1 Deutsch, Karl W., et al., Political Community and the North Atlantic Area (Princeton: Princeton University Press, 1957).Google Scholar

2 Etzioni, Amitai, Political Unification (New York: Holt, Rinehart and Winston, 1965).Google Scholar

3 Holsti, K. J., “Change in the International System: Interdependence, Integration, and Fragmentation,” in Ole Holsti, Randolph Siverson and Alexander, George (eds.), Change in the International System (Boulder: Westview Press, 1980), 24.Google Scholar

4 Naomi, Black, “Absorptive Systems are Impossible: The Canadian-American Relationship as a Disparate Dyad,” in Andrew Axline, W., et al., Continental Community: Independence and Integration in North America (Toronto: McClelland and Stewart, 1974), 92–108.Google Scholar

5 Moran, Theodore H., “Multinational Corporations and Dependency: A Dialogue for Dependentistas and Non-dependentistas,” International Organization 32 (1978), 79–100.CrossRefGoogle Scholar

6 See von Riekhoff, Harald, “The Third Option in Canadian Foreign Policy,” in Tomlin, Brian W. (ed.), Canada's Foreign Policy: Analysis and Trends (Toronto: Methuen, 1978), 87–109;Google Scholar and Peyton, V.Lyon, and Tomlin, Brian W., Canada as an International Actor (Toronto: Macmillan, 1979), 122–39.Google Scholar

7 Tomlin, Brian W., Dolan, Michael B., Riekhoff, Harald von, and Molot, Maureen A., “Foreign Policies of Subordinate States in Asymmetrical Dyads,” The Jerusalem Journal of International Relations 5 (1981), 14–40Google Scholar. For a summary statement see Dolan, Michael B., Tomlin, Brian W., Molot, Maureen A., and Riekhoff, Harald von, “Foreign Policies of African States in Asymmetrical Dyads,” International Studies Quarterly 24 (1980), 415–49.CrossRefGoogle Scholar

8 Dolan, See Michael B., Tomlin, Brian W., Riekhoff, Harald von and Molot, Maureen A., “Foreign Policies of African States in Asymmetrical Dyads” and "Asymmetrical Dyads and Foreign Policy: An Empirical Analysis of Canada-United States Relations, 1963-1972,” Journal of Conflict Resolution, forthcoming;Google Scholar and Dolan, Michael B. and Tomlin, Brian W., “Foreign Policies of Subordinate States in Asymmetrical Dyads: Theoretical Reformulation and Empirical Analysis, Canada-U.S. Relations,” a paper presented at the National Science Foundation Conference on Dependency Reversal, New Mexico State University, 1981.Google Scholar

9 Haas, Ernst B. and Schmitter, Phillipe C., “Economics and Differential Patterns of Political Integration: Projections About Unity in Latin America,” in International Political Communities: An Anthology (New York: Doubleday Anchor, 1966), 259–300.Google Scholar

10 Lyon, and Tomlin, , Canada as an International Actor, 95–121.Google Scholar

11 The rules are presented in A Manual for the Identification and Abstraction of Canada-United States Events. Research Report No. 2 of the Asymmetrical Dyads and Foreign Policy Project, Carleton University, 1979, 27 pp. Extensive pre-testing of the Abstraction Manual produced successive refinements in the rules for identifying events. A separate test of inter-abstractor reliability was undertaken to ensure that events could be identified in the source within acceptable limits of reliability. Setting such limits is problematic, however, because little attention has been devoted to the problem of identifying events in original sources, as opposed to indexes and chronologies, and because the few treatments which have been given the issue indicate considerable difficulties in locating identical universes of events (see Sigler, John H., “Reliability Problems in the Measurement of International Events in the Elite Press,” in Sigler, John H., Field, John O., and Adelman, Murray L. [eds.], Applications of Events Data Analysis [Beverly Hills: Sage, 1972]).Google Scholar Using the familiar coefficient of reliability (2M/N₁ + N₂, where M is the number of events abstracted by both abstractors, N₁ is the total number of events abstracted by the first abstractor and N₂ is the total number of events abstracted by the second abstractor), CR > .5 was achieved for 91 per cent of the sample months selected for the test, and we deemed this level acceptable. The coefficient is comparable to a variance-explained term and may be interpreted as a squared correlation coefficient (see Rummel, Rudolf J.“Dimensions of Conflict Behavior Within Nations,” Journal of Conflict Resolution 10 [1966], 65–73)CrossRefGoogle Scholar. Our coefficients are higher than the CRs that have been reported previously for abstraction from an original source (see Rummel, "Dimensions of Conflict" and Sigler, "Reliability Problems," 16). (Inter-coder reliability, typically much higher, is reported in footnote 14, below.) The implications of having different samples, which would occur if abstractor reliability were low, have not been addressed systematically in event literature, and we intend to undertake such an analysis with our data at a later stage.

12 The Globe and Mail was chosen as the primary source after consultation with members of the international studies community in Canada. Although it is not nationally oriented, the Globe and Mail comes closest to being a “paper of record” forthe country. A sample of events has been drawn from each of three regional press sources as well, and a source comparison study is planned. The Monthly Report on Canadian External Relations (titled International Canada since 1970) is published by the Canadian Institute of International Affairs and provides a “summary of important developments in the field of Canadian external policy.” It was selected as the next most comprehensive source to supplement the daily coverage provided by the Globe and Mail. Both sources were used in the compilation of events forthe 1963-1972 period (events reported in both sources were abstracted only once for the collection). This period was chosen for our initial focus because it includes episodes of intensified relations following the Liberal party's return to power in 1963 and captures the intense expressions of Canadian nationalism during the middle and late 1960s. Furthermore, this period culminates in the articulation of the Third Option strategy. Thus, the shifts in foreign policy which our theory is intended to explain are well-represented in this 10-year period. We intend to update the collection of events to the present using International Canada as a primary source.

13 The variables, and the coding rules, are described in A Manual for the Coding of Canada-United States Events. Research Report No. 3 of the Asymmetrical Dyads and Foreign Policy Project, Carleton University, 1979, 93 pp.

14 The order of the seven categories was confirmed through a paired-comparison test in which 69.2 percent of 504 ranks were correctly placed. The modal choice was correct for 24 of 28 ranks, or 85.7 per cent. The test was conducted using 18 senior undergraduate and graduate students in international politics. Inter-coder reliability, using the same coefficient as was previously employed for abstraction (note 11), is as follows for the foreign policy dimensions: CR = .722, over all dimensions; .675, transactions; .791, policy coordination; .675, alignment; .866, distribution of benefits; and .603, affect. The coefficients include decisions concerning both the placement of the action on the dimension and the selection of the appropriate category within the dimension. The CR levels are directly related to the proportion of “not applicable” values on the dimensions; distribution of benefits, the dimension with the smallest number of real values, has the highest CR, and affect with the greatest number has the lowest CR. Apparently it is easier for the coders to agree on the nonapplicability of an event than it is to agree on the specific value to be given to an event judged to be applicable to the given dimension. Comparisons with other event projects are difficult because there are few event projects using ordinal scales that report CR levels (some report correlation coefficients, but they are not comparable because they measure whether coders found a similar number of references to a particular category or value, not whether specific events were coded identically). One major event project, CREON, does include a five-point scale of affective behaviour on which all events were deemed applicable. Although, ceterisparibus, the reliability of a five-point scale should be greater than our seven-point scale, CREON's CR is .57; see Hermann, Charles F., et al., CREON: A Foreign Events Data Set (Beverly Hills: Sage, 1973), 88.Google Scholar

15 Typically, each abstracted event contains information concerning the actor, the target and the action, as well as additional information on the setting, the issue, and the like. To give a simple illustration, an event may have the following format: The Canadian prime minister expresses unreserved support for the policy of the United States on a particular issue. Here Canada is the actor, the United States the target, and the expression of support for American policy constitutes the action. Unequivocal support for American policy, as cited here, would be assigned a +3 value on the alignment dimension; qualified support, with reservations attached, would be assigned a +2 value; a zero value would be assigned to a statement which is expressly noncommittal; and complete disagreement or lack of support would be assigned a −3 value. The action in any particular event may be relevant to more than one dimension.

16 The quarter was selected as the initial unit of aggregation for the examination of the data. In future analyses, more directly related to a test of the theory, we intend to explore the effects of using various periods of aggregation.

17 The weights employed for intensity values vary from +3 to -3, with a mid-point of zero. The index generated with these weights correlates r = .99 with an interval scale constructed by expert judges; see Dolan and Tomlin, “Foreign Policies of Subordinate States in Asymmetrical Dyads: Theoretical Reformulation and Empirical Analysis, Canada-United States Relations,” 19.

18 Disparate issues and actors within each government are aggregated in this measure. Individual agents and specific issues are identified in our more comprehensive data set and these will be analysed in later stages of the research project.

19 This is approximately the period for which our theory will be tested eventually. The extended period figures are presented simply in the interest of providing the information to the reader.

20 Azar, Edward E. and Sloan, Thomas J., Dimensions of Interaction (Pittsburgh: International Studies Association, 1975). The Canada-United States events for the COPDAB collection are drawn primarily from the New York Times.Google Scholar

21 The large differences between the number of Canadian and American events will not distort the results because the total number of events, respectively, in each quarter is placed in the denominator of the measures of the dimensions.

22 Useful sources on polynomial functions and on their limitations in forecasting are Gerard Tintner, Econometrics (New York: John Wiley & Sons, 1965);Google ScholarNelson, Charles R., Applied Time Series Analysis (San Francisco: Holden-Day, 1973)Google Scholar: and Lowden, D. J. “The Perils of Polynomials,” Management Science 9 (1963), 546-50Google Scholar. Polynomial functions and moving averages have become passe in the field of forecasting as more accurate techniques based on stochastic processes, such as Box-Jenkins modelling and spectral analysis, have been developed. Because our concern is not with forecasting, difficulties with polynomials in that pursuit are irrelevant. We rejected the use of Box-Jenkins models for this article because the Box-Jenkins models we estimated were in effect too close to mirroring the actual data and therefore were not a useful means of highlighting the trends in the data.

23 Appendix A illustrates the polynomial function by including both the actual and predicted values for the coordination dimension.

24 Gilpin, Robert, “Integration and Disintegration on the North American Continent,” International Organization 28 (1974), 862.CrossRefGoogle Scholar

25 Citing relevant historical cases, Peyton Lyon questions the validity of the assumption that the establishment of an economic free trade area would make political integration inevitable. See Canada-United States Free Trade and Canadian Independence (Ottawa: Economic Council of Canada, 1975).

26 See for example Kari Levitt, Silent Surrender: The American Economic Enterprise in Canada (Toronto: Macmillan, 1970), and Wallace Clement, Continental Corporate Power: Economic Linkages between Canada and the United States (Toronto: McClelland and Stewart, 1977). Clement does note the existence of an indigenous Canadian economic elite located primarily in the small-scale manufacturing sector. This particular elite group supports Canadian economic nationalist policies in opposition to continental integration. But, overall, this elite is regarded as too weak to prevail against the powerful pro-integrationist forces composed of Canada's dominant, indigenous elite, which is concentrated heavily in finance and transportation, and American-controlled multinationals operating in the manufacturing and resource sectors.

27 Deutsch, Karl W., Nationalism and Social Communication (Cambridge, Mass.: MIT Press, 1953) and Deutsch, et al., Political Community and the North Atlantic Area.Google Scholar

28 Keohane, Robert O. and JrNye, Joseph S.., “Introduction: The Complex Politics of Canadian-American Interdependence,” International Organization1 28 (1974), 595–607;CrossRefGoogle Scholar and Power and Interdependence: World Politics in Transition (Boston: Little, Brown, 1977).Google Scholar

29 Pentland, Charles“Political Integration: A Multidimensional Perspective,” in Axline, et al., Continental Community, 60.Google Scholar

30 Riekhoff, Harald vonSigler, John H., and Tomlin, Brian W.Canadian-United States Relations: Policy Environments, Issues, and Prospects (Montreal: C. D. Howe Research Institute, 1979), 1–57.Google Scholar

31 Holsti, K. J. “Canada and the United States,” in Spiegel, Steven L. and Waltz, Kenneth N. (eds.), Conflict in World Politics (Cambridge, Mass.: Winthrop Press, 1971), 375–96.Google Scholar

32 Fora discussion of the role of crisis in Canada-United States relations, see Dolan, Michael B.Tomlin, Brian W. and Riekhoff, Harald von “Instability and Sudden Change in Canada-United States Relations,” a paper presented at the annual meeting of the Canadian Political Science Association, Halifax, 1981.Google Scholar

33 See George Grant, Lament for a Nation (Toronto: McClelland and Stewart, 1965):Google ScholarLaxer, James (ed.), Canada Ltd. (Toronto: McClelland and Stewart. 1973);Google ScholarLevitt, Silent Surrender; and Ian Lumsden (ed.). Close the 49th Parallel (Toronto: University of Toronto Press, 1970).Google Scholar

34 Gilpin, “Integration and Disintegration on the North American Continent,” 857.

35 Keohane and Nye, “The Complex Pattern of Canadian-American Interdependence,” 595-602, and Power and Interdependence, 3-57.

36 Wright, Gerald and Molot, Maureen A., “Capital Movements and Government Control,” International jOrganization 28 (1974), 671-88.CrossRefGoogle Scholar

37 Keohane and Nye (Power and Interdependence, 165-93) attribute part of Canada's success in bargaining with the United States to the undivided attention accorded to the latter in Canadian policy considerations.

38 An alternative explanation is statistical. The erratic quarterly fluctuations across four dimensions exhibited in the United States set may be a function of the differential number of events in the two data sets, with the Canadian set containing more than twice as many events. The adequacy of this explanation will be investigated by drawing a random sample from the Canada set, and then comparing the quarterly fluctuations of the reducedCanadian set with the American foreign policy set.