October 27, 2004 | Center for Data Analysis Report on National Security and Defense
The proposed global redeployment of U.S. troops coupled with the open question of how long U.S. forces will stay in Iraq highlight the need for objective data on force deployments, objectives, and results. Surprisingly, no comprehensive time series data on U.S. troop deployments by year and country seem to exist in a single dataset. This report remedies that need by introducing a comprehensive troop deployment dataset for 1950–2003. The Troops dataset is available here: www.heritage.org/Research/NationalSecurity/troopsdb.cfm.
Five Decades of Extensive Foreign Deployments
Those who closely follow U.S. defense posture know that President George W. Bush’s August 2004 announcement of a proposed global redeployment of U.S. troops was predated by many years of preparation and planning. "The world has changed a great deal and our posture must change with it," President Bush said. The plan will withdraw 70,000 troops from Europe and Asia, continuing the departure from the Cold War stance of 1950–1990.
Although the announced moves have been characterized by some as dramatic -- even risky -- they are actually part of an ongoing process of threat assessment and long-range planning at the Pentagon. For example, the major post-Cold War drawdown of U.S. troops in Europe occurred in the early 1990s, when hundreds of thousands of American soldiers were brought home from Germany.
The current proposed realignment is thus neither the first nor the last step in modernizing America’s global military posture, although it does raise the question anew: What is the role of the U.S. military in the world? To begin, we should ask what the global role of the U.S. military has been in the past, and whether it has been effective.
However, no comprehensive source exists that could show where U.S. troops have been deployed for the past 50 years. The Center for Data Analysis has remedied that by providing a comprehensive troop deployment dataset for 1950–2003. Although the Pentagon publishes annual troop deployment figures, it was necessary to compile the data into an integrated, comprehensive database to compare troop locations across different years. The integrated data series described below offers a unique perspective on the constant change involved in troop deployments abroad.
On average, 2.3 million U.S. troops were on duty per year from 1950–2000. Of this average, 535,000 troops (23 percent of all military personnel) were deployed on foreign soil. Chart 1 shows the time series of the percent of troops deployed abroad. The pattern of deployment varies over time, as deployment locations and levels respond to changing threats as well as foreign wars.
Using the integrated data, we can measure how many countries host a given number of troops every year and how that pattern of deployment breadth evolves. Chart 2 shows the number of countries hosting 100 or more troops per year over time and the number of countries hosting 1,000 or more troops per year.
The dataset does not distinguish the type of personnel by military branch or mission. Researchers may find it useful to make such a distinction. For the most part, U.S. troops were stationed abroad as part of the Cold War system of deterrence through alliances with nations such as Japan, South Korea, and NATO member nations. While such defensive missions were the norm, troops sent to Korea in the early 1950s and to Vietnam during the 1960s were actively at war, as are those currently in Iraq. A qualitative description of troop deployments would have to distinguish these situations from the norm. Combat troops can be further categorized in two ways (at a minimum) -- attacking or defending. Non-combat missions could be further described in terms of deterrence, pacification, engineering, diplomacy, and so on. The most important classification is whether U.S. troops are welcome or unwelcome by the host nation. In the overwhelming percentage of cases, they have been welcome as part of a mutual agreement.
Asymmetric Posture in Five World Regions
The commitment of American soldiers has been consistent in Europe, varied in Asia, and shallow in the other parts of the globe. Chart 3 gives a decade-by-decade accounting of troop deployment by major region. The number of U.S. troops in Europe and Asia dwarfs the scant troops stationed in Africa, the Middle East, and the Americas (excluding the United States). During the second half of the 20th century, 52 percent of deployed troops were in Europe and 41 percent in Asia. Most of those deployments can be traced to a handful of countries: Germany, Japan, and Vietnam. Many nations still have thousands of U.S. troops on their soil as part of long-standing alliances, but many of the countries that have been home to American forces would be a surprise to contemporary perceptions, because they include France, Spain, Portugal, and even Libya.
Germany, by a wide margin, has the highest troop-year total for every individual decade and for the entire 1950–2003 period. Chart 4 is a graphical list of the 20 countries in order of the total amount of troops. Japan, South Korea, Vietnam, and the United Kingdom round out the top five countries that have hosted the most U.S. military personnel and are included in Chart 5 as time series over the entire period. Evident in Chart 5 is the rapid rise and decline of the number of troops sent to Vietnam -- from less than 1,000 personnel in 1961 to 537,000 in 1968, then quickly down to zero seven years later. Chart 6 shows time series for the next five countries with the highest total deployments: France, the Philippines, Italy, Panama, Thailand, and Spain.
The following subsections describe in some detail the changing force composition in the five regions over the past 54 years.
Region 1: East Asia. The historical roots of the U.S. are almost entirely European, but the 20th century saw a deepening engagement into the affairs of Asia. Most of that engagement focused on Japan and the postwar occupation, but the 1950s were also dominated by the Korean War. As a consequence of these two developments, U.S. bases have become a seemingly permanent feature of East Asian security. The data ironically indicate a sharp pullback of U.S. forces from Asia in the 1970s and 1980s, but this is primarily driven by the end of the Vietnam conflict and complete withdrawal from there. Troop commitments to Japan and South Korea formed a bedrock of U.S. engagement in Northeast Asia.
What might have been viewed as an occupation of Japan after 1945 was instead transformed by the threat of communist China and the Soviet Union into a mutually beneficial security arrangement and strong alliance. Other postwar relationships also evolved into long-standing alliances involving heavy U.S. troop commitments, notably in the Philippines and Taiwan.
The U.S. role in Korea was motivated by the sudden attack from North Korea, which ultimately served to validate the strategy of containment, that is, the need to aggressively contain communist ambitions. Only 510 servicemen were based in South Korea in 1950, prior to the attack. U.S. Department of Defense (DOD) records show that 326,863 troops were deployed in South Korea in 1953, a number that stabilized between 50,000 and 60,000 in the 1960s and 1970s. A slow drawdown continued as troops averaged 40,000 in the 1980s and 35,000 in the 1990s.
America positioned thousands of troops in Hong Kong and Taiwan, abruptly starting in 1954. At their peak, there were 4,539 American soldiers in Hong Kong in 1957, which were quickly drawn down to levels of 290 and then 26 over the next two years. The situation was similar in Taiwan, with a sudden buildup from 811 to 4,174 troops in 1954, peaking in 1958 at 19,000 and then stabilizing between 4,000 and 10,000 until 1977. All American forces were pulled out in 1979, a withdrawal that clearly began in 1973 after President Nixon’s diplomatic opening with the People’s Republic of China.
The war in Vietnam brought hundreds of thousands of American military personnel from all branches to many countries in Southeast Asia. Hundreds of soldiers had been in Vietnam since the end of World War II, but 1956 marked the beginning of new deployments into Cambodia and Laos. Troop levels in Vietnam proper accelerated sharply in 1962, peaked in 1968 at over 500,000 servicemen, and then declined sharply every year after, ending with complete pullout in 1975. Thailand also saw large deployments of tens of thousands of Americans during 1962–1975, when it hosted large U.S. airbases that were heavily used to support combat operations in Vietnam.
The Philippines hosted a steady level of 15,000 U.S. billets per year throughout four decades, with large Air Force and Navy bases. Then a series of protests, capped by a devastating volcano eruption that devastated the bases, severed the deployment relationship after 1992.
Region 2: Europe.It is impossible to consider the defense posture of American forces in Europe without being impressed by the dominant position of Germany, which was split in half by post-1945 occupations of Soviets in the East and the other Allies in the West. As relations hardened between the Cold War adversaries, Germany became the primary potential battleground for World War III and U.S. troop levels tripled from 1950 to 1953. Every year for nearly four decades, one-quarter of a million troops were billeted in West Germany, but by 1993 the number had dropped to 105,254. In 2000, just 69,203 American military personnel remained.
France had some 50,000 U.S. troops based on its soil for decades, but that number was reduced to less than 100 during the mid-1960s.
Across the rest of Europe, there have been many different hosting relationships over the half-century. Britain saw the withdrawal of tens of thousands of U.S. soldiers after 1990. In Italy, the number of U.S. servicemen has remained remarkably stable to this day, averaging about 12,000. In contrast, a similar deployment level in Spain of 9,000 troops was trimmed sharply to 2,000 after 1989. The end of the Cold War also led to U.S. personnel reductions of 50 percent or more in countries such as Portugal, Iceland, Greece, and the Netherlands. The major exception was the deployment of 15,003 soldiers to Bosnia and Herzegovina in 1996 and the steady presence of 3,000 troops there ever since.
Region 3: Middle East. Turkey is categorized here as a Middle Eastern country, but has equally as much European heritage. Turkey’s close friendship with the U.S. during the Cold War included the basing of 5,000 to 10,000 American troops from 1957 to 1992, when a slow drawdown began. Roughly 5,000 American troops were in Libya per year during the 1950s and 1960s. The only other major deployments to the Middle East were in Kuwait and Saudi Arabia beginning in 1990. In recent years, the U.S. has deployed troops to new bases in Qatar and Bahrain.
However, it would be a mistake to think American forces have not been widely deployed in the Middle East. Thousands have been based in Iran, Pakistan, Morocco, and Saudi Arabia over the decades. Hundreds more have been based consistently in Egypt, Lebanon, Israel, and India.
Region 4: Africa. Africa is the clearest case of non-involvement by U.S. military forces during the past five decades. A handful of African nations saw very minor deployments beginning in the 1960s, but nothing on the scale of strategic basing seen in other continents. The exceptions to this rule are Libya (described above) and Ethiopia, which hosted over 1,000 American soldiers every year from 1957 to 1973. By the 1980s, most African nations had minor troop deployments of five to 50 American soldiers. This may signal more engagement and is certainly attributable to more than better recordkeeping by the Pentagon, given that small deployments in other regions in the 1950s and 1960s were recorded.
Region 5: Americas.American forces have been broadly, lightly based throughout Latin America since 1950. The strongest relationships in the hemisphere have been with Canada, Panama, and Bermuda, each with basing commitments of 3,000 to 15,000. The number of U.S. troops stationed in Canada declined sharply from 1965 to 1975 and is now only about 150. Most other nations in the region hosted U.S. forces year after year -- usually less than 100 -- and none had annual billets over 1,000.
The President’s proposed redeployment of 70,000 troops from foreign countries to domestic bases has been greeted as a major movement, but it needs to be kept in perspective. An average of 311,870 troops were stationed in Europe per year during 1986–1990. That force was slashed by two-thirds after the Berlin Wall fell, to an average of 109,452 troops per year during 1996–2000.
U.S. troops were disengaged (i.e., reduced) from Africa in the early 1970s. In Latin America, the disengagement was sharper, declining from an average of 37,000 soldiers annually in the 1950s to 28,000 in the 1960s to a steady 14,000–16,000 thereafter. Troop deployments to the Middle East were cut in half during the 1970s and 1980s, but redoubled in the 1990s.
A final analytical consideration is the impact of U.S. troops on foreign countries. In postwar Germany and Japan, the goal of the U.S. presence was Cold War peace through deterrent strength. One can argue that the mission was largely successful, not only for the U.S., but also for the economic development of the two respective regions.
On the other hand, one could also argue that the dominance of American forces can serve as a crutch of inaction by others. For example, the tragedy of genocide in Bosnia was unchecked without U.S. leadership. Yet is the primacy of U.S. engagement in addressing obvious atrocities a healthy development?
No other military in world history has been so widely deployed as that of the United States. Troop deployments are overwhelmingly supportive of host countries, and warm relations between soldiers and local populations are the norm. However, the first priority in deployment strategy is not a particular foreign government’s desire to keep a certain number of American troops in its country, but the American need to align its forces against contemporary and future threats. Better data about the deployment levels of American forces will hopefully contribute to an understanding of the consequences of past strategies and the development of future strategies.
Tim Kane, Ph.D., is Research Fellow in Macroeconomics in the Center for Data Analysis at The Heritage Foundation.
Data and Methodology
The original annual data come from the Statistical Information Analysis Division (SIAD) of the Directorate for Information Operations and Reports (DIOR) in the U.S. Department of Defense. The SIAD makes the information publicly available via its Web site, although it is available only on an annual basis, not as a time series. Data for years 1950 through 1977 are in Microsoft Excel format, and data for following years are in Adobe Acrobat format.
Our first research objective was to create a comprehensive disaggregated table showing how many U.S. troops were deployed to each unique country during every year studied. The disaggregation allows one to correctly account for countries that underwent changes in name or geographic boundary. For example, modern Germany is a fusion of former East Germany with West Germany, while at the end of 1991 the Soviet Union disintegrated into 15 independent republics. The disaggregated table includes 250 countries in rows, along with 54 columns representing troop counts for the years 1950–2003.
The second objective was the creation of an integrated table which aggregates varied country names and associated data into a single time series. For example, the Japan data include data on Japan, as well as Okinawa and the Bonin Islands, which were counted separately during the 1950s and 1960s. Scotland (1954–1955), Trieste (1950–1954), and the Azores (1950–1969) were uniquely identified in DOD records and were integrated into the United Kingdom, Italy, and Portugal records, respectively, for the "integrated" spreadsheet. Others like Myanmar have integrated data for Burma, its longstanding country name in earlier records. Then there are cases of continuation on the integrated sheet, such as Russia, which denotes contemporary Russia and data from the former Soviet Union. The resulting integrated sheet gives a clearer overview of U.S. troop deployments over the past 55 years.
The integrated data have some inherent limitations:
Tim Kane, Ph.D., Troop Deployment Dataset, 1950-2003, The Heritage Foundation, Center for Data Analysis, October 2004, at www.heritage.org/Research/NationalSecurity/troopsdb.cfm . Lucia Olivera and Jon Casale provided invaluable research assistance in preparing this data set. More detailed data are available upon request. See the Appendix for data and methodology details.
U.S. Department of Defense, Directorate for Information Operations and Reports, Statistical Information Analysis Division, "Military Personnel Historical Reports: Active Duty Military Personnel by Regional Area and by Country," at www.dior.whs.mil/mmid/military/history/309hist.htm October 19, 2004).