The Comparative Effects of Power, Economic Development, and Political System on International Conflict and Cooperation Dr. Mary Lou Moore and Dr. Jack E. Vincent abstract

National attributes, such as those relating to power, economic development and political system may be important to understanding both the level of, and changes in, international conflict and cooperation. In this study we will first examine the most significant national attribute predictors (measured both within time and across time) of EG (economic development growth), PG (power growth), AUTH (Authoritarian growth), COOP (cooperation growth) and CON (conflict growth). We believe that an examination of such possible linkages is quite testable and useful for testing hypotheses as to the possible effects of such changes upon the international system. For example, power growth is predicted, in this study, to be linked to growth in national conflict behavior. If this is shown to be the case, then, with the methods employed, “critical attribute variables” can be identified in this regard and the policy implications of such linkages evaluated.

To achieve these objectives, 144 nations were scored on the attribute variables found in the University of Idaho Martin Institute Archives (1970-l989). They were also scored on 22 “conflict” and “cooperation” variables derived from the same source. Several factor analyses were performed in this regard. The resulting factor scores were then correlated (using Spearman's Rho) with the “growth variables,” of EG, PG, AUTH, COOP and CON described above. This research tests hypotheses both from existing literature and from our own assumptions. For example, our analysis indicates that total population appears to be a critical variable related to EG (economic growth) defined in terms of changes in GNP/per capita. Those nations that started with higher populations at the beginning of the period under consideration (1970-1989), tended to exhibit more EG than nations that started the period with smaller populations. The central purposes of this project, then, are to identify “critical attribute variables” linked to the growth indicators as well as evaluate the policy implications of each correlational linkage.

Introduction

National economic development appears to be a primary concern of national leaders, businessmen, investors, academics and the general public. Economic development theory has two distinct components: pure economic theory (factors relating to supply and demand, etc.) and infrastructure considerations (factors such as population size, educational attainment, etc.) that likely impinge upon and affect economic growth. We focus, in this study, on the infrastructure factors that may account for why nations differ so radically in respect to their development. The identification of such “critical variables” may assist states in developing effective policies to foster their economic development. Much of the infrastructure literature is rich in “theory” but weak in “testing” and “rank ordering” (in terms of importance) of the critical variables that may prove useful for national policy decision-making.

To summarize: it is the purpose of this research project to develop a methodology to empirically test infrastructure development propositions relating to power, economic development, cooperation and conflict growth as well as political system changes (toward or away from authoritarianism). This is done for the purpose of identifying and ranking potentially important critical variables in each case. The final step will be to suggest possible policy implications from such “critical” linkages.

Development and its Consequences

One important research question has concerned the relative importance that economic development, power, and political system have in respect to international behavior. Vincent (1996) undertook one such study. He has argued:

Scholars do not agree on what “drives” the international relations system, in the sense of weighting the relative importance of several potentially important factors, such as power, economic development and political system. This was particularly true during the cold war where a set of power configurations and relations occurred which may not be repeated for the foreseeable future. We now have an opportunity to evaluate whether the importance of such factors as power, economic development and political system, were unique to the cold war era or may tend to prevail in a similar way in new systemic arrangements, as in the present pattern, or in still other patterns likely to develop in the future. The purpose…is to set the stage for such an evaluation by focusing on the “middle period” of the cold war (the years from l966 to l978) where data is rich enough to attempt such an evaluation (p.2)….. it may be concluded that power is more important than economic development and economic development is more important than political system across all three-time periods (p.5).
Some projects have evaluated of the linkage of “quality of life” (defined primarily in terms of economic development) to “political liberties.” For example, Vincent and McCluskie, (1977: 77) found:

… a positive association in the 1970s between shifts in quality of life indicators and shifts in democracy (viewed as a dependent variable). However, examination of the1980s, in this study, did not indicate the same strong relationships. Nevertheless, at the end of the decade, the most democratic states stood far above the most undemocratic states on most quality of life indicators. When positive quality of life shifting occurs, it will likely move the world system in a democratic direction. This possible relationship supports foreign policy decision-makers concerned about increasing democracy and peace on a worldwide scale.

As noted above, both the “causes” and the “effects” of “rising economic development” are primary research focus in this project. Thus, infrastructure considerations, such as fertility, could be very important to a nation's economic development. For example, high fertility may be associated with less economic growth. It is also possible that rising economic development may also have a positive association with both cooperation and conflict. In addition, linkages between increases in power indicators and increases in democracy might also be revealed.

Since the beginning of the industrial revolution, it might be noted that “nation states” have been significantly transformed, in part, from the advances in technology, communication, transportation, etc. In this connection, Rostow suggests that nations develop in stages: that is, “traditional society, preconditions for ‘take off’ appear, then take off occurs, a drive toward maturity, and the age of high mass consumption takes place” (in Chilcote, 2000: 12). Earlier stages of development may only afford survival whereas later stages of development may offer true progress in “quality of life” issues.

Organski has focused on the role of government in economic development theory and finds four stages: “primitive national unification, industrialization, national welfare, and abundance” (Chilcote, 2000: 12). The importance of increasing government efficiency to mobilize human and material resources toward national ends is also one of his basic themes. In line with this, Johnson (1982: 17) argues that some countries, such as Japan, depend upon governmental intervention to become industrially strong. It is possible that government type, democratic or autocratic, may be important in this regard.

Russett and Starr (1995: 587) argue that states can be classified as not free, partially free, or free (classified as authoritarian to democratic) are combined with economic practices from free market to socialistic. The causal relationship of government type to economic development, however, remains an important, but seldom empirically based, research issue.

To approach the research objectives defined above, several factor analyses were performed on the attribute and behavior data (consisting of twenty- three attributes scores, nine “cooperation” scores and the thirteen “conflict” scores) measured both within time (1989) and across time (1989-1970). The resulting factor scores were correlated (using Spearman’s Rho) with the growth indicators, that is, with the 1989-1970 (difference scores) for the growth variables discussed above. These correlations were then used to test hypotheses derived from the literature and our own assumptions about the dynamics of economic development, power increases, and shifts toward democracy as well as the possible implications for conflict and cooperation.

It should be noted that the variables defining growth are dropped (to avoid problems of auto-correlation) in each of the factor analyses of the remaining variables. For example, in the analysis of EG (economic growth), V1 (measuring GNP per capita growth) becomes the variable focused upon (it is isolated as the dependant variable) and all other variables are entered into the factor analysis. The same logic, of course, applies in the case of the other growth indicators.

In this connection, our chosen factor analysis procedure creates dimensions that “optimally” fit the multidimensional space of the correlation matrix, between the variables analyzed, under the condition that each factor is orthogonal (or uncorrelated) with every other factor in the analysis. Our Kaiser rotation procedure rotates the factor dimensions until an “efficient” axis set is found where the factor loadings (which are Pearson r correlations of the variables with the factor scores) tend to be either large or small in order to facilitate interpretation. The square of each variable loading on a factor represents the proportion of variance “accounted for” by that factor.

All analyses were carried out on our data by using the SPSS package (versions 9.0 and 10.0). The factor method involves the solution of a set of linear equations; these are called “normal” equations, and serve to minimize the least-squares error terms. The parameters for any particular set of data are found by matrix inversion of either a covariance matrix or a correlation matrix of the variables. In this procedure the variables are assumed to be distributed normally, and have (essentially) equal variance. However, this is not likely to be the case with large and diverse data sets like ours. On the other hand, many studies have shown the procedure is remarkably robust in spite of deviations from such assumptions (Green, 1978).

Thus, even if the normality and equality of variance assumptions are not strictly maintained, our linear least-squares approach, as employed in factor analysis, can still be used to describe the most critical trends in the data. In this connection, it should be noted that Spearman's Rho is employed in all of our significance tests. Thus, assumptions of normality and equality, relating to certain correlation tests, do not apply to the ranked factor scores, computed from our factor analyses. This is because since Spearman's Rho is a distribution free, nonparametric method.

Throughout this research we present examples of contrasting types, for example a state standing high on a factor may be classified as a economic “winner” while a state measuring low on this same factor may be classified as a “nonwinner” and so forth, through out, for all other factors. The nations selected for the contrasts, as winners or nonwinners, of course must have appropriate scores on the heavy loading variables, generated from the factor analysis, that are consistent with predictions. If they do locate as predicted, then the research project has turned out well, in the sense that conclusions can then be postulated concerning possible causal linkages between the factors and the growth variables. In this connection, since the factor scores are orthogonal, any variance explained by one factor (the factor scores of that factor) cannot be explained by any another factor in the factor set. That is, the Pearson r squares of each factor (in respect to a dependant variable) may be added to generate the Multiple R square for the entire set. This being the case, we therefore assume that the added Rho squares closely approximate or at least estimate Multiple R squared. In fact, the empirical difference (between Pearson’s and Spearman’s Rho) tends to be quite small (about .02 on the average with the dependant variable) when the factor scores approximate a normal distribution if a large number of variables load on a dimension. Thus, we can only “approximate” R squared because of the rank transformations employed in applying Spearman's Rho. In general, however, this only modestly disturbs the perfect orthogonality of the raw factor scores, generated by the factor analysis, prior to the use of Rho. In this connection, we felt it was necessary to proceed in this manner to avoid the argument that is likely to develop that deviant cases may be driving our correlation results.

Rules of Interpretation

We provide the following interpretive rules to our factor analyses:

A. Nations that score high on factors are predicted to have high scores on positively

loading variables and low scores on the negative loading variables.

B. Nations that score low on factors are predicted to have low scores on

positive loading variables and high scores on the negative loading variables.

3. 
   Nations that score high on a dependent index (such as Economic Growth) indicate high growth.
   

4. 
   Nations that score low on dependent index (such as Economic Growth) indicate low growth.
   
5. 
   +Rho indicates high scores on variable “A” (referring to type A above) are linked to high scores on variable “C” (referring to type C above) relating to economic
   

F. -Rho indicates high scores on “A” are linked to low scores on “C” relating to economic growth and so forth.

As noted, the data used in this research has been obtained from the Martin Peace Institute Archives at the University of Idaho,¹ which includes the Vincent Scale of Cooperation, ² and the Vincent Scale of Conflict.³ (See the Appendix, Part Eight.)

Sullivan indicates that, “the presence of a statistical correlation between two variables does not necessarily imply a causal relationship between them”(l976: 5). This being the case, can a “causal link” between economic development cooperation be found by applying our Rho correlation technique? Clearly the case for cause may be stronger when dynamic (change variables) variables are linked, although, even then, in the absence of manipulation, it is difficult to establish cause, since even the dynamic linkages may be due to “other” variables outside the study, which are not examined.

In this connection, Vincent has also argued: “There is an important issue concerning ‘cause’ in the social sciences. In the case of Einstein's theory, ‘curved space,’ created by the large mass of the sun, helps ‘explain’ Mercury's orbit. Or, depending upon one’s philosophical orientation might be viewed as a ‘cause of Mercury’s orbit. Cause, however, is very difficult to deal with in both the hard sciences and the social sciences. In the latter case, social scientists are seldom in a position to provide ‘definitive tests of presumed causal relationships” (Vincent, Model Building, 2002: 1).

For example, in this paper, we will assert that population size may be a “critical” variable regarding PG. We, of course, cannot actually test (by changing nation’s population sizes) to see if PG then increases. If we assume that the two variables, in fact, co-vary over time, we still cannot conclude that one variable “caused” the other to change. The reason why we cannot simply conclude that variable “a” caused variable “b” to change is the fact that an exogenous factor, variable “c” may have been the one “causing” the changes in both “a” and “b.” Our probabilistic theoretical model, then, may be viewed as a “plausible” (but not actually proven) explanation of what we are observing. The “policy recommendation equation” that comes later in the paper must be understood with these observations in mind. They are offered with the caveat that these policy recommendations can be viewed as ‘reasonable’ policy recommendations, since the observed correlations are, in fact, consistent with our probabilistic theoretical model. “Causal" relationships, then, are only “assumed,” so that possible policy recommendations can be spelled out. It should be noted that without such “causal assumptions,” the findings generated here could appear to be without social significance. That is, why bother to change anything if it is assumed that there are no causal connections between the variables linked in the research? Further, without making such causal assumptions, it then follows that a great deal of the probabilistic social science research, done to date, would wind up in the same dubious position of being viewed as disconnected to actual social policy for the same reasons. This is because, as noted above, most social scientists are not usually in a position to definitely “prove” possible posited causal connections related to their research, for the various reasons stated above.

Thus, if the results match the direction predicted in this study, the strongly associated dynamic variables may be prime candidates for “manipulation” based on the arguments presented above. For example, if a variable such as “newspaper circulation” is deemed critical to EG, then policies to promote that circulation in lesser-developed nations may assist in promoting EG. In the absence of actual manipulation, however, such correlational fits will only be “suggestive of,” not “proof of” the causal connection. The approach here is similar to the numerous government policies based on correlational linkages relating to seat belts, speed limits, intoxication levels, tobacco advertising, etc., where the exact “cause” can not be “proved” but the suggestive correlational linkages are strong enough to allow significant changes in public policy.

Attribute Theory and Economic Development

Attribute theory is one research approach directly related to this study. Vincent explains attribute theory (originally formulated by Rummel) this way:

Attribute theory works with two collections of variables, A-Space, referring to attribute variables, and B-Space, referring to behavior variables. It argues that A-space, expressed as factor scores, should account for B-space variables, also expressed as factor scores. Reducing the number of variables through factor analysis is integral to the theory since it is not possible to perform a multiple regression analysis on a variable set where v, the number of variables, exceeds n, the number of subjects. The term monad refers to the subject of study, i.e., individual nations. Thus when individual nations are scored on the factor dimensions of a factor analysis of attribute variables; the scores should account, in correlational terms, for the factor scores generated from an analysis of behavioral variables. To put it another way, if we know a nation’s attribute location in factor score terms, we can predict its behavior (also expressed in terms of factor scores) (1977: 3-4).

In this study, attribute theory suggests a possible foundation for statistical evaluation and is basically consistent with, but not perfectly so, the methodology explained above. That is, in this application, both attributes and behavior are factored together, with the exception of the variable under consideration, such as economic growth. However, reducing a large number of variables to a smaller orthogonal space is a primary concern here, as in attribute theory. The reduction to factors allows the evaluation of linkages and patterns that would not be possible with purely bivariate methods. Although our primary concern is with the shifts in economic development, power, and political system for international conflict and cooperation, the methods employed may also allow the discovery of linkages which previously may not have been discussed or even suspected.

In this connection, we can assume that virtually every state, and nearly every person, wants more wealth; and also, it would seem, nearly every person and/or state has ideas about how to achieve wealth. For example, Rummel argues “Freedom produces wealth and prosperity” (1996: 3). In this connection, authors, such as Rummel, appear to see democracy as a conflict-reducing factor and may assume, therefore, that it also promotes cooperation. As noted, Rummel also posits that economic development is a direct result of freedom: “Freedom is an economic engine of jobs, wages, increased earnings, new technology, and greater human choice” (1996: 3). In addition, a nation may have to develop economically to enjoy both political and civil rights (Vincent, 1987). For many states, democracy and economic development appear to go together. However, why is it that some nations who develop strong economies remain stable and peaceful while other nations divert their wealth toward power objectives and militaristic applications? In this connection, some scholars predict that increases in force capability (FC) will stimulate both conflict and cooperation. Cerven argues,

“…cooperation and conflict appear… highly linked and both are related to force

usage. That is, states that exhibit a high degree of cooperative behavior may also tend to engage in conflictual behavior and tend to use force. Conflict and cooperation appear to be part and parcel with one another. It is instructive to view these two seemingly disparate characteristics as possible part of a circle of behavior.” (2000: 7).

In this connection he also maintains, “states that frequently engage in the use of force may also frequently engage in negotiation…as the US and North Vietnam did during the war in Vietnam”(2000: 7).

Also, in this connection, Vincent argues “the expected effect of changes in power base should be roughly equal to the combined effect of economic development or political system changes on conflict behavior. It appears to have direct relevance to the current debate over the wisdom of altering political systems in a democratic direction to promote peace”(1996: 1).⁴

Regarding AUTH (our concept variable, where a high score indicates an Authoritarian system), Hewitt and Wilkenfeld have argued, “the presence or absence of democratic norms of conflict resolution will dictate whether or not such crises will be likely to escalate to violence.” They argue that there is a “dampening effect of democratic composition on the escalation of violence” (1996: 123). This also appears to support Rummel’s argument that “Democracies do not make war on each other, without exception” (1996: 2). It follows that (AUTHG, indicating authoritarianism growth) is expected to suppress cooperation and promote conflict.

At the same time, we also know that the static measures of conflict and cooperation, in fact, tend to be strongly correlated (Cerven, 2001). In the time frame we are using, the actual correlation is .82 (see the Appendix, Part Five). Thus, we assume that the empirical correlations may vary depending on whether the relationship examined is static (within time) or dynamic (across time). That is, it is possible that the dynamic correlations may support the predictions for democracy, in spite of the strong static positive co-variation between conflict and cooperation in both time frames examined.

Interpreting Growth

Our first research question deals with identifying the attributes and attribute differences that are linked to economic growth. This can be complicated, however, due to the measurement systems involved. For example, based on the literature and our own assumptions, we expect that previous high economic performance as measured by GNP/per capita growth by states will likely be associated with the highest levels of economic growth measured later in time. As noted, however, static predictions, such as ‘high’ or ‘low,’ (within time) can be distinguished from dynamic predictions (across time), such as a state moving ‘up’ or ‘down’ on a measurement scale. In contrast to nongrowth variables, growth variables can be all positive, all negative, or mixed. These outcomes, coupled with negative or positive Rhos and negative or positive factor loadings, can make interpretation very difficult. Our “Growth Interpretation Chart” (found in the Appendix) assists the reader in determining why the “findings language” takes the form it does, by interpreting of the possible mix of positive, negative, or mixed growth scores, positive or negative factor loadings and positive or negative Rhos). That is, every possible combination needs a separate rule to correctly interpret the results. By listing all the permutations for “GNP Per Capita and Population Growth,” and numbering each one as a different rule, we provide the reader with a very powerful, quick tool for the interpretation of any combination in the analyses that follow as well as aid in interpreting the “predictions” and “assumptions” portions of the research.

That is, it may also be the case that certain static indexes (such as cooperation) may be positively associated with another static measure, such as power, as predicted by Cerven, but at the same time, the growth indexes of the same variables can be negatively associated. For example, in our empirical analysis of power growth, we found such negatively associated linkages:

-.70 those with high PG had the “least policy support” growth (Rule 14),

-.95 those with high PG had the least “extend economic aid” growth (Rule 14),

-.86 those with high PG had the least “makes agreements” growth (Rule 14),

• 
  -.74 those with high PG had the least “total cooperation” growth (Rule 14).
  

The CON indicators are: reject, accuse, protest, deny, make demands, warn, threaten, hold demonstrations, reduce diplomacy, expel from country, seize property, use of force, and total conflict (Vincent scale).

Assumption 1, (relating to EDP): Economic growth will be linked to crude birth rate (low and/or down) crude death rate (low and/or down), infant mortality (low and/or down), fertility (low and/or down) and life expectancy (high and/or up) population per physician (low and/or down), population urban percent (high and/or up) population growth rate annual percent, (high and/or up), population growth rate urban annual percent, (high and/or up) because each indicator is assumed to be linked to economic growth. For example, cities can be viewed as the engine of economic growth, compared to rural areas.⁵ High infant mortality reduces the total possible work force and so forth.

Assumption 2, (relating to FC): GNP is assumed to force capability growth because the higher (high and/or up) the GNP, of a nation the more tools it possesses to further its force capability. The same also applies to the remaining variables listed below.⁶ These are population total (high and/or up), population urban total (high and/or up) and urban population percent of total (high and/or up) power (high and/or up) and passenger cars (high and/or up) arms imports in millions (high and/or up), armed forces in thousands (high and/or up), and military expenditures in millions (high and/or up).

Assumption 3, (relating to AUTH): Authoritarianism (high and/or up) will be linked to diminished economic growth because states that are repressive may have lower stability and economic yield.⁷ The variables that are used to measure authoritarianism are less civil rights (LCR) and less political rights (LPR).

(Type 1) Positive growth outcome, positive variable outcome; we recommend,

(Type 2) Negative growth outcome, positive variable outcome; we do not recommend,
(Type 3) Positive growth outcome, negative variable outcome; we do not recommend, and (Type 4) Negative growth outcome, negative variable outcome; we do not recommend.

The term “agree” indicates that the empirical results fell in line with the predictions made in the general equation.

The EDP variables empirically associated with economic growth are: low fertility, (.92)** agree, high life expectancy, (.91)** agree, low infant mortality, (.90)** agree, low crude birth rate, (.94)** agree, low population per physician, (.49) agree, low crude death rate, (.61) agree, high population urban percent, (.82)** agree, and high population growth rate annual percent, (.83) agree, high population growth rate urban annual percent (.88) agree. 

Rho was significant for the factor and the variable listed above loaded above .49 (many in the .80 range) on that factor. The percent of success for these critical variables is 100.0 percent.  The number of successful predictions is arrived at by dividing the number of agrees by the number of agrees plus the number of disagrees.  Since there were no disagrees, there is 100.0 percent agreement with the direction the theoretical model presented above for these variables.

Findings for Assumption 2

In this analysis of these findings, evaluating the relationship of Force Capability (FC) to Economic Growth, the interpretive rules that applied for Assumption 1 also apply here.

The Force Capability (FC) variables associated with economic development growth are: GNP, (.89)** agree, population total, (.88)** agree, high numbers of passenger cars, (.90)** agree, high population urban total, (.72) agree, high urban population percent of total, (.81)** disagree, (See Footnote 5), high power, (.84)** agree, high arms exports in millions, (.74) agree, high armed forces in thousands, (.67) agree, high military expenditures in millions, (.89)** agree. The percent of success for these variables is 88.8 percent.

Findings for Assumption 3

In this analysis of these findings, evaluating the relationship of Authoritarianism (AUTH) to Economic Growth, the interpretive rules that applied for Assumptions 1 and 2 also apply here. In this tabulation, AUTH variables associated with economic development are high civil rights, (.65) agree, and high political rights, (.61) agree. The percent of success for these critical variables is 100.0 percent.

The above list indicates the critical static indicators of economic growth. It is clear from this listing, that in the time period studied, the fastest economically growing states were those that were already developed. As mentioned above in Findings 1, we attempt to point out our recommendations, regarding certain variables, if certain conditions are met, in the conclusion sections. Here, for a number of variables, Type 1 condition is observed, as described above. A number of positive variables outcomes appear linked to economic growth. However, if this pattern holds over time, the consequence should be a greater and greater separation, of the rich nations from the poor nations in terms of wealth. That is, the rich nations have the very characteristics that will make them richer, compared to the poorer nations. This will very likely occur unless there is a willing massive transference of wealth from the rich nations to the poor nations, something that appears very unlikely given the current world political context. For example, Owen (2001:15) argues that world leaders are clearly unwilling or unable politically to implement the needed remedies that could bring economic equity between nations.

Antonakis (1997: 89) argues that military expansion neither hinders nor helps economic growth. This conclusion clearly does not appear to be supported by this data set. Many FC variables do appear linked to EG. However, in this connection, we recommend policies that might not make the world situation even worse off than it already is. For example, we do not think it is desirable for states to endlessly increase their populations. The negative results from such an expansion of population, with its resultant draining the environment of its natural resources make this a Type 3 solution. A type 3 condition is apparent for a number of these variables. That is, we argue that even if certain variables are linked to something, valued such as EG, we will not recommend them because of the likely negative effects in other areas; such as more and more passenger cars, higher and higher urban populations and more and more standing military forces. Further, even though FC appears strongly linked to economic growth, it also appears linked to force usage and conflict (Cerven, 2000: 7). This, as well, is the basis of our nonrecommendation for a number of the militarily related variables that appear to be associated with economic growth.

The above list indicates the critical indicators of economic growth and authoritarianism. Lower economic growth is associated with authoritarian types of political systems. Conversely states with higher civil and political rights have stronger economic growth. We recommend policies to promote higher civil and political rights in nations that lack them. Therefore, a Type 1 condition exists. This data supports Rummel’s (1996: 3) position that greater freedom brings greater prosperity.

Assumption 4, relating to EDP (Economic Development Promoters), will be linked to PG (power growth) because it is assumed that EDP provides the needed infrastructure base to develop additional power. ⁸

Assumption 5, relating to FC, (Force Capability) will be linked to PG (power growth) because it is assumed that force capability allows the protection of the state and its infrastructure opening the door to additional PG.⁹

Assumption 6, relating to LCR and LPR, (Authoritarianism) will be linked to lower PG (power growth) because states that are repressive may have lower levels of stability and lower productivity. Therefore, such states are assumed to be worse position to increase their force capability. ¹⁰

From the above assumptions, we can then predict the linkages of the variables listed for power growth. The predictive equation is PGmax = EDPmax + FCmax + AUTHmin. That is, we predict that as the measures of EDP and FC go up and the measures of AUTH go down, we expect PG to go up.

EDP indicators linked to PG are: low fertility, (.92)** agree, high life expectancy, (.91)** agree, low infant mortality, (.90)** agree, low population per physician, (.49) agree, low crude birth rate, (.94)** agree, low crude death rate, (.61) agree, and high population urban percent, (.82)** agree, and high population growth rate annual percent, (.83) agree, high population growth rate urban annual percent (.88) agree. The percent of success for these critical variables is 100.0 percent. Thus, many of our original EDP assumptions are confirmed for these variables.

The (FC) variables associated with PG (Power Growth) are: GNP, (.89)** agree, population total, (.88)** agree, high numbers of passenger cars, (.90)** agree, high population urban total, (.72) agree, high urban population percent of total, (.81)** (disagree), (See Footnote 5) and high power, (.84)** agree, high arms exports in millions, (.74) agree, high armed forces in thousands, (.56) agree, high military expenditures in millions, (.89)** agree. The percent of success for these critical variables is 88.8 percent. Thus, again, many of our assumptions have been confirmed with these variables.

The AUTH variables associated with PG (Power Growth) are high civil rights, agree (.68)**, high political rights, (.63)**agree. The percent of success for these critical variables is 100.0 percent. Again, our assumptions have been confirmed for these variables.

 Conclusions for Assumption 5

Assumption 7, relating to EDP (Economic Development Promoters), will be negatively linked to AG (authoritarian growth) because the needed infrastructure to promote civil and political system rights appears related to economic performance and authoritarian regimes do not generally perform well economically.¹¹

Assumption 8, relating to FC (Force Capability) will be negatively linked to AG (authoritarian growth) because military capabilities are linked to economic success, which appears to be a characteristic of more democratic states.^{12, 13}

Assumption 9, relating to LCR and LPR (low civil and political rights measures)) will be positively linked to AG (authoritarian growth) because authoritarian regimes tend to perpetuate themselves and exert efforts to expand their control over their political system. ^{14, 15}

The EDP variables empirically associated with authoritarian growth are high fertility, (.92)** agree, low life expectancy, (.91)** agree, high infant mortality, (.90)** agree, high population per physician,{underline “per physician”} (.49) agree, high crude birth rate, (.94)** agree, high crude death rate, (.61) agree, and high population urban percent (.82)** agree, and high population growth rate annual percent, (.83) agree, high population growth rate urban annual percent (.88) agree. The percent of success for these critical variables is 100.0 percent and our assumptions have been supported.

In this analysis of findings, there were no FC variables associated with AG and, and therefore, no list of critical force indicators was created.

In this tabulation, specific variables associated with authoritarian growth are low civil rights, (.85)** agree, and low political rights, (.83)** agree. The percent of success for these critical variables is 100.0 percent. Our assumptions have been supported for these variables.

We recommend that democratic states encourage further full civil and political rights in authoritarian nations. It should be noted that people in AUTH states might not be able to be in a position to demand better civil and political rights or change the authoritarian institutions which are so ingrained in their daily lives. Whether this should include a forceful overthrow of AUTHmax systems by AUTHmin systems (a hotly debated issue in the current world system) is another matter. It is possible that the destabilizing effects of such “democratic” interventions, into AUTHmax states, may lead to a Type 4 condition, or lose-lose condition, for all parties involved.

In this tabulation, the FC variables associated with conflict growth are: GNP, (.89)** agree, population total, (.88)** agree, high numbers of passenger cars, (.90)** agree, high population urban total, (.72) agree, high urban population percent of total, (81)** disagree, (See Footnote 5) and high power, (.84)** agree, high arms exports in millions, (.74) agree, high armed forces in thousands, (.56) agree, high military expenditures in millions, (.89)** agree. The success specified for these critical variables is 88.8 percent.

Assumption 15, relating to LCR and LPR: Authoritarianism will be negatively linked to COOP (cooperation growth) because authoritarian regimes tend to be more self contained which tend to limit positive cooperative actions toward other nation/states. ^{21, 22}

In this tabulation, the EDP variables associated with cooperation growth are low fertility (.92)** agree, high life expectancy, (.91)** agree, low infant mortality, (.90)** agree, low population per physician, (.49) agree, low crude birth rate, (.94)** agree, low crude death rate, (.61) agree, and high population urban percent, (.82)** agree, and high population growth rate annual percent, (.83) agree, and high population growth rate urban annual percent, (.88) agree. The percent of success for these critical variables is 100.0 percent. Our EDP assumptions have been supported.

In this tabulation, the FC variables associated with cooperation growth are: GNP, (.89)** agree, population total, (.88)** agree, high numbers of passenger cars, (.90)** agree, high population urban total, (.72) agree, high urban population percent of total, (.81)** disagree, (See Footnote 5) and high power, (.84)** agree, high arms exports in millions, (.74) agree, high armed forces in thousands, (.56) agree, high military expenditures in millions, (.89)** agree. The percent of success for these critical variables is 88.8 percent, supporting our assumptions.

In this analysis of findings there were no static LCR and LPR variables associated with cooperation growth, and therefore, no list of critical force indicators was created. It follows that Assumption 15 is not supported by the data.

The above list indicates the FC critical static and dynamic indicators of cooperation growth. Since FC is also linked to CONmax, a Type 3 condition appears in this case, and no policy recommendations are made. Cerven’s argument that conflict and cooperation are strongly linked appears relevant to this finding.

Since there were no findings for Assumption 15, it follows that no conclusions are necessary.

In general, the most economic growth occurs for those nations that are already economically developed. It appears that economically developed nations are more likely to have the resources, or attributes, to maintain and develop economic growth.  The paradox is that many of the things we value associated with EG and PG may be leading us in the direction of more and more conflict. This is apparent from the equations that were supported by the findings. These are:

EGmax = EDPmax + FCmax + AUTHmin,