Going the Distance: The Price of Projecting Power

Abstract

The central purpose of this article is to establish the relationship between power projection, technology, and economic power. How economically powerful does a state need to be before it can afford the capital intensive technologies, foreign bases, and military and logistical forces associated with global power projection? The specific research question we focus on in this article is: What determines how far states send their military forces? We argue that as the costs associated with projecting power decrease or as the wealth necessary to project power increases, states will project power more frequently and at greater distances. We use a system level time series analysis from 1870–1936 and a dispute level analysis on all militarized international disputes from 1870–2000 to test these propositions. This article is the first to demonstrate empirically that the distance and frequency of power projection is a function of the cost of projecting power. We close with a discussion of contemporary states building power projection capabilities and how future research might build from our research to explain this behavior.

Keywords:

• distance
• geography
• loss of strength
• militarized interstate disputes
• power projection

Acknowledgments

Earlier versions of this paper were presented at the 2011 meeting of the International Studies Association and as a poster at the 2011 IGCC So-Cal Symposium at the University of California, Irvine, and the 2010 meeting of the Peace Science Society (International). We would like to thank Megan Becker, Alex Braithwaite, Scott Bennett, Erik Gartzke, Miles Kahler, David Lake, Yon Lupu, Paul MacDonald, Blake McMahon, Alan Rozzi, Christina Schneider, and Branislav Slantchev for comments on earlier drafts of this paper. All data and code used in this project are publicly available at a Dataverse archive maintained by the authors, http://hdl.handle.net/1902.1/18567, and are cross posted at the Dataverse archive maintained by International Interactions: http://dvn.iq.harvard.edu/dvn/dv/internationalinteractions. All questions regarding the data and code should be directed to the authors. Both authors contributed to study design, data collection, analysis, and preparation of the manuscript.

Notes

¹See Mearsheimer (2001:135) for a discussion.

²See Lake (1999:280) for a discussion.

³See Blechman and Kaplan (1978:8) for a discussion.

⁴Assuming that the returns for security cooperation are constant, a decrease in the cost of power projection should result in an increase in the number of cooperative security arrangements. See Lake (1999:281) for further discussion.

⁵See Farewell (1972:XVII) for further discussion.

⁶We define internal rents as garnered from inside the state; external rents are rents extracted from outside the border of the state.

⁷This is not to say firms do not seek rents, but that rent-seeking is a primary method of generating profits for states.

⁸We selected this subset of cases to gain a more accurate measure of whether a state had actually projected military power. At lower levels of hostility, such as 3 or below, states can simply make a show of force, or threaten to use force and therefore might not actually be projecting power.

⁹Note that our objective is to measure whether a state projects power past a given distance threshold, not whether states initiate a MID at a certain distance. Consider for example, the United States, which might project power 8,000 miles to the Taiwan Straights, but if the Chinese initiate the MID, then they would be coded as the aggressor. We are more interested in capturing the fact that the United States projected military power 8,000 miles rather than the fact that once the United States carriers were there, a MID was initiated by China. Who initiates the MID is often the result of chance or error (Gartzke 1999), whereas the decision to project force is a more deliberate choice. For this reason, and because we are trying to capture the decision to deploy military force over distance, we code the state that projected the greatest distance as the projecting state.

¹⁰In future research, we develop a measurement model that estimates the “true” value of global shipping using computational measurement tools developed for comparative and international relations research (Fariss and Schnakenberg 2013; Schnakenberg and Fariss 2012; Treier and Jackman 2008). This research builds on insights from historical research on the development of diesel engines and gas turbines periods Smil (2007).

¹¹Note that the shipping data from Lake (1999) begin in 1869 but the system level tests begins with the year 1870 because of the inclusion of lagged values of these variables.

¹²We only repeat the tests of H2 for the frequency of MIDs occurring in a system year greater than a set distance. We do not repeat the frequency analyses for H1 because we were interested in the total number of MIDs in the system each year for those tests.

¹³We estimated these models without dyads that contained the United States since this state projects power in many dyadic disputes during the time period of our analysis. The results are robust to the exclusion of these dyads.

¹⁴This is a military acronym for “command, control, communications, computers, intelligence, surveillance, and reconnaissance.”