|
Article Excerpt
On February 28, 2007, former Rep. Martin Meehan (D-MA) and a bipartisan group of co-sponsors reintroduced the Military Readiness Enhancement Act in the House of Representatives to amend title 10 United States Code [section] 654 ("Policy Concerning Homosexuality in the Armed Forces") to enhance the readiness of the Armed Forces by replacing the current "don't ask, don't tell" policy with a policy of nondiscrimination on the basis of sexual orientation. In the recent DUKE JOURNAL OF GENDER LAW AND POLICY article, "Constructing the Co-Ed Military," Elaine Donnelly, president of the Center for Military Readiness, asserts that "nothing has changed that would justify the turmoil that would occur in and outside of Congress if Meehan's legislation were seriously considered or passed." (2) But on what evidence is she basing her claims that turmoil would ensue if 10 U.S.C [section] 654, the ban on openly gay service members, were repealed?
The outcomes of repeal are exactly the points that an informed public conversation about the "don't ask, don't tell" policy should be engaging, but in a serious, evidence-based debate. And the focus of that debate should be not on fears of what would "occur in and outside of Congress," but on the impact of any legislation on military readiness. Signed into law in November of 1993 by President Clinton, the Defense Department issued the first set of comprehensive regulations in February of 1994. The "don't ask, don't tell" policy, once considered an interim measure by policy makers, has remained in force, relatively unchanged, for over fourteen years. (3)
Contrary to Donnelly's assertion, however, much has changed in the military, political, and cultural landscape since 1993. Military opinion and public opinion have experienced dramatic shifts, which have been well documented by scholarly research and in the media. The most recent evidence of these shifts, and perhaps the most telling, was a statement released by a group of twenty-eight retired U.S. generals and admirals urging Congress to repeal the current ban on openly gay, lesbian, and bisexual troops. (4) The support of such a large number of senior military officers for an end to the so-called gay ban reflects nothing less than a sea change in military opinion on the issue. In 1993, when the current policy was formulated, some surveys found that 97 percent of generals and admirals opposed lifting the ban. (5) When General John Shalikashvili, who served as Chairman of the Joint Chiefs of Staff from 1993 through 1997, published an op-ed in The New York Times on January 2, 2007 calling for the end of "don't ask, don't tell," he cited polls showing that a large number of younger enlisted personnel also favor letting openly gay soldiers serve. (6) That poll of 545 troops who served in Afghanistan and Iraq by Zogby International, found that 72 percent of service members are personally comfortable interacting with gays and lesbians. (7) Public opinion polls show similarly strong indicators of change. Furthermore, data indicate that the policy is now harming the military's reputation because it is out of step with public opinion. (8)
Donnelly contests all of this evidence. She says that the research supporting the claim that discrimination undermines the military, and that integration would enhance military effectiveness, is not compelling. And she suggests that, "A closer look at materials produced by the activist groups usually reveals questionable methodology and unsupported conclusions." (9) The Palm Center, a research institute at the University of California, Santa Barbara, stands at the core of her offensive against full integration of openly gay and lesbian service members. She repeatedly attempts to dismiss the credibility of the Palm Center's research by labeling our data the product of "social engineers" and "activists." (10)
In the following pages, we respond to the substance of Donnelly's critique, addressing the factual errors in her analysis, addressing her unsupported assertions about the quality and integrity of research in this area, in particular by the Palm Center, and commenting on the stakes raised by the rhetoric Donnelly chooses to deploy in presenting her position in lieu of evidence to support that position. (11) An analysis of the substance of her complaints shows that her critique is without merit, that the methodologies behind the studies she cites are in fact sound, and that the data show that discrimination compromises military effectiveness, while integration enhances it. We agree with one aspect of Donnelly's argument: when lives and national security are at risk, basic assumptions must be challenged and objective analysis is at a premium. And that is the place from which the Palm Center approaches our research.
THE CENSUS KNOWS: ESTIMATING THE NUMBER OF GAY AND LESBIAN SERVICE MEMBERS
One of the basic questions that inform public conversation about "don't ask, don't tell" involves the number of gays and lesbians serving in the armed forces. Some commentators have suggested that ten percent of the American public is gay or lesbian (12), and have applied this figure to the military, suggesting that ten percent of the armed forces, or 250,000 troops, may be gay. Unfortunately, the "don't ask, don't tell" policy itself precludes conducting a survey of active-duty military personnel that asks respondents about their sexual orientation or behavior. As a result, scholars have had to consider other approaches and methods to try to identify this otherwise hidden population of lesbians, gay men, and bisexuals serving in the United States armed forces. In 2004, Dr. Gary Gates of the Urban Institute published a study which found that approximately 65,000 currently-serving troops are gay. (13) Donnelly attacks Gates' study, which she says amounts to "urban legend" rather than "a serious piece of scholarship," (14) but virtually every assertion Donnelly makes in describing the study is incorrect.
Donnelly's first incorrect assertion is that the Urban Institute analysis includes a "speculative claim" that in the U.S. four percent of men and three percent of women are lesbian or gay. (15) In fact, findings from a study published after the Urban Institute study suggest that these figures actually underestimate the fraction of lesbian, gay, and bisexual identified men and women in the United States. (16) A federally-funded nationally representative survey of men and women aged 18-44 found that 4.1 percent of both men and women self-identified as either "homosexual" or "bisexual." (17) This suggests that the Urban Institute study likely underestimates the size of the lesbian, gay, and bisexual population.
Donnelly continues with a claim that the study "speculated that household-mates of the same-sex are homosexual." (18) The fact that same-sex "unmarried partners" identified in the U.S. Census are likely lesbian and gay couples is not speculation. To begin with, the same-sex couples identified in the U.S. Census are not simply "household-mates" (19) as the author suggests. Rather, they are same-sex couples in which one partner is explicitly identified as either the "husband/wife" or "unmarried partner" of the other partner. (20) Therefore, those identified as "roommates," "boarders," and "unrelated adults" are not included among these couples. Nor are any same-sex couples with blood relationships such as siblings, cousins, or children. Social scientists and government officials have closely considered this question of categorization and are in broad agreement that same-sex "unmarried partner" couples identified in the Census are indeed lesbian and gay couples. (21)
In the specific case of military service, same-sex couples appear to be a reasonable proxy for the broader lesbian, gay, and bisexual population. Black and colleagues compare military service rates among same-sex couples in the Census to men and women in the General Social Survey (a nationally representative sample of adults) who indicate that they have had exclusively same-sex sexual partners in the last five years. (22) The estimates, it turns out, are virtually the same.
Donnelly incorrectly states that the analyses are based on the fraction of individuals in same-sex couples who identify as "veterans." (23) In fact, the estimates for the number of lesbians, gay men, and bisexuals in the military are based on the fraction of individuals in same-sex couples who indicate that they are on active duty in the military and, separately, on those who indicate that they are in the ready reserve forces (National Guard and Reserve). The report uses these figures to estimate the number of lesbian, gay men, and bisexuals on active duty and in the ready reserve. The analyses also separately consider the number of lesbian, gay, and bisexuals who are veterans using only those in same-sex couples who state that they have prior military service.
Finally, Donnelly claims that the Urban Institute report uses "questionable methodology." (24) In fact, the report uses well-known statistical procedures and conservative assumptions to reach its conclusions. The primary statistical analyses use Bayes' Rule, a common procedure frequently used by statisticians. To estimate the size of the lesbian, gay, and bisexual population currently serving in the military using Bayes' Rule, the analyses rely on two key assumptions, both of which err on the side of a conservative estimate:
1. Four percent of men and three percent of women in the U.S. population are lesbian, gay, or bisexual. As already stated, Mosher et al. (2005) find that 4.1 percent of both men and women identify as such. (25)
2. The fraction of individuals in same-sex couples who report being on active duty or in the ready reserve is the same as the fraction of all lesbian, gay men, and bisexuals serving in the military. As the report clearly states, this likely underestimates the latter as it seems reasonable to believe that single lesbians, gay men, and bisexuals can more easily serve and hide their sexual orientation (as the "don't ask, don't tell" policy dictates) than can their coupled counterparts. As such, the statistic derived from same-sex couples likely underestimates the real fraction of lesbians, gay men, and bisexuals presently serving in the military.
Far from "urban legend," the Urban Institute study is based on high-quality data, sound statistical techniques, and conservative assumptions. The report offers a credible estimate for the size of the lesbian, gay, and bisexual population serving in the U.S. military: an estimated 65,000 active and reserve military members are currently serving. (26)
THE COST OF FIRING GAYS AND LESBIANS
What is the financial cost of firing gays and lesbians from the military? In February 2005, the Government Accountability Office (GAO) released a report that found that during its first ten years, the "don't ask, don't tell" policy cost $190.5 million to implement. (27) The Palm Center organized a Blue Ribbon Commission to study the GAO's report, and the Commission included distinguished experts including professors at military universities, a retired U.S. Army Colonel with a Ph.D. in economics, and a former Secretary of Defense. The Commission's work was vetted by an accounting professor from the Naval Postgraduate School.
The Commission found that the GAO's findings were based on questionable data and methodology, and determined that the GAO's errors led to both over- and under-estimations of the total cost of implementing "don't ask, don't tell." When the over- and under-estimations were reconciled, it was determined that the GAO's overall estimate was 91 percent too low, and that the "don't ask, don't tell" policy cost the Pentagon at least $363.8 million to implement during its first ten years. (28) Because the Commission used conservative assumptions, it argued that its finding should be seen as a lower-bound estimate. (29)
Donnelly questions the Commission's findings on two grounds. First, she says, the issue is not the cost of implementing "don't ask, don't tell." Rather, she claims, "It is the cost of recruiting and training individuals who are not eligible to serve in the military because they are homosexual." (30) She argues, "losses related to the homosexual conduct law, whatever it is, could be reduced to near-zero if all potential recruits were fully and accurately notified that the 1993 law means that homosexuals are not eligible to serve." (31) The flaw in Donnelly's reasoning, however, is that even if "don't ask, don't tell" were interpreted, as she would have it, to prohibit all gays and lesbians from serving (which we show below it does not), the financial cost of the policy would still be large. From the 1940s until 1993, gays and lesbians were prohibited from serving as a result of a complete and outright ban, and during this time the military fired approximately 2,000 people each year for being gay. (32) (Approximately 1,000 per year have been fired under "don't ask, don't tell.") (33) Hence, even under a complete and total ban, the financial cost of firing gays and lesbians would be as high or higher than it is under "don't ask, don't tell."
Second, Donnelly questions the Blue Ribbon Commission's findings and defends the GAO's original report by noting that the GAO "'stood by'" its original analysis even after the Blue Ribbon Commission had critiqued it. (34) It is correct that the GAO "stood by" its original report, but this does not mean that its original report was sound. As the Blue Ribbon Commission first explained in its report, and then reiterated in an analysis of the GAO's defense by Blue Ribbon Commission Chair Aaron Belkin, the GAO misrepresented a critical piece of cost-of-training data that led to a $150 million dollar error. (35) In previous GAO studies, and as is widely reported throughout the military manpower / personnel literature, the cost to train one enlisted service member, conservatively, was approximately $25,000 to $30,000 in the 1990s. (36) In its original report, however, the GAO contradicted its own prior studies and reported that it cost only $6,400 for the Army to train one soldier. (37) By using inaccurate data to estimate the cost of training enlisted personnel who were subsequently fired for being gay or lesbian, the GAO missed approximately $150 million of costs. Donnelly and the GAO can only stand behind the original GAO study if they believe that it cost $6,400 to train one soldier in the 1990s. This figure, however, is not supported in the literature, and seems particularly off-target when one considers, as the Commission found, that service members fired for being gay received, on average, more initial skill and mid-career training than other members of the military.
WEIGHING THE EVIDENCE FROM FOREIGN MILITARIES
One central question in the U.S. debate over gays and lesbians in the military is whether cohesion, readiness, morale, and recruiting have suffered in foreign military forces that have lifted their gay bans. According to some social scientists, foreign military experiences provide an opportunity for...
|
