Writing Armageddon

Writing Armageddon
Furious writing or writing furiously?

Saturday, October 22, 2016

Managerial norms, Peace and the onset of Types of Wars

Managerial norms, Peace and the onset of Types of Wars (a brief look)
by Konstantinos Travlos

Introduction

The First Hague Peace Conference, 1899

Recently I have been reading a lot on the Congress of Vienna. This is for the next article I am working on, on the theme of the causes and consequences of interstate managerial coordination (for the earlier foundations of this concept see Travlos 2014 ). My ideas on this were partly influenced by Peter Wallensteen's early work on universalism and particularism (Wallensteen 1984). In my exploration of those ideas I came face to face with the argument that any inter-state peace built on managerial coordination will have the price of increased warfare of other types, mainly wars between states and polities that are not recognized as states, wars within states, and war between non-state polities. The logic behind this position is that thanks to inter-state peace a) states are more able to engage in aggression against non-states, or interfere in non-state interrelations b) states are more likely to intervene in other states during their civil wars, rather than dealing with the results of the wars on the inter-state level.

Both Wallensteen, and later research by Schahczenski (Schahczenski 1991) produced some empirical indicators on this question. It is also a common theme among historians and scholars who are critical of the Congress of Vienna and the regulatory regime it ushered (see Zamoyski 2008). This is also a general "common sense" position among both radical (i.e Chomsky) and realist (i.e Mearsheimercritics of international order.

In a paper under review, I explain that the Wallensteen and Schahczenski findings were driven by coding decisions concerning Wallensteen's  concepts of "universalism" and "particularism". I argue that if one uses a more sophisticated instrument to capture the quality of managerial coordination, the resultant picture is more optimistic. A replication of the analyses of Wallensteen and Schahczenski provides some support for my argument.

Here I want to do something more simple. First I want to show the amount of war onsets per year in the 1816-2007 period. Such graphs and data is readily available in the debate over the relative bellicosity of different periods of international relations, but usually focused only on either inter-state wars (wars between polities that are legally recognized as states), or intra-state wars (wars within polities recognized legally as states). Here I also include non-state wars (Wars between polities not legally recognized as states), and extra-state wars (Wars between polities legally recognized as states and polities not legally recognized as states). The arguments I am interested in are 1) that there is a relationship between a decrease in inter-state war onset and increase in the onset of other types of war 2) Periods of strong interstate managerial coordination are associated with this phenomenon.


Analysis 

I take my data from the venerable Correlates of War War data-sets (COW-War Data v.4.0). I am focusing in this post on the onset of war, instead of the equally defensible number of ongoing wars or number of wars terminated, for two reasons. First, the main argument of supporters of managerial regimes is that these make the outbreak of war less likely. Second it is easier to collect that data quickly.  I decided to focus on an annual count, and I also deal with each war as an observation, as opposed to breaking it down to participant dyads.

First here are some graphs with the variation in annual war onset in the 1816-2007 period (click for larger versions).








The story these figures might be telling is that the onset of non-state wars and extra-state wars have had a downwards moving trend in the last 200 years of international history. Inter-state wars have followed a curvilinear trend with an increasing trend for the first 100 years culminating in the period between 1900-1950 and then decreasing again. Intra-State Wars had a U trend, with a low period in the middle of the 200 year period. In all cases the 2000-2007 period saw a downward trend.

What happens when we combine them all in one graph (click for a clearer picture)




In general there does not seem to be a general relationship between the variation of the onset of different types of wars. While the U and inverted U trends of Intra-State War and Inter-State War would point to a potential relationship, the graph does not indicate that. Further indicators of the lack of a relationship between the variation in onset of different types of war is garnered by conducting correlations (with chi-square tests of statistical significance at alpha=0.05 - a measure of how likely it is that any relationship we find is due to random chance). Here are the results (correlations of more than 60% are considered worth further exploration)

Extra-State War and Non-State War=  27% (not statistically significant)
Extra-State War and Inter-State War= -0.4% (not statistically significant)
Extra-State War and Intra-State War= 2.7% (not statistically significant)
Non-State War and Inter-State War= -13.90% (not statistically significant)
Non-State War and Intra-State War= -35.73% (statistically significant) 
Intra-State War and Inter-State War=  2.51% (statistically significant) 

Only the variation in the onset of Non-State Wars and of Intra-State Wars were found to have a statistically significant correlation.  In all other cases the variation in the onset of one type of wars had no statistically significant relationship with the variation in onset of other types of war.

These indicators question the argument that variables that affect the variation in the onset of one type of wars (as managerial coordination is argued to have on inter-state war) will have negative(or positive) externalities on the variation in onset of other types of war. This casts some doubt on the position that the costs of interstate peace associated with managerial regimes is an intensification of other types of war (again a favorite argument of radicals and realists against attempts to promote interstate peace).

What about specific operationalisations of managerial regimes? As I said earlier, Wallensteen found that in the 1816-1976 period while universalism was associated with less inter-state war, especially among the major powers, it did not seem to have an effect on other types of warfare.  The table below shows it. I include the raw number of annual onsets in each period, as well as the ratio of total onsets over total years of international relations in the period.  

Wallensteen
1816-1976
Extra-State War Onsets
Non-State War Onsets
Inter-State War Onsets
Intra-State War Onsets
All Types
Onsets
Particularism (75 years)

46 (0.60)


18(0.24)


37(0.5)


49(0.65)


150
(2 onsets a year)
Universalism
(86 years)


46 (0.50)



29(0.33)



24(0.3)



60(0.70)



159
(1.8 onsets a year)

As can be seen, only in the category of Interstate Wars, do periods of universalism see fewer war onsets than period of particularism (universalist periods are those of intense regulatory coordination between the major powers, those of particularism are of intense major power competition and unilateralism). Thus we can defend Wallensteen on the grounds that per his operationalisation of managerial regulation, the empirical facts did indicate that weakness of such norms. However as I argue in the paper under review, there are problems with that operationalisation that require its revision.

My proposed revision is the Interstate Managerial Coordination Scale (IMaC). What story does the data tells us if we use this operationalisation? The following table has the results for the 1816-1976 period of Wallensteen's original analysis.


Interstate Managerial Coordination
1816-1976
Extra-State War Onsets
Non-State War Onsets
Inter-State War Onsets
Intra-State War Onsets
All Types
Onsets
Adversarial Particularism
(23 years)
13(0.56)
5(0.21)
14 (0.60)
12(0.52)
44 (1.9 onsets a year)
Particularistic Regulation
(79 years)
44(0.55)
20(0.25)
29(0.36)
58(0.73)
151
(1.9 onsets a year)
Anemic Regulation
(14 years)
3(0.21)
1(0.07)
4(0.28)
11(0.78)
19 (1.3 onsets a year)
Managerial Regulation
(33 years)
26 (0.78)
15 (0.45)
13 (0.39)
23(0.69)
77
(2.33 onsets a year)
Universalist Regulation
(12 years)
6(0.50)
6(0.50)
1(0.08)
5(0.41)
18
(1.5 onsets a year)

The categories of IMaC are from that of least managerial coordination, to most managerial coordination. Once we break down international relations history by the categories of  the more sophisticated IMaC, we see that the categories of IMaC that are characterized by the expectation of more intense managerial coordination (Managerial Regulation and Universalist Regulation) see fewer onsets of all types of war than the two expected to have the least intense managerial coordination (Adversarial Particularism, Particularistic Regulation). The raw numbers are 195 war onsets of all types in 102 years vs. 95 onsets in 45 years. However we must note that Managerial Regulation, while having fewer raw war onsets compared to Particualaristic Regulation, had more on average per year (roughly 2.3 to 1.9).

The picture that arises is more nuanced than either the argument I make, or the one that the critics of managerial regulation make.
  • Periods of Universalist Regulation are associated the least with  Inter-State War and Intra-State War onsets compared to other periods. They are the most associated with Non-State War onset. 
  • Periods of Managerial Regulation are the most associated with Extra-State War onset.
  • Periods of Anemic Regulation are the most associated with Intra-State War onset, and the least associated Extra-State War onset and Non-State War onset.
  • Periods of Particularistic Regulation are less associated with Non-State War onset compared to periods of more intense managerial coordination, but more associated with Intra-State War onset.
  • Periods of Adversarial Particularism are the most associated with Inter-State 
Discussion of 1816-1976 findings

The findings are thus mixed. The period we would expect to have the most intense managerial coordination norms and regimes (Universalist Regulation)  is indeed less likely to experience Inter-State, and Intra-State War onset. This decrease does not seem to be associated with increased Extra-State war onset in comparison to periods of ambiguity (Anemic Regulation), though it is still less than periods expected to be characterized by very little managerial regulation. However, it does seem to experience more Non-State War onsets than other periods. One explanation might be that this period is associated with the Congress of Vienna (1816-1823), a period during which European imperialism and state centralization, had not destroyed other non-state polities yet, creating space for them to fight each other.

However when we move to the period that we do not expect to be characterized by the most intense managerial coordination, but still have higher levels than the other periods (Managerial Regulation), the record is mixed.  While still experiencing less Inter-State War onsets than the periods we would expect to be characterized by  less managerial coordination, it experiences more Intra-State War onsets than the period we expect to be characterized with the least managerial coordination (Adversarial Particularism). It also outperforms all periods in Extra-State War onset, and outperforms the less managerial periods in Non-State War onset. It is here that Wallensteen's and Schahczenski's finding reside, and it is the events of this period that give the critics of managerial regulation their ammunition. 

The ambiguous patterns on regulation of Anemic Regulation are associated the most with Intra-State War onset, while other types of war are either less likely than all other categories (Extra-State Wars,Non-State Wars) or less likely than all other categories bar Universalist Regulation (Inter-State Wars). One hypothesis might be that such periods characterize transition periods for the inter-state system from a cycle of concentration and regulation, to a cycle of de-concentration and warfare (Thompson 1986). The increased incidence of civil war might indicate the collapse of  the domestic status-quo of states that were too associated with the fraying international status-quo.

The period to be characterized by the presence of intense adverseness among states, but still retaining at least the shell of regulatory norms (Particularistic Regulation) has more Inter-State, Intra-State, and Extra-State War onsets than the period of most intense regulation (Universalist Regulation), and more Intra-State War onset than Managerial Regulation. However, it has a similar association (or perhaps a bit less likely) to Inter-State war onset with Managerial Regulation. It has more Extra-State War onsets than Anemic Regulation, but less Non-State War onset compared to the periods associated with more regulation.

The period we would expect to be characterized the least by managerial regulation (Adversarial Particularism) has the most Inter-state war onsets of all the periods and the second highest incidence of Extra-State War.

Taken together all of the above should caution both sides of the debate over world order. There are types of managerial coordination that while being less associated with Inter-State War , are more associated with violence against non-state polities, or between non-state polities. But that is not the case with all types of managerial coordination. Conversely, periods of frequent Inter-State warfare are not necessarily going to be associated with the avoidance of other types of war. Essentially, managerial coordination in itself is likely associated with less Inter-State and Intra-State war onset compared to periods that eschew it, but whether that pacific influence applies to other types of war will depend on the specific tuning of the norms and institutions associated with it, in another name politics (Compare the Iroquois Confederacy as a managerial system to the Vienna Congress for an example of the importance of politics. See Neta C. Crawford 1994.

What Happens if we extend IMaC to 2007


The above findings are located in the 1816-1976 temporal domain. That was the time period of the data available at the time of Wallensteen's and Schahczenski's analysis. But we can extend the analysis of IMaC and onset of type of war to 2007.


Interstate Managerial Coordination
1816-2007
Extra-State Wars Onsets
Non-State Wars Onsets
Inter-State Wars Onsets
Intra-State Wars Onsets
All Types
Onsets
Adversarial Particularism
(23 years)
13(0.56)
5(0.21)
14(0.60)
12(0.52)
44 (1.9 onsets a year)
Particularistic Regulation
(92 years)
48(0.52)
21(0.22)
33(0.36)
70(0.76)
172(1.86 war onsets a year)
Anemic Regulation
(14 years)
3(0.21)
1(0.07)
4(0.28)
11(0.78)
19 (1.3 onsets a year)
Managerial Regulation
(52 years)
28(0.54)
15(0.29)
24(0.46)
41(0.79)
102(1.1 war onsets per year)
Universalist Regulation
(12)
(12 years)
6(0.50)
6(0.50)
1(0.08)
5(0.41)
18
(1.5 onsets a year)

The inclusion of the 1977-2007 period brought in more years that fit the Managerial Regulation and Particularistic Regulation categories of IMaC. One big change is that now Managerial Regulation  is the period with the least number of average war onsets per year (1.1). Furthermore it is less likely to experience Extra-State War onset than Adversarial Particularism, but has similar associations to that onset to Particularistic Regulation. It's previous association with Non-State wars is now less pronounced. These are indicators in support of the pacification thesis. However it has become as likely, or more likely, to experience Intra-State war onset as all other less regulatory periods. This is an indicator in support of the thesis that the price for inter-state peace is increased chance of the onset of other types of war.  Thus the discussion of the 1816-1976 findings is still relevant.

Finals Thoughts-Conclusion

I wrote this post in reaction to criticisms of the Vienna Congress I read in books I am reviewing for a paper on the theme of the causes of consequences of managerial coordination in international relations. The criticism is that while managerial coordination regimes can decrease the likelihood of the onset of inter-state war, the cost for this is a greater likelihood for the onset of other types of war, especially wars associated with state expansionism against non-state polities, and civil wars.  I decided to start a preliminary evaluation of this argument by first looking at the straight numbers of annual war onsets, for all types of wars, using COW Data, in the 1816-2007 period and seeking relationship. The simple evaluations here did not indicate support for an argument of relationship.

In the next step, I presented a less sophisticated version of my under-review analysis of the findings of  Wallensteen (1984) and Schahczenski (1991) that found support for the thesis that the decrease in inter-sate war onset associated with universalism (periods of managerial coordination) was associated with increase in the onset of other types of war. I argued that a more sophisticated measure of major power managerial coordiantion (IMaC) would provide evidence against the thesis. The findings in the 1816-1976 and 1816-2007 period were mixed. They indicated that different types of managerial regulation might have differing effects on the likelihood of the onset of different types of war. While in general the periods of more managerial coordination tended to be more pacific than those of less managerial coordination, the effect was not always uniform across all types of war onset. Both side of the debate will need to be more careful, and put more importance on the actual politics of regimes.

Some important points before ending
1) The above is not a full social scientific analysis. There is a lack of control variables, and the evaluations are very basic. The results should not be considered more than tentative indicators.

2) Just looking at the likelihood of onset of war, or even the raw numbers as I do here, is not enough to evaluate the pacific character of interstate managerial coordination regimes. Evaluations of number of ongoing wars, and of war terminations would also be needed. Equally important is a discussion of the character of peace (for a recent discussion see Goertz, Balas and Diehl 2016), and the quality of peace (for recent discussion see Wallesnteen 2015.)

3) It should be noted that social scientific findings are probabilistic, and that just because there was a certain trend in the past one should not extrapolate it to the future. The period in question is the period of the global expansion of a liberal capitalist system globally. The fact that this system's liberal component seems to be retreating might completely invalidate all lessons from the past.  All we can do as social scientists if provide potential negative and positive consequences for various suggested policies.  

No comments:

Post a Comment