography (a factor not easily modified). For example, although both Ontario and British Columbia have a forest industry, because the products being produced are for Japanese consumption, Japanese-owned plants tend to locate in British Columbia. If the market is Canada (or North America) then location is in Ontario with its central location and substantial industrial activity.  lain Brown JAPANESE MANUFACTURING GREENFIELDS: The Provincial Location Decision ^page 41  Our results indicate that it is difficult for a province to try to attract Japanese greenfield FDI without a market for the product, or without an existing industrial base. For a province to try to attract this type of FDI with neither, it is near impossible, unless governments intervene. However, even government intervention may do little to encourage location preferences. Our study suggests that corporate tax rates and unionization were not significant. Therefore, rather than offer subsidies or direct government resources to reducing unionization, perhaps tax dollars would be better spent on developing an industrial strategy to encourage clustering of industry.  If tax dollars are efficiently spent on needed infrastructure that creates a competitive advantage, then this may encourage clustering of firms that benefit from this infrastructure. Should Canadian industries cluster, then Japanese manufacturing greenfields in the same industry are more likely to be attracted to this area. With a base of Japanese manufacturing greenfields, more Japanese manufacturing greenfields are likely to be attracted. This possible Japanese agglomeration effect could then create industry diversification.  7.4. Further Study and Extensions  This study questions government efforts to attract Japanese greenfield investments to areas that do not already have a base of firms in similar industries or Japanese manufacturing greenfields. Our study also drew our attention to a possible free trade effect, in that we noticed that 75% of greenfields locating in B.C. during the last decade did so after free trade. Contrary  lain Brown JAPANESE MANUFACTURING GREENFIELDS: The Provincial Location Decision ^page 42  to British Columbia, only 25% of Ontario's Japanese manufacturing greenfields located in Ontario after free trade. While this may be a free trade effect, Ontario's relative fewer investments could reflect the change in business climate due to the more socialist government being elected in 1990. We feel it is also worth studying why the majority of Japanese manufacturing greenfields locate in Ontario, which is close to Quebec's wealthy urban population, yet out of its jurisdiction. This too may be a free trade effect in that the Japanese manufacturing greenfields wish to have access to North American free trade, which is something that is not guaranteed if Quebec should separate from Canada. Alternatively, the possible avoidance of Quebec could be a language preference.  Future work will take into account non-free trade issues such as language or changing political climates so that we may study whether the Canada/U.S. Free Trade Agreement has shifted FDI amongst the provinces and/or away from Canada to the U.S. We plan to study this by comparing Japanese manufacturing greenfields locating in Canada to those locating in the U.S. We will seek to not only determine whether one country has attracted relatively more Japanese manufacturing greenfields than the other, but also to determine if industry location is due to comparative advantage as FDI for tariff jumping is no longer required within the new North American trade zone.  lain Brown JAPANESE MANUFACTURING GREENFIELDS: The Provincial Location Decision ^page 43  8. BIBLIOGRAPHY  T. J. Bartik, "Business Location Decisions in the United States: Estimates of the Effects of Unionization, Taxes and Other Characteristics of States," Journal of Business & Economic Statistics, vol. 3, no. 1, pp. 14-22. 1985. C.C.H. Canadian Limited, Canadian Tax Reporter, vol. 1, pp. 1602-1603. 1992. D. W. Carlton, "The Location and Employment Choices of New Firms: An Econometric Model With Discrete and Continuous Endogenous Variables," Review of Economics and Statistics, vol. 65 pp. 440-449. 1983. C. Coughlin, et. al., "State Characteristics and the Location of Foreign Direct Investment with the United States," Review of Economics and Statistics, pp. 675-683. 1991. Dodwell Marketing Consultants, Auto Parts Industry of Japan (Tokyo: 1989). External Affairs Canada, Canada - U.S. Free Trade Agreement. 1988. N. Glickman & D. P. Woodward, "The Location of Foreign Direct Investment in the United States: Patterns and Determinants," International Regional Science Review, vol. 11, no. 2, pp. 137-154. 1988. E. M. Graham & P. Krugman, Foreign Direct Investment in the United States (Washington D.C.: Institute for International Economics 1990). L. C. Hamilton, Regression with Graphics: A Second Course in Applied Statistics, (Pacific Grove, California: Brooks/Cole Publishing 1992). K. Head, J. Ries, and D. Swenson, "Agglomeration Benefits and Location Choice: Evidence from Japanese Manufacturing Investments in the United States," University of British Columbia Working Paper. March 1993. Investment Canada, Annual Report. 1991. Investment Canada, International Investment's Canadian Developments in a Global Context. 1991. Japan, Japanese Direct Investment Aboard in Fiscal 1990. Ministry of Finance. 1991. JETRO, Directory: Japanese-Affiliated Companies in USA & Canada. 1991-1992.  lain Brown JAPANESE MANUFACTURING GREENFIELDS: The Provincial Location Decision ^page 44  M. I. Luger & S. Shetty, "Determinants of Foreign Plant Start-ups in the United States: Lessons for Policymakers in the Southeast," Vanderbilt School of Transnational Law, vol. 18, pp. 223-245. 1985. P. Krugman "Geography and Trade," Cambridge, MA: The MIT Press. D. McFadden, "Conditional Logit Analysis of Qualitative Choice Behaviour," in P. Zarembka, ed., Frontiers in Econometrics (New York: Academic Press 1974). D. McFadden, "Cost, Revenue, and Profit Functions," in D. McFadden and M. Fuss, eds., Production Economics: A Dual Approach to Theory and Applications, vol. 1 (Amsterdam: North-Holland 1978). R. J. Newman & D. H. Sullivan, "Econometric Analysis of Business Tax Impacts on Industrial Location: What Do We Know, and How Do We Know It?" Journal of Urban Economics, vol. 23, pp. 215-234. 1988. Ontario, Canada, Japanese Investment Profile. 1992. Roads and Transportation Association of Canada, Canada's Roadway Infrastructure Selected Facts and Figures, (Ottawa, Canada 1990). A. M. Rugman, International Business in Canada Strategies for Management (Scarborough: Prentice-Hall 1989). R. W. Schmenner, J. C. Huber, and R. L. Cook, "Geographic Differences and the Location of New Manufacturing Facilities," Journal of Urban Economics, vol. 21, pp. 83-104. 1987. D. Smith & R. Florida, "Agglomeration and Industry Location: An Econometric Analysis of Japanese-Affiliated Manufacturing Establishments in Automotive-Related Industries," H. John Heinz III School of Public Policy and Management (Carnegie Mellon University) Working Paper. 1992. Statistics Canada, "CALURA," Catalogue 71-201 Annual. 1991. Statistics Canada, "Canada Year Book," Annual. 1991. Statistics Canada, "Canada's International Investment Position." 1991. Statistics Canada, "Cansim Data Retrieval Program," (On-Line Database System, 1993). Statistics Canada, "Consumption of purchased fuel and electricity," Catalogue 57-208 Annual. 1991.  lain Brown JAPANESE MANUFACTURING GREE1VFIELDS: The Provincial Location Decision ^page 45  Statistics Canada, "Energy Statistics Hand Book," Catalogue 57-601 Quarterly. September 1992. Statistics Canada, "Manufacturing industries of Canada: national and provincial areas," Catalogue 31-203 Annual. 1991. Statistics Canada, Pacific Region Advisory Services, "Inquiry for Union Statistics," February 9, 1993. Statistics Canada, "Quarterly report on energy supply-demand in Canada," Catalogue 57-003 Quarterly. June 1992. Toyo Keizai Company, Japanese Overseas Investment (Tokyo: 1990). Transport Canada, "Airports Group - Airport Site Directory." 1993. United Nations, Foreign Direct Investment and the Service Sector and International Banking, the United Nations Centre for Transnational Corporations. 1991. United Nations, World Investment Report: the Triad in Foreign Direct Investment, the United Nations Centre for Transnational Corporations. 1991. D. P. Woodward, "Locational Determinants of Japanese Manufacturing Start-Ups in the United States," Southern Economic Journal. January, 1992.  APPENDIX 1 FIFTY-FIVE JAPANESE MANUFACTURING GREENFIELDS IN CANADA (Dependent Variable)  PROV  YEAR  FIRM #  CANADIAN COMPANY  JAPANESE INVESTOR  OWNERSHIP  SIC CODE  #CDN FIRM IN SIC  100.0  2512  1093  PRODUCT DESCRIPTION  AB  1990  1  Tomen Alberta Timber Industries  Toyo Menka Kaisha  BC  1989  2  Advanced Energy Technology  NTT  45.0  3391  27  R&D rechargeable batteries  BC  1989  3  Atsugi Nylon Canada Inc  Atsugi Nylon  24.0  1811  31  Nylon  BC  1991  4  Campbell River Fibre Ltd  C Itoh and Co Ltd  90.0  2512  1093  BC  1983  5  Canadian Autoparts Toyota Inc  Toyota Motor Corp  100.0  3255  47  BC  1988  6  Canadian Chopstick Mfg Co Ltd  Mitsubishi Corp  100.0  2599  272  Chopsticks  BC  1980  7  Daiwa (Canada) Ltd  Daiwa Seiko Inc  3931  209  Golf clubs  BC  1983  8  Dominion Malting Ltd  Sumitomo Corp  35.0  1131  48  BC  1991  9  I.S. Forest Products  Inland Kogyo  17.0  2512  1093  BC  1990  10  M.C. Forest Investment  Mitsubishi Corp  100.0  2711  39  Pulp  BC  1988  11  Primex Fibre Ltd  Sanyang Pulp  50.0  2711  39  Pulp, chips  BC  1989  12  S.M. Cyclo of Canada  Sumitomo Heavy Ind.  3199  772  Speed reducers & variators, motors  NB  1989  13  Ampal Pallets Inc  Mitsui & Co Ltd  57.9  3099  463  Steel pallets  NB  1980  14  NBIP Forest Products Inc  Oji Paper Co/Mitsui & Co  33.0  2712  ON  1987  15  ABC Nishikawa Industries  Nishikawa Kasei Co Ltd  49.0  3256  96  ON-OP  1988  16  Bellemar Parts Ind. Canada  Honda Motor Co Ltd  100.0  3259  190  ON  1986  17  CAMI Automotive Inc  Suzuki Motor Co Ltd  50.0  3231  27  - 46 -  Lumber  Woodchips Aluminum wheels  Liquor malt Forest product  Newsprint Plastic autoparts & armrests Seats for vehicles & tire assembly Vehicles  ON-OP  1989  18  Canada Mold Technology Inc  Nagase Ltd  100.0  3062  539  Prototype molds  ON-OP  1990  19  Cangel Inc  Nitta Gelatin Inc  100.0  1011  526  Gelatin & lard  ON  1987  20  Copar International  Toyo Radiator Co Ltd  46.0  3251  50  Radiator & oil coolers  ON  1988  21  DDM Plastics Inc  Daikyo/Suzuki/Mitsui  100.0  3256  96  Automotive plastics  ON-OP  1985  22  Denon Canada Inc  Nippon Columbia Co Ltd  95.0  3341  25  Car stereo, cassette & CD software  ON  1990  23  DNN Galvanizing Corp  Nippon Kokan (NICK)  40.0  2912  26  Hot dip galvanizing steel sheets  ON  1981  24  Epson Canada Ltd  Seiko Epson Corp  19.0  3361  147  Computers, printers, software products  ON  1980  25  Epson Manufacturing Ltd  Seiko Epson Corp  3361  147  Printers, ribbons & technical products  ON  1980  26  F&P Mfg Inc  F. Tech Inc  55.0  3257  20  ON  1987  27  General Seating of Canada Ltd  NHK Spring Co  65.0  3259  190  ON  1986  28  Honda of Canada Mfg Inc  Honda Motor Co Ltd  100.0  3231  27  Automobiles  ON  1988  29  Inoac Canada Ltd  Inoue MTP Co  50.0  3257  20  Automotive interior panels & armrests  ON  1990  30  IDS Fitel Inc  Furukawa Electric Co Ltd  50.0  3562  153  ON  1986  31  Kao-Didak Ltd  Kao Corporation  93.0  3399  72  Floppy disks  ON  1984  32  Kuriyama Canada Ltd  Kuriyama Corp  100.0  1621  77  Industrial plastic hose & plumbing  ON  1983  33  Mitsubishi Electronic Ind  Mitsubishi Electric Corp  100.0  3341  25  Colour cathode ray tubes  ON  1987  34  Miura Boiler Co  Miura Boiler Co  99.3  3011  43  High pressure steam boiler  ON  1980  35  Murata Erie North America Ltd  Murata Mfg Co Ltd  100.0  351  90  Ceramic capacitors  ON-OP  1987  36  Nichirin Inc  Nichirin Co Ltd  100.0  3259  190  Hydraulic hoses for autos & motorcycles  ON  1985  37  NKC of Canada Inc  Nakanishi Metal Works  100.0  3192  533  Conveyor systems  ON  1986  38  Quality Safety Systems Co  Tokai Rika Co Ltd  40.0  3259  190  Seat belts & auto components  ON  1986  39  Rockwell Int'l Suspension Syst  Mitsubishi Corp  40.0  3254  35  ON  1982  40  Sanyo Cdn Machine Works Inc  Sanyo Machine Works Inc  100.0  3081  1464  - 47 -  Auto part, pedal bracket Seats for automobiles  Passive fibre optic components  Coil springs & torsion bars Automatic assembly & welding machine  ON-OP  1985  41  SM Yttrium Canada Ltd  Shin-Etsu Chemical Co  ON  1987  42  SMC Pneumatics Canada Inc  SMC Corporations  ON  1986  43  Toyota Motor Mfg Canada Inc  Toyota Motor Corp  ON  1985  44  Trutech Canada Inc  Nihon Parkerizing Co  ON  1989  45  UCAR Carbon Canada Inc  Mitsubishi Corp  ON  1980  46  UNIC International Corp  ON  1987  47  Vdo-Yazaki Ltd  Yazaki Corporation  ON  1985  48  Woodbridge Inoac Inc  Inoue MTP Co  ON  1989  49  Yachiyo of Ontario Mfg Inc  Yachiyo Industries  PQ  1986  50  Cree Yamaha Enterprise Ltd  Yamaha Motors  PQ  1986  51  H Aida Enterprise Inc  PQ  1989  52  PQ  1989  SK SK  3731  94  Silicon  3092  44  Cylinders & valves  3231  27  Automobiles  3041  288  3399  72  319  1601  50.0  391  603  50.0  3257  20  100.0  3259  190  Fuel tanks  40.0  3281  327  FRP boats  Tokia  100.0  1021  414  Processing seafood  Kobe Aluminium Canada  Kobe Steel Ltd  100.0  2961  71  Aluminium  53  Miura Boiler Company  Miura Company  3011  43  Boilers  1988  54  Hitachi Canadian Ind Ltd  Hitachi Ltd  100.0  337  281  Electric power equipment  1987  55  SK Turbine Ltd  Marubeni Corporation  100.0  3194  116  Turbines  NOTE: "-OP" signifies that this is the year of operation, rather than year of establishment  100.0  100.0  50.0  Paint finishing system, rolling oil conc Artificial graphic electrodes Sandblasting equipment Meters Automotive interior trims  APPENDIX 2 PROVINCIAL CHARACTERISTICS (Independent Variables) INDEPENDENT VARIABLE  COLUMN POP % Metro Crime UI Union % Area MAN$/CAP GDP/CAP Highway KM/SQ.ICM  Population in 1000's Percentage of Population Living in Metropolitan Areas Reported Offenses per 100,000 People Unemployment Rate Unionization Rates Land Area (excluding fresh water) Square KM (constant 1980-90) Average Annual Manufacturing Pay per Worker Provincial GDP per Capita Highway Kilometres per Square Kilometre of Area (Constant 1980-90)  AVG NRG $/MBtu  Weighted Average Cost of Crude Oil, Natural Gas, and Electricity Prices per Million Btu's  Low NRG $/MBtu  Lowest Cost of Either Crude Oil, Natural Gas, or Electricity Prices per Million Btu's  AIRPORT  Number of Airports with Control Towers  TAX  Provincial Corporate Tax Rate  % PROV LAB in MFG  Percentage of Work Force in Manufacturing (Manufacturing Labour Force/Total Labour Force)  % CDN MFG LAB/PROV  Percentage of Canadian Manufacturing Work Force in the Province (Provincial Manufacturing Labour Force/Total Canadian Manufacturing Labour Force)  Ship Rank  Shipping 1=BC (yearly deep water; no canal) 2=NS, NB, NFL (yearly deep water) 3 = Que (mostly open deep water) 4= Ont (mostly open shallow water) 5 =Man (partly open deep water) 6 =PEI (deep water but no port facilities) 7 = Ab, Sask (no water access)  APPENDIX 2 CONT'D PROV YEAR  %METRO  1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990  POP 000's 2666.0 2744.2 2787.7 2813.8 2847.7 2870.1 2889.0 2925.0 2980.2 3048.3 3132.5  AB AB AB AB AB AB AB AB AB AB AB  1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990  SK SK SK SK SK SK SK SK SK SK SK MN MN MN MN MN MN  54.4 54.0 54.5 54.8 55.2 55.5 56.7 57.2 57.8 58.3 58.3  CRIME RATE 141.77 150.63 161.03 156.34 155.39 154.99 161.88 165.63 160.57 166.60 177.51  UI % 6.8 6.7 12.1 13.8 14.7 14.1 12.5 11.9 10.3 9.1 8.3  UNION % 39.2 40.0 40.0 40.4 38.3 36.9 41.2 37.5 38.1 36.4 38.0  2140.6 2237.3 2314.5 2338.7 2338.5 2348.5 2375.1 2377.7 2388.7 2425.9 2473.1  60.7 58.1 56.4 56.1 55.9 55.7 61.4 61.7 62.1 62.3 62.6  153.14 157.16 149.12 145.16 129.87 128.19 134.52 145.05 148.41 143.24 145.49  3.7 3.8 7.7 10.6 11.1 10.0 9.8 9.6 8.0 7.2 7.0  22.0 23.3 22.7 23.9 23.6 22.7 26.4 24.8 26.1 25.8 26.6  644390 644390 644390 644390 644390 644390 644390 644390 644390 644390 644390  19003 21466 24248 25989 27241 27864 29027 29132 29453 30638 31000  20156.97^0.265^2.67 22318.87^0.265^3.48 22854.18^0.265^3.84 23682.39^0.265^4.31 25204.62^0.265^4.56 27826.70^0.265^4.61 24132.46^0.265^4.13 25051.52^0.265^4.24 26059.36^0.265^3.87 27059.24^0.265^3.79 28533.02^0.265^3.87  1.36^7 1.92^7 1.96^7 2.16^7 2.07^7 2.13^7 2.12^7 2.14^7 1.86^7 1.40^7 1.30^7  11.0 11.0 11.0 11.0 11.0 11.0 11.0 15.0 15.0 15.0 15.0  7.25 7.21 6.47 5.86 5.76 5.98 6.03 6.17 6.72 7.02 7.10  4.39 4.66 4.64 4.33 4.33 4.23 4.22 4.20 4.46 4.66 4.69  7 7 7 7 7 7 7 7 7 7 7  1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990  959.4 968.3 977.0 989.3 1000.5 1008.4 1010.2 1015.8 1013.5 1006.7 997.1  32.5 32.8 33.3 33.7 33.8 34.2 38.3 38.6 39.0 39.2 39.7  142.96 152.39 139.13 135.83 139.08 144.83 153.95 155.71 152.73 148.37 145.92  4.4 4.6 6.1 7.3 8.0 8.1 7.7 7.4 7.5 7.4 7.0  18.8 28.9 28.3 28.2 28.4 27.1 32.7 30.5 31.1 32.2 32.3  570700 570700 570700 570700 570700 570700 570700 570700 570700 570700 570700  17526 19690 21926 23495 25385 25272 25356 26141 27557 28379 29000  12924.74^0.341^3.05 14808.43^0.341^3.76 15107.47^0.341^4.54 15399.78^0.341^4.91 16381.81^0.341^5.26 17290.76^0.341^5.40 16971.89^0.341^5.28 16954.12^0.341^5.23 17945.73^0.341^4.54 19326.51^0.341^4.66 20332.97^0.341^5.05  1.70^5 2.28^5 2.90^5 3.12^5 3.17^5 3.15^5 3.15^5 2.72^5 2.52^5 2.49^5 2.90^5  14.0 14.0 14.0 14.0 16.0 16.0 17.0 17.0 17.0 15.0 15.0  4.85 4.82 4.37 4.03 3.98 3.97 3.95 4.04 4.19 4.43 4.50  1.15 1.16 1.17 1.13 1.14 1.09 1.07 1.06 1.05 1.08 1.07  7 7 7 7 7 7 7 7 7 7 7  1024.9 1026.2 1033.3 1045.6 1055.1 1064.0 1071.2 1079.0 1084.1 1086.3 1089.0  56.8 57.2 57.3 57.4 57.2 57.5 58.4 58.6 58.9 59.0 59.4  116.36 124.75 131.48 133.19 130.40 132.60 139.29 143.43 124.06 115.93 121.84  5.5 5.9 8.5 9.4 8.4 8.2 7.7 7.4 7.8 7.5 7.2  30.7 30.2 30.0 26.6 28.2 28.7 35.5 35.4 35.3 36.7 36.8  548360 548360 548360 548360 548360 548360 548360 548360 548360 548360 548360  15637 17546 19460 20393 21325 22157 23009 23775 24653 25926 27000  10916.19^0.154^4.22 12824.01^0.154^4.62 13562.37^0.154^5.03 14260.71^0.154^5.37 15657.28^0.154^5.65 16598.68^0.154^5.87 17196.60^0.154^5.49 18064.87^0.154^5.86 19895.77^0.154^5.88 21135.05^0.154^6.22 21769.51^0.154^6.85  2.25^4 2.81^4 3.48^4 3.67^4 3.69^4 3.63^4 3.40^4 3.26^4 2.94^4 2.84^4 2.76^4  15.0 15.0 15.0 16.0 16.0 16.0 17.0 17.0 17.0 17.0 17.0  12.00 11.67 10.78 10.13 10.00 9.93 10.05 10.18 10.70 10.55 10.40  3.14 3.08 3.13 3.07 3.11 2.91 2.92 2.90 2.94 2.88 2.91  5 5  MN MN MN MN  1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990  ON ON ON ON ON ON ON ON ON ON ON  1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990  8569.7 8624.7 8702.5 8798.0 8901.7 9006.4 9113.0 9265.0 9431.1 9589.6 9749.6  64.8 65.2 65.4 65.6 65.8 66.1 69.4 69.6 69.7 69.9 70.2  116.24 120.10 116.39 111.40 109.71 106.13 111.70 113.75 113.42 112.68 116.12  6.8 6.6 9.7 10.3 9.0 8.0 7.0 6.1 5.0 5.1 6.3  29.7 29.5 30.2 32.5 32.1 31.8 31.6 30.9 31.1 31.1 31.7  891190 891190 891190 891190 891190 891190 891190 891190 891190 891190 891190  18083 20251 22375 24179 27996 27009 27870 28743 30303 31559 32500  13418.67^0.187^3.83 15285.29^0.187^4.61 15778.22^0.187^5.42 17270.40^0.187^5.92 19265.87^0.187^6.18 20381.17^0.187^6.46 22244.05^0.187^6.32 24162.12^0.187^6.15 26758.07^0.187^6.23 28380.85^0.187^6.54 28420.14^0.187^6.81  2.45^16 3.13^16 3.87^16 4.33^16 4.33^16 4.31^16 3.56^16 3.68^16 3.24^16 3.51^16 3.46^16  14.0 14.0 14.0 15.0 15.0 15.0 15.5 15.5 15.5 15.5 15.5  20.92 20.44 18.81 18.39 17.38 19.16 19.23 19.18 19.41 19.53 19.70  49.19 49.20 49.75 50.20 49.03 51.60 51.70 51.36 51.04 51.69 51.34  4 4 4 4 4 4 4 4 4 4 4  BC BC BC BC BC BC BC BC BC BC BC  MN  AREA MAWCAP GDP/CAP HIGHWAY AVG NRG LOW NRG AIRPORT SQ.KM KM/SQ.KM $/MBtu $/MBtu^i 929730 22231 14343.21^0.071^3.17 1.64^26 929730 24388 16285.62^0.071^4.20 2.39^26 929730 27059 16542.31^0.071^5.32 3.29^26 3.31^26 929730 29242 17112.45^0.071^5.73 3.49^26 929730 29905 17950.98^0.071^6.14 929730 31389 18988.54^0.071^6.33 3.52^26 929730 3.31^26 31895 19848.39^0.071^5.89 929730 32375 21496.07^0.071^5.56 2.33^26 929730 33542 23325.28^0.071^5.44 2.36^26 929730 34841 25234.06^0.071^5.37 1.97^26 929730 35000 25764.09^0.071^5.74 2.28^26  - 50 -  TAX %PROVLAB%CDN MFG % IN MFG LAB/PROV 8.65 15.0 12.30 11.69 16.0 8.58 10.25 8.24 16.0 7.97 9.65 16.0 8.00 9.45 16.0 9.49 7.63 16.0 7.38 9.19 16.0 7.62 9.59 15.0 7.82 14.0 10.06 10.07 8.07 14.0 10.08 8.10 14.0  SHIP RANK  1 1 1 1 1 1 1 1 1 1 1  5  5 5 5 5 5 5 5 5  PQ PQ PQ PQ PQ PQ PQ PQ PQ PQ PQ  1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990  6386.1 6438.2 6462.2 6474.9 6492.0 6514.2 6540.2 6592.6 6640.8 6698.2 6768.2  61.5 61.3 61.4 61.7 61.8 61.9 63.4 63.4 63.6 63.7 63.9  77.80 81.52 78.66 73.88 74.29 77.42 78.23 79.77 78.93 78.50 85.18  9.8 10.3 13.8 13.9 12.8 11.8 11.0 10.3 9.4 9.3 10.1  35.9 1356790 37.9 1356790 36.5 1356790 38.2^1356790 38.7 1356790 38.7 1356790 38.5 1356790 37.9 1356790 37.8^1356790 40.2^1356790 40.0 1356790  16720 18720 20586 21859 23054 24040 24833 25830 27133 28810 30000  11308.94 12660.84 13343.44 14251.03 15556.22 16570.57 17964.74 19683.13 21372.73 22342.12 22763.22  0.080 0.080 0.080 0.080 0.080 0.080 0.080 0.080 0.080 0.080 0.080  4.91 5.81 6.73 7.23 7.23 6.95 6.90 7.05 7.09 7.21 7.51  2.69 3.38 4.41 4.73 5.06 4.73 3.60 4.26 3.28 3.86 4.11  15 15 15 15 15 15 15 15 15 15 15  NB NB NB NB NB NB NB NB NB NB NB  1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990  695.4 696.4 696.6 703.2 707.9 709.9 710.4 712.3 714.3 717.8 722.4  16.4 16.4 16.5 16.4 16.3 16.5 17.1 17.1 17.2 17.1 17.2  88.41 89.37 91.22 89.14 85.79 87.10 90.53 90.59 89.99 89.00 96.51  11.0 11.5 14.1 14.8 14.8 15.1 14.3 13.1 12.0 12.5 12.1  26.8^72090 32.3^72090 34.8^72090 37.5^72090 31.8^72090 29.3^72090 30.0^72090 34.6^72090 34.1^72090 35.3^72090 36.9^72090  16353 18081 19690 20878 22128 23494 24026 25856 26997 27517 28000  7217.43 8548.25 9369.8C 10655.57 11830.77 12687.70 14187.78 15274.46 16470.67 17617.72 18403.9 1  0.284 0.284 0.284 0.284 0.284 0.284 0.284 0.284 0.284 0.284 0.284  6.18 7.39 8.02 8.57 8.74 9.42 8.04 8.37 7.88 8.17 8.61  3.21 4.82 5.77 6.22 6.29 6.54 3.60 4.26 3.28 3.86 4.76  NS NS NS NS NS NS  NS NS  1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990  845.1 847.4 849.5 857.0 864.4 871.0 873.2 878.0 881.9 888.3 895.1  32.4 32.7 32.9 32.9 33.1 33.4 33.9 34.2 34.2 34.5 34.9  115.07 119.89 123.32 105.88 99.58 100.24 103.41 103.97 103.86 106.62 116.69  9.7 10.1 13.1 13.2 13.0 13.6 13.1 12.3 10.2 10.3 10.5  27.0^52840 29.5^52840 30.4^52840 29.5^52840 28.0^52840 28.3^52840 27.8^52840 30.9^52840 30.5^52840 30.8^52840 30.4^52840  16069 17829 19437 20747 22170 22039 23206 23981 24688 26413 27100  7445.27 8667.69 9961.15 11235.71 12379.69 13701.49 14923.27 15897.49 17084.70 18089.61 19011.28  0.486 0.486 0.486 0.486 0.486 0.486 0.486 0.486 0.486 0.486 0.486  6.62 7.40 7.86 9.29 9.58 9.70 8.79 9.28 8.81 9.37 10.11  PE PE PE PE PE PE PE PE PE PE PE  1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990  122.8 122.5 122.4 123.7 125.1 126.0 126.6 127.3 128.5 129.9 130.7  0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0  92.39 93.51 97.04 98.42 101.78 100.87 96.47 88.66 100.71 104.26 107.74  10.6 11.2 12.9 12.2 12.8 13.3 13.4 13.2 13.0 14.1 14.9  30.0^5660 30.1^5660 26.4^5660 31.2^5660 22.9^5660 21.6^5660 23.0^5660 29.0^5660 29.5^5660 31.3^5660 33.7^5660  13003 14074 14964 16226 16766 16966 16975 17195 18855 19375 20000  6889.25 8236.73 8586.60 9417.95 10367.71 10476.19 11832.54 12482.33 13859.92 14603.54 15233.36  0.919 0.919 0.919 0.919 0.919 0.919 0.919 0.919 0.919 0.919 0.919  NF NF NF NF NF NF NF NF NF NF NI'  1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990  565.6 567.7 566.2 571.4 572.4 571.5 568.3 568.1 568.8 571.1 572.7  27.1 27.1 27.6 27.7 28.0 28.1 28.5 28.6 28.5 28.6 28.6  71.96 79.74 80.13 78.19 77.98 74.60 76.99 77.37 75.74 82.30 82.93  13.2 13.8 17.3 19.2 20.6 20.9 18.7 16.8 15.2 15.4 17.1  43.7^371690 49.2^371690 49.4^371690 51.9^371690 47.5^371690 40.3^371690 45.6^371690 48.8^371690 51.8^371690 52.0^371690 55.1^371690  15645 17631 19707 20605 20308 20125 20723 21336 22835 24252 25200  7240.10 8178.62 8935.01 9599.23 10389.59 11142.61 11933.84 12997.71 13990.86 14841.53 15343.11  0.032 0.032 0.032 0.032 0.032 0.032 0.032 0.032 0.032 0.032 0.032  NS  NS NS  - 51 -  13.0 13.0 13.0 13.0 13.0  11.76 (3' 15.86 16.00 16.28 15.68 16.00  28.54 28.37 28.22 28.42 29.39 27.78 27.84 27.92 27.69 26.60 26.83  4 4 4 4 4 4 4 4 4 4 4  12.0 13.0 14.0 14.0 15.0 15.0 15.0 15.0 16.0 16.0 16.0  11.51 11.02 10.11 9.87 9.84 9.82 10.39 10.36 10.85 11.12 11.30  1.74 1.70 1.67 1.70 1.73 1.67 1.76 1.74 1.77 1.83 1.86  2 2 2 2 2 2 2 2 2  3.21 4.82 5.77 6.22 6.29 6.54 3.60 4.26 3.28 3.86 4.76  3 3 3 3 3 3 3 3 3 3 3  13.0 13.0 15.0 15.0 15.0 15.0 15.0 15.0 15.0 15.0 16.0  10.94 10.81 9.57 9.30 9.10 9.13 9.10 9.43 9.75 9.73 9.70  2.11 2.09 2.02 2.05 2.12 1.99 1.97 2.02 2.04 2.04 2.07  2 2 2 2 2 2 2  6.63 9.43 11.35 11.66 12.30 12.89 8.33 7.44 7.25 7.51 8.69  3.21 4.82 5.77 6.22 6.29 6.54 3.60 4.26 3.28 3.86 4.76  1 1 1 1 1 1 1 1 1 1 1  10.0 10.0 10.0 10.0 10.0 10.0 10.0 15.0 15.0 15.0 15.0  5.62 5.74 5.47 5.57 5.29 5.84 5.79 5.88 6.06 5.94 5.90  0.16 0.16 0.17 0.18 0.18 0.19 0.19 0.19 0.19 0.19 0.19  6 6 6 6 6 6 6 6 6 6 6  6.81 7.99 9.60 10.19 10.55 13.10 12.20 11.26 11.94 11.13 11.02  3.21 4.82 5.77 6.22 6.29 6.54 3.60 4.26 3.28 3.86 4.76  7 7 7 7 7 7 7 7 7 7 7  15.0 15.0 16.0 16.0 16.0 16.0 16.0 16.0 16.0 16.5 17.0  8.54 8.75 8.10 7.47 7.30 7.36 7.82 8.35 8.30 7.83 8.00  0.95 0.98 0.99 0.95 0.96 0.92 0.96 1.00 0.99  2 2 2  13.0 13.0 8.0 5.5  Z.  .  17.67 17.31 16.09 15.63  0.95.  0.99  3 3 3 3 3 3 3 3 3 3 3  2 2  2 2  2 2  2  2 2 2 2 2 2 2  APPENDIX 3 NOTES TO JAPANESE MANUFACTURING GREENFIELDS AND PROVINCIAL DATA  1. If the year of establishment was 1991 then for our purposes we classified this establishment as being in 1990, because data on provincial variables was incomplete for 1991. 2. If the year of establishment was actually the year of operation, then to approximate the year of establishment we subtracted one year. 3. Highway miles were assumed to remain constant over the eleven year period. Provincial figures for 1988 were used. 4. Crude oil prices were not available from Statistics Canada for New Brunswick and were incomplete for Quebec. However, given the national energy policy and data that was available these missing figures were estimated. The Alberta amount plus $10.00 was substituted for Quebec and New Brunswick was assumed to face the same oil prices as Quebec.  APPENDIX 4 JAPANESE MANUFACTURING GREENFIELDS PROVINCIAL INDUSTRIAL LEVEL FOR 1987 WAGES, FUEL, and REVENUE (In Thousands) SIC  Firm #  Prov  Estabs  2512  1  1 2 3 4 5 6  45 336 66 185 338 10  3391  2  1 2 3 4 5 6  0 4 0 16 5 0  1 2 3 4 5 6  1 1 0 16 12 0  1811  3  Wages  Fuel  Revenue  86281 1158091 58145 202220 365629 15995  12998 137610 8953 31180 53416 2303  357328 5356553 270520 867672 1794828 66928  45225 7308  4347 986  345477 38924  163922  33971  1004544  86281 1158091 58145 202220 365629 15995  12998 137610 8953 31180 53416 2303  357328 5356553 270520 867672 1794828 66928  191871  14738  1006041  2512  4  1 2 3 4 5 6  45 336 66 185 338 10  3255  5  1 2 3 4 5 6  0 1 1 33 10 0  1 2 3 4 5 6  14 9 8 99 80 4  3712  267 22  12342 3462  27802 549  2840 12  123278 955  1 2 3 4 5 6  45 336 66 185 338 10  86281 1158091 58145 202220 365629 15995  12998 137610 8953 31180 53416 2303  357328 5356553 270520 867672 1794828 66928  2599  2512  6  7  - 53 -  1131  8  1 2 3 4 5 6  7 8 3 15 4 3  24954 45047  2680 3486  132941 219923  170444 194485 11664  14605 14208 1199  1180198 679285 61786  2512  9  1 2 3 4 5 6  45 336 66 185 338 10  86281 1158091 58145 202220 365629 15995  12998 137610 8953 31180 53416 2303  357328 5356553 270520 867672 1794828 66928  2711  10  1 2 3 4 5 6  2 16 6 5 8 1  322950 125641 125540 94275  149825 68704 63380 44371  2629804 892708 944065 716661  1 2 3 4 5 6  2 16 6 5 8 1  322950 125641 125540 94275  149825 68704 63380 44371  2629804 892708 944065 716661  2711  11  3199  12  1 2 3 4 5 6  26 62 7 450 191 7  7903 30489 3300 526267 207077  314 1057 111 25434 8114  31486 112847 9812 2681006 771907  3099  13  1 2 3 4 5 6  39 43 2 249 103 10  13182 12532  837 832  55700 45750  205147 52915 1254  19611 3613 133  902069 217599 6365  1 2 3 4 5 6  0 4 3 9 21 0  262219  171838  1719818  348143 727865  169559 401159  1808450 3679968  1 2 3 4 5 6  2 6 0 75 10 0  244844 21581  19772 2040  1119061 118368  2712  3256  14  15  3259  3231  3062  1011  3251  3256  16  17  18  19  20  21  1 2 3 4 5 6  3 12 4 137 23 1  1 2 3 4 5 6  0 3 0 15 5 1  1 2 3 4 5 6  4 14 0 427 86 0  1 2 3 4 5 6  64 45 6 198 138 27  1 2 3 4 5 6  0 4 1 37 5 0  1 2 3 4 5 6  2 6 0 75 10 0  3341  22  1 2 3 4 5 6  0 1 0 16 8 0  2912  23  1 2 3 4 5 6  3 4 0 12 5 2  644112 20550  37826 925  3300384 90632  1849987  89923  36834766  1284 2368  44 99  2422 6907  271055 27339  9612 982  679773 62324  119395 74661  9747 5597  2340450 625625  336150 183860 34637  28638 25182 3314  3237259 2198977 477239  594076 544  51832 53  3407927 2142  244844 21581  19772 2040  1119061 118368  3361  3361  3257  3259  3231  3257  3562  3399  24  25  26  27  28  29  30  31  1 2 3 4 5 6  8 17 0 86 29 1  1 2 3 4 5 6  8 17 0 86 29 1  1 2 3 4 5 6  3 0 0 10 4 2  1 2 3 4 5 6  3 12 4 137 23 1  1 2 3 4 5 6  0 3 0 15 5 1  1 2 3 4 5 6  3 0 0 10 4 2  1 2 3 4 5 6  13 17 2 65 41 4  1 2 3 4 5 6  2 3 0 43 21 1  7716 33102  77 339  18670 79239  238360 60226  6017 1370  1380276 261013  7716 33102  77 339  18670 79239  238360 60226  6017 1370  1380276 261013  499  17  1866  195106 167  3818 14  873087 858  644112 20550  37826 925  3300384 90632  1849987  89923  36834766  499  17  1866  195106 167  3818 14  873087 858  4372 7451  325 316  24050 27461  116417 28854  12157 1835  464558 150520  83529 37673  9900 6340  366685 178511  1621  32  1 2 3 4 5 6  8 9 3 36 15 3  3341  33  1 2 3 4 5 6  0 1 0 16 8 0  3011  34  1 2 3 4 5 6  4 3 2 24 6 2  1 2 3 4 5 6  10 5 5 45 14 3  1 2 3 4 5 6  3 12 4 137 23 1  351  3259  35  36  3192  37  1 2 3 4 5 6  117 68 7 222 88 13  3259  38  1 2 3 4 5 6  3 12 4 137 23 1  1 2 3 4 5 6  1 5 0 23 6 0  3254  39  7149 7794  1063 616  51757 39474  44759 23691  6299 2564  366865 312982  6615  230  21898  111223 27561  3336 1029  464761 158376  64697 21030  20306 7955  289807 69494  644112 20550  37826 925  3300384 90632  127396 57044 7331 384532 73083 5407  6455 2224 316 14476 3165 301  539744 247175 36916 2115972 363639 28464  644112 20550  37826 925  3300384 90632  1265  139  5164  4637  825  28465  3081  40  1 2 3 4 5 6  113 180 22 652 369 45  33221 61610 7846 212299 84173 12519  1491 2556 418 9074 4073 588  87487 171660 22358 556718 241621 45003  3731  41  1 2 3 4 5 6  9 7 0 52 26 0  33247 6225  11876 685  524571 71206  154734 52056  44597 18250  1723356 642122  1 2 3 4 5 6  11 4 1 22 5 0  2185  46  8212  26161  1062  165016  1 2 3 4 5 6  0 3 0 15 5 1  1849987  89923  36834766  1 2 3 4 5 6  20 24 1 178 50 6  11069 7894  1527 865  43682 19699  172338 22309 2190  24917 2964 287  705846 69130 5997  1 2 3 4 5 6  2 3 0 43 21 1  83529 37673  9900 6340  366685 178511  1 2 3 4 5 6  163 177 20 820 345 22  1134868 345967 11326  49033 14774 701  5697737 1383431 57585  1 2 3 4 5 6  51 69 8 276 139 19  26408  516  94609  459214  15372  1985639  3092  3231  3041  3399  319  391  42  43  44  45  46  47  3257  3259  3281  1021  2961  3011  337  3194  48  49  50  51  52  53  54  55  1 2 3 4 5 6  3 0 0 10 4 2  1 2 3 4 5 6  3 12 4 137 23 1  1 2 3 4 5 6  3 79 18 101 65 0  1 2 3 4 5 6  1 48 73 19 40 1  1 2 3 4 5 6  4 9 0 32 23 1  1 2 3 4 5 6  4 3 2 24 6 2  1 2 3 4 5 6  17 22 2 154 71 4  1 2 3 4 5 6  3 10 4 63 33 1  499  17  1866  195106 167  3818 14  873087 858  644112 20550  37826 925  3300384 90632  717 20594  51 841  2079 90377  31106 26920  1823 1122  149543 141114  100725 54517  6435 5384  1067023 452911  38372  4005  207321  127082 71529  19981 11928  878524 707052  6615  230  21898  111223 27561  3336 1029  464761 158376  14877  481  56592  479350 88879 5267  17245 3601 934  1752783 399546 23034  8303 1175 139225 36346  166 84 5218 2311  29236 3698 493724 167758  PROVINCE CODES ^ ^ 3 = New Brunswick 2 = British Columbia 1 = Alberta ^ ^ 5 = Quebec 6 = Saskatchewan 4 = Ontario  - 59 -  APPENDIX 5 JAPANESE MANUFACTURING GREENFIELDS NATIONAL INDUSTRIAL LEVEL FOR 1987 WAGES, FUEL and SHIPMENTS (In Millions) SIC  Establishments  Wages  Fuel  Shipments  319  1601  1851  79  8638  337  281  624  23  2390  351  90  107  33  413  391  603  591  18  2451  1011  526  851  82  9811  1021  414  530  43  4111  1131  48  502  42  2557  1621  77  89  11  801  1811  31  211  43  1199  2512  1093  1919  252  8862  2599  272  63  5  275  2711  39  757  355  5817  2712  43  1572  927  8537  2912  26  81  13  204  2961  71  215  33  1701  3011  43  148  4  652  3041  288  217  30  850  3062  539  303  10  753  3081  1464  436  19  1194  3092  44  71  2  274  3099  463  290  25  1257  3192  533  666  27  3387  3194  116  187  7  703  3199  772  802  36  3724  3231  27  2116  100  39093  3251  50  598  52  3430  3254  35  168  16  720  - 60 -  3255  47  209  16  1086  3256  96  270  22  1252  3257  20  196  3  877  3259  190  672  39  3429  3281  327  93  5  442  3341  25  83  1  822  3361  147  367  8  2006  3391  27  66  6  420  3399  72  122  16  552  3562  153  161  14  688  3731  94  246  75  2961  3931  209  144  8  788  ^  APPENDIX 6 CONDITIONAL LOGIT REGRESSION RESULTS VARIABLES & THEIR SOURCE  ^LW  =^log of manufacturing wages as listed in Appendix 2. LE =^log of weighted average energy prices as listed in Appendix 2. LE2 =^log of low energy prices as listed in Appendix 2. SHIPRANK =^as listed in Appendix 2. LMETGDP =^log of (metropolitan times GDP per Capita times population, from Appendix 2). LCAN =^log of the number of Provincial firms in like industry as listed in Appendix 4. LJPN =^log of the number of Japanese manufacturing greenfields in the province as per Table 1. QUEBEC =^A dummy variable to measure the effect of this large province with few Japanese greenfield investments. LABOR = Using data from Appendix 5 to estimate the labour. factor input coefficient for each specific industry for the Cobb Douglas Production Function. ENERGY =^Using data from Appendix 5 to estimate the energy factor input coefficient for each specific industry for the Cobb Douglas Production Function. UNION_P =^Unionization rate in the province as listed in Appendix 2. TAX =^Provincial Corporate Tax Rate as listed in Appendix 2. CRIME =^Provincial crime rate per 1000 divided by the provincial population. Both of these numbers are in Appendix 2.  55 Japanese Manufacturing Greenfields  CHOICE 1 (AB) 2 (BC) 3 (NB) 4 (ON) 5 (PQ) 6 (SK)  FREQUENCY 1 11 2 35 4 2  PERCENT 1.8182 20.0000 3.6364 63.6364 7.2727 3.6364  Without a Measure of Market Size (i.e. GDP) LOG OF LIKELIHOOD FUNCTION^: -58.3579 NUMBER OF CASES^: 55 NUMBER OF CHOICES^: 330 Standard Parameter^Estimate^Error^t-statistic LCAN^.489813^.28027^1.74765 LJPN^1.21409^.435358^2.78872 SHIPRANK^-.324323^.196032^-1.65444 LW^-5.31907^4.93843^-1.07708 LV^-4.91472^2.68017^-1.83373 lw = log(avgmfgy), lv = log(avgengy)  A better fit is achieved with GDP LOG OF LIKELIHOOD FUNCTION^: -55.1337 NUMBER OF CASES^: 55 NUMBER OF CHOICES^: 330 Standard Parameter^Estimate^Error^t-statistic LCAN^.677792^.285464^2.37435 LJPN^2.26426^.641879^3.52755 SHIPRANK^-.316939^.203115^-1.56039 LW^-8.23081^5.30678^-1.5510 LV^-6.42729^2.72486^-2.35876 LGDP^-1.29039^.501009^-2.57558  

Cite

Citation Scheme:

    

Usage Statistics

Country Views Downloads
United States 9 5
Japan 6 0
China 5 4
Germany 3 10
United Kingdom 1 0
City Views Downloads
Tokyo 6 0
Unknown 4 10
Beijing 4 2
Ashburn 3 0
Wilmington 2 0
Chicago 1 0
Dolton 1 0
Boardman 1 0
Redmond 1 2
Hangzhou 1 0

{[{ mDataHeader[type] }]} {[{ month[type] }]} {[{ tData[type] }]}
Download Stats

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0086082/manifest

Comment

Related Items