THE PUBLIC VIEW OF THE MILITARY. John D. Blair and Jerald G. Bachman

THE PUBLIC VIEW OF THE MILITARY 4066 Jhn D. Blair and Jerald G. Bachman Public attentin was increasingly fcused n the U.S. military during the curse f the Vietnam War. Many negative aspects f the military and its persnnel were highlighted. These included financial scandals amng senir enlisted men, blatant mistreatment f recruits, the specter f. widespread drug use, anp the nightmare^f My Lai type atrcities^f! unknwn' dimensins'. These revelatins and the failure t achieve' military ( supremacy despite massive expenditures f resurces (bth' men and material), the increased utrage ver the war and the draft expressed n campuses and in mass rallies and marches, and the grwing disaffectin with the war evidenced in the plls led many t believe that the military had encuntered a "crisis f legitimacy"'in which it had lst its supprt amng the public. In ther wrds, it was felt that the disaffectin with the war and what it revealed abut the military had generalized int disaffectin fr the military. It is time t'veassess this' assumptin abut what has happened t the public view and assessment f the military. Hward Schuman (1972), in an'insightful analysis f the "Tw Surces f Antiwar Sentiment in America," demnstrated empirically that the grwing dissent cncerning the Vietnam W^r mii'llt lit: nly langiilially related t the'srts f arguments being advanced in'the antiwar mvement especially as expressed n university campuses. He distinguished "mral" ppsitin t the war frm "pragmatic" dissent based n the frustrating expenditure f vast resurces cupled with nly very meager results t be fund in that particular war. Therefre, grwing disaffectin fr Vietnam did nt, fr mst peple mean \ increasing mral rejectin 'f the war, but prbably reflected grwing pragmatic' ppsitin, t a cstly disappintment. AUTHORS' NOTE: Hie data fr this paper were cllected by the Survey Research Center and the Center fr Research n the Utilizatin f Scientific Knwledge f the Institute fr Scial Research, University f Midiigan. Prtins f the research were supprted by the Office f Naval Research and the Army Researcli Institute fr the Behaviral and Scial Sciences. The views expressed herein are thse f the authrs and d nt represent institutinal psitins f the Department f the Navy r the Department f the Army. 215

216 THE SOCIAL PSYCHOLOGY OF MILITARY SERVICE Hence, the generalizatin f antiwar sentiment t antimilitary sentiment fund n campuses and reflected in the destructin f ROTC facilities and/r the remval f R O T C frm full academic participatin might well nt be reflected in many pqbli attitudes tward the military, where antiwar sentiment was based primarily n pragmatic cnsideratins rather than mral nes. In fact, a number f studies at the Institute fr Scial Research (which asked respndents t rate varius grups n a "feeling thermmeter" ranging frm 0.t 100, with 50 as the neutral pint) have fund that the mean level f feeling tward the military has been warmer, r mre psitive, than fr mst ther scietal grups investigated; mrever, this level shwed minimal decline frm 1964 t 1972 (D. Segal, 1975a) and had returned t very near 1964 levels by 1974 (Inglehart and Barnes, 1975). A 1973 ISR study had respndents rate a number f institutins n hw gd a jb they were ding fr the cuntry as a whle. Sme f the findings have been reprted in Rdgers and Jhnstn (1974). The U.S. military, alng with the natin's clleges and universities, headed the list f 15 institutins. Abut 60% rated the military as ding a gd r very gd jb, while nly 10% rated it pr r very pr. In the same study, the peple running the military were given ne f the highest ratings fr being "hnest and mral." And when asked whether the military shuld have mre influence in sciety, r less, ver half the respndents said it shuld be "the same as nw," and the rest were split almst exactly between thse preferring mre military influence and thse preferring less. N ther institutin in the study gt such evenly balanced ratings f influences. The relative gd ratings f the military in the 1973 ISR study were als fund in a recent survey cnducted by Luis Harris and Assciates (1973) fr the Senate Subcmmittee n Intergvernmental Relatins. In that study 40% said that they had "a great deal f cnfidence" in the military. The nly institutins rated higher were televisin news (41%), higher educatinal institutins (44%), and medicine (57%). The several branches f the federal gvernment, majr cmpanies, rganized labr, and even rganized religin all received lwer cnfidence ratings than the military. Hwever, the Harris study als indicated that in 1966 almst all the institutins had substantially higher ratings. In the case f the military, the prprtin wh expressed a great deal f cnfidence fell frm 62% in 1966 t 35% in 1972, and then rse slightly t 40% in 1973. JOHN D. BLAIR and JERALD G. BACHMAN 217 In shrt, it appears that dissatisfactin with the Vietnam War did nt always lead t a general antimilitary psture, althugh the influence f the war cannt be ignred. Tliep: is, hwever, cnsiderable evidence tliiit tliu actllilt Ucnd f trust in gvernment has paralleled that f supprt fr the Vietnam War; i.e., bth have decreased remarkably thrughut the perid frm 1964 t the early 1970s (A. Miller, 1974). This is nt t say that ther factrs such as the civil rights mvement, urban rits, recessinary perids, inflatin, unemplyment, and finally Watergate have nt been equally r perhaps even mre invlved in the decline f trust in gvernment. It is, hwever] t'-say; that pragmatie'disaffectin fr the,]war might.have been directed-less'tward a.military whse hands-were dirtied in an immral war and mre tward the gvernment which was perceived t have been respnsible fr an unwn war. (See Mdigliani, 1972, fr a distinctin between attitudes abut "interventinism" and attitudes dealing with "administratin distrust" fund amng the public during the Krean War.) Anther issue invlved in the assumed generalizatin frm revulsin with the hrrrs f war t antimilitary feelings is the effect f the war n its veterans. The Vietnam Veterans Against the War were elquent spkesmen fr the pint f view, that antiwar sentiment generalized int antimilitary sentiment, but again ne must be careful in extraplating t ther Vietnam era veterans. Tw separate lngitudinal studies which included yung veterans d nt supprt such an extraplatin (Jennings and Markus, 1974; Liachman' and Jennings, 1975). The pint, therefre, is that it is time t reassess public views f the military at the end f the Vietnam era. Data Surces This paper addresses the basic questin f what the public view f "the military" is. The findings are based n survey data cllected frm tw samples: (1) a representative natinal crss-sectin f 1,355 civilians age 16 and lder and (2) a sample f 2,522 Navy persnnel stratified s as t be representative f majr Navy entities (ships and shre statins). Sixteen-page, self-cmpleted questinnaires, identical except fr certain persnal backgrund measures, were administered t bth sample during late 1972 and early 1973. A detailed descriptin f sampling techniques as well as a descriptin

218 THE SOCIAL PSYCHOLOGY OF MILITARY SERVICE JOHN D. BLAIR and JERALD G. BACHMAN 219 f the fit f the samples t their respective ppulatins has been prvided in an earlier reprt (Michaelsen, 1973). This paper extends and elabrates several analyses reprted elsewhere (Bachman, 1973, 1974; Blair, 1975). Interrelatinships Amng Military Values, Preferences, and Perceptins This sectin f the paper is cncerned with the ways in which values, preferences, and perceptins abut military matters are interrelated. We begin by utlining the prcess f data reductin and describing the resulting dimensins. Then we explre sme f the ways in which these dimensins are intercrrelated, lking separately at a number f different analysis grups in the Navy and amng civilians. We review evidence t supprt ur cnclusin that the scales are interrelated in basically similar ways fr the several grups. It shuld be added, hwever, that there are substantial differences in the way that grups are psitined alng sme f the scales; we review sme f these differences later. THE MEASURES OF VALUES, PREFERENCES, AND PERCEPTIONS > I The questinnaire segment dealing with military values, preferences, and perceptins (Sectin C) includes 57 items, designed t measure a cnsiderable number f different, but interrelated, cncepts. An imprtant phase f ur analysis invlved the cnslidatin f these items int a smaller number f indexes in rder t prduce multi-item variables, which are generally mre stable and reliable than single items, and t reduce the cmplexity f the material t a mre manageable level. An early stage in ur effrts tward data reductin invlved a number f factr analyses including nearly all the items in Sectin C f the questinnaire. These analyses cnfirmed mst f ur prir expectatins abut sets f variables t be cmbined int indexes; in a few ther cases, the analyses enabled us t lcate items which did nt meet ur expectatins. These analyses were cnducted separately fr civilians, the Navy sample taken as a whle, and three subgrups within the Navy ^ sample (fficers, first-term enlisted men, later-term enlisted men). The patterns f factrs which emerged frm these several grups were quite similar; thus we felt cnfident that the indexes we were develping were applicable acrss all the grups examined. The measures f values, preferences, and perceptins cncerning military service which resulted frm ur data reductin effrts are shwn in the left hand clumn 1 f the tables and figures, althugh in Table 1, the separate perceived and preferred measures are excluded. As a matter f cnvenience, the measures are rganized in the table accrding t the cnceptual categries fllwed in ur preceding papers. Mst f the measures are indexes based n tw r mre items. Three ne-item measures are included because they are cnceptually imprtant but d nt lend themselves t cmbinatin int indexes. Of the 57 items in Sectin C f ur basic instrument, 42 are included in the 17 measures.! 1] A wrd is in rder cncerning the names given t the measures. An effrt was made t capture the essence f an item r index in relatively few wrds, while at the same time cnveying a gd deal f the meaning. Sme f the measures were better suited t this effrt than thers; in a few cases the names may seem a bit strained. In all cases, the name crrespnds t a high scre n the measure..'.'!!:;! i I ilh,. ; A GENERAL FACTOR.OF PROMILtTARY, SENTIMENT, ' H 'Wj, > Our earlier explratins f the data, and sme examinatin fthe crrelatin matrices described abve, led us t feel that there is a "general factr" f prmilitary (r antimilitary) sentiment underlying mst f the measures that we have been discussing. In an effrt t test this ntin we perfrmed a set f factr analyses. As a first step, prduct-mment crrelatins were cmputed amng all f the measures in Table 1. The crrelatins were cmputed fr each f the fllwing analysis grups separately:, Navy first-term enlisted men Navy later-term enlisted men Navy fficers Civilian men Civilian wmen The cmplete crrelatin matrices are presented in Bachman (1974).

"ijohn D.'BLAIR and JERALD G. BACHMAN 221 CC H O < IL < CC UJ 2 ai O < 2 O W z < r C Lfl CD O I» c 2 M CO ^ II O 2 E S- is * - 2 c UJ E m 5 O M IN r c r» r- ^ *t c in tn t (N Cl N O O) ti ifl n tn r- r- 0> r-> CN cn (O r-. ID "3- m m n ID i- c t r- in in c f- CN 00 CN T O r- t- (D lfl IN a.q «0 O " Z * n v. XI TJ 9- Q- 00 Ol LT) r- tf O) r- «- t in in l' \ ^ r- (N r- CM *J CN r- in t 1 I r-- «O O in r h n H) r- in to (0 t m CO 00 T (N N CO CO CN t in cf 00 CN m r- CN t in m a CT c > ra c cn «C» ^ 1 a j; m O > cd? ra a a> 5 e < v atw» II u ^ C ra.2 5 S. a) C T3.Ol _CC < t t CN CN r t c c CN in CD f N N 0) CN CO «- t~ i to ID (Ti cn i CO N to -3- [*- U) i j I- i' in c c T to CN 5 5 8 a > CO «5 2 = => 1 0 QJ «u c 2 Q.HJ & CO D- r- r c m UD ^ CN CM M D rs t m v >- CN i r r r r S c - j.c If r*. c >- tn m >- «t r c in t in m *- r r r r \ r r c ai c r- i *- CN CN CN CN O r- m r rr (-) LO i in r- CN *r r- c in CN r- c t in c c CN I I 1 I in CN 3 * in c in c m CN c in in m r r r f r Xia> a E > 3.si CT CD r E C ai cn ra c a \ E ui ra > V) r 3 03 C SI D O j O O S U a. a a a in O O H r- I 1 >, In this series f factr analyses we were lking fr the largest and mst general single factr underlying the military value, preference, and perceptin measures. Accrdingly, we used the principal cmpnents methd and fcused attentin n the first factr (unrtated). The factr ladings fr each f the five analysis grups are displayed, in Table 1., The results shwn in Table 1 clearly cnfirm ur view that there is a rather substantial general factr f "prmilitary sentiment" which cntributes t ur measures f military views. It accunts fr r "explains" between 23% and 30% f, the variance in these measures fr Navy enlisted men and civilians. It accunts fr 36% f'the variance fr Navy fficcru, and [\\\\ fru'tik ladings tr this grup tend t be smewhat higher than is true fr ther grups. This appears t be a reflectin f educatin, fr we have shwn similar factr ladings fr civilian cllege graduates with 40% f the variance explained. There is a (l cnsiderable degree j f similarity in the patterns f factr ladings fr all fivej analysis grups. Withut exceptin, the directin f lading is the-same fr/alllanalysis grpups^le., a measure is either psitively laded 'fr- all grups r negatively fr all. Mrever, thse measures which, lad mst strngly are the same acrss all grups.. _, Let us cnsider what it meant (in early 1973) t be high in ur general factr, f prmilitary sentiment. Nt surprisingly, thse highest in prmilitary sentiment rated ur military leaders as quite cmpetent, gave the military services high markes fr jb pprtunity and fair treatment," stated a preference fr higher levels f military spending and influence, and saw the rle f the military in sciety as predminantly psitive. Their freign plicy views were rather "hawkish" they were relatively supprtive f U.S. military interventin in ther cuntries, they preferred a psitin f military supremacy (rather than parity with the U.S.S.R.), and they were mst likely t supprt past U.S. invlvement in Vietnam. Finally, they placed a.high value n- bedience t military authrity they tended t agree that "servicemen shuld bey rders withut questin," and sme maintained this psitin even when faced with a My Lai-type incident., Tw measures which shw little assciatin with the general factr f military sentiment are the dimensins mst clsely linked t the debate abut the draft versus the all-vlunteer frce; these measures 1 220

222 THE SOCIAL PSYCHOLOGY OF MILITARY SERVICE JOHN D. BLAIR and JERALD G. BACHMAN 223 are Preference fr Citizen-Sldiers (versus "Career Men") ( and Preference fr Wide Range f Plitical Views amng Servicemen. These tw dimensins seem t stand smewhat apart frm mst f the ther measures and are less integrated int an verall prmilitary r antimilitary cntinuum. Our basic cnclusin is that ur respndents shwed little "plarizatin" alng these dimensins perhaps indicating that mst peple had nt given much thught t the issues they represent. ), Amng all the dimensins summarized abve, the measure f Vietnam Dissent has a particularly strng lading n the general factr f military sentiment. One pssible inleiprehilin fr this relatinship is that thse generally supprtive f the military establishment were, as a result, least critical f ur past invlvement in Vietnam. In ther wrds, Vietnam views were shaped by brader attitudes abut the military. An alternative interpretatin is that views abut the Vietnam invlvement were generalized t the larger military establishment, s that negative feelings abut Vietnam led t negative views abut military spending, influence, leadership, and the like. These tw interpretatins are nt mutually exclusive indeed, it is likely that bth patterns f causatin were at wrk. Returning t the basic questin f the nature f the public view f the military, we cnclude that Vietnam sentiment is, in fact, clsely related t the verall view f "the military." Mre generally, we have fund that mst f ur items distribute respndents alng a prmilitary-antimilitary cntinuum; i.e., individuals are similarly rdered frm ne item t anther. But des this cnsistent set f relatinships amng an individual's attitudes mean jthat the anti- Vietnam War sentiment reprted in the plls has "generalized" l ther aspects f the military? It is t an examinatin f the distributins f the different items that we turn next fr an answer t this questin. The Public View f the Military,, < The first prtin f this paper stressed the intercrrelatins amng ur measures f military views a pattern indicating that individuals wh were mst favrable tward the military alng ne dimensin tended t be relatively favrable alng ther dimensins. But this des nt mean that fr a typical individual r fr the public in ] general views abut the military were cnsistently psitive r cnsistently negative. T the, cntrary, there were substantial differences in verall ratings f the several aspects f the military that we explred; alng sme dimensins the dminant public view was quite favrable, while alng thers it was nt. In this sectin f the paper we will lk first at mean scres n the varius measures fr all civilians'in ur'sample the public as a whle. Then we will.turn t sme variatin in the respnses t the measures resulting frm differences in age, and, educatin. Finally, we will examine the effect f persnal experience with Die military n the values, perceptins, and preferences f civilian men. THE PUBLIC AS A WHOLE. - i. '.» ; In Figure 1,' ( we have summarized the basic findings fr the ttal civilian sample. The. mean scre n each measure is reprted in terms f its relatinship t the substantive midpint.[2] This relatinship is determined bth by the differences f the sample mean frm the midpint and by the standard deviatin f the ttal civilian sample n that item. Hence if a mean scre were +.25 scale pints different frm the midpint and the standard deviatin were.50, then n the figure, the bar t wuld extend t the first line (1/2 S.D.) n the "prmilitary" side. The end f the bar represents the lcatin f the mean scre with respect t the substantive midpint r divisin between prmilitary and antimilitary, respnses expressed as prprtins f the standard deviatin fr the sample as a whle. This rather cmplex prcedure was used because the measures differ cnsiderably in terms f the lengths f the' scales used as well as the variance frm ne scale t anther, and we j wanted t increase the 1 cmparability acrss scales. i _, t In Figure 1 w,e (find that althugh the items are highly interrelated and display,,a cmmn underlying factr!,their distributins d nt ' demnstrate' that in early 19731.theipublic was(primarily either pr- r antimilitary. The public as a'whle evaluated different aspects f the military quite differently, even thugh each measure may rder individuals similarly. Our analysis leads us t believe that there are three fundamentally different aspects f the military that are evaluated by the public and reflected by these findings. The first f these deals with the military rganizatin itself. The scales labeled "the military wrk rle" and ' I

224 THE SOCIAL PSYCHOLOGY OF MILITARY SERVICE JOHN D. BLAIR and JERALD G. BACHMAN 225 THE MILITARY WORK ROLE Perceived Military Jb Opprtunities "Anti-Military' Perceived Fair Treatment in Services LQIlljllllHU Perceived Discriminatin Against Wmen nd Blacks MILITARY LEADERSHIP Perceived Cmpetence f Military Leaders MILITARY INFLUENCE OVER NATIONAL Preference fr Higher Military Spending Rle f Militry in Sciety Perceived s Perceived Military (Versus Civilian) POLICY Negative Influence Preferred Military (Versus Civilian)Influence Adequacy f Military Influence (Perc. Minus FOREIGN POLICY ANO MILITARY Supprt fr Military Interventin Preference fr U.S. Military Supremacy Vietnam Dissent POWER Pref.) LO LO mum HI 'Pr-Military 1 llllllll! TTTTULO EI nihi HI I m rrrrrm 1 LO ISSUES INVOLVED IN AN ALL-VOLUNTEER FORCE imim LO Supprt fr Amnesty Oppsitin t Unquestining Military Obedience imn LO Oppsitin t Obedience in My Lai-type Incident HI HE Preference fr "Citizen Sldiers' (Vs. "Career Men") HI &j Preference fr Wide Range f Views Amng Servicemen - HifTM - 1/2 S.D. MID 1/2 S.U. I7TTTTT1 All Civilians ; Center line shws midpint f each measure; ther lines shw 1/2 S.D. fr all civilians n each measure. Firjure V. MEAN SCORES OF ALL CIVILIANS CONTRASTED WITH MIDPOINT OF EACH MEASURE IH, llo HI HI "military leadership" and the item dealing with "unquestining military bedience" all deal with dimensins f the military as an rganizatin: hw cmpetently) it is managed, what it prvides, hw its members are treated, and what can be expected f the members. Generally this aspect f the military is psitively evaluated. On the ther hand, its members are nt seen as having much recurse if unfairly treated-especially inj cmparisn t that enjyed by civilians. ( j. j ( - i <l - I!' 1 i<l, A '' j The secnd. ^undamentaljaspect^fj the, military ;deals with the use f military frce. The measures fund'"under; the label, "freign plicy and military pwer" as well as' the ne asking abut "bedience in a My Lai-type,incident" reflect attitudes abut the apprpriate cnditins fr ging t war, the necessity f maintaining military superirity, and the utcme f the Vietnam War. The findings suggest that.the mass public is reluctant t g t war except t prtect the United States frm actual attack, but that it will supprt effrts t deter such an attack thrugh a high level f military superirity., Our, measure f Vietnam dissent did nt shw in early 1973 the verwhelming rejectin f the war generally reprted in pll findings. We feel that this is a reflectin f ur nt just asking respndents the cmmn questin f whether the war was a mistake, but asking instead a series f quefitiiih Hull uuilluln blh (, ''mral" and "pragmatic" cnsideratins and that cmbine issues f interventin and natinal defense. Unfrtunately these measures were develped lng befre Schuman's seminal wrk appeared (1972), and the mral and pragmatic aspects f sentiment tward the war were nt clearly delineated in ur questins. The end result f the cmbinatin fthe varius items is t prvide a measure that indicates ambiguity in the public view f v the Vietnam [War (Bachman, 1 1973). j The 1 third fundamental aspect f the military, cncerns civilmilitary relatins. The measures-we see as reflecting values, perceptins, and preferences in this area are fund under the heading "military influence, ver natinal plicy" and the three items nt yet examined under "issues invlved in, all-vlunteer frce." Civilmilitary relatins invlve an interchange between the civilian and military sectrs f sciety r, expressed smewhat differently, between sciety as a whle and the military rganizatin. Hw is the military's relatinship t sciety evaluated? Our data reveal that the rle fthe military in sciety is seen as quite psitive.

226 THE SOCIAL PSYCHOLOGY OF MILITARY SERVICE JOHN D. BLAIR and JERALD G. BACHMAN 227 In additin, althugh tlie military is seen us having cqtisk)t;r! i t> influence vis-a-vis civilians in areas dealing with natinal security and internal rganizatinal cncerns, the public prefers it that way, and hence the influence f the military is seen as generally adequate neither excessive nr inadequate. These findings cme frm data cllected abut 6 mnths earlier than thse that prvided the basic findings reprted in the Washingtn Pst by Chapman (1974) and later by Rdgers and Jhnstn (1974). The summary f the data prvided by the Washingtn Pst's headline, "Military Mst Admired US Institutin," certainly verstates the case, but ur findings supprt the basic thrust f Rdgers and Jhnstn's findings, which are that the military rganizatin and the rle it plays in sciety are still psitively evaluated by the "average persn." Althugh ur data cannt speak directly t the questin, the cnsistently high levels that the military received n the thermmeter scres (reprted abve) thrughut the entire Vietnam era lead us t believe that the findings cming frm these tw studies are nt a reflectin f a change in the public image f the military after the war, but indicate a cntinuing high level f supprt fr the military rganizatin. The ther side f the cin in civil-military relatins is what sciety des fr the military. Here the picture is smewhat mixed in terms f public attitudes: There is a cnsiderable reluctance t prvide mre mney fr the military.[3] In additin, there seems t be sme reluctance abut having nly career sldiers r limited plitical views within the military. These last tw items, as nted abve, are imprtant issues in an all-vlunteer frce but d nt seem t be highly tied t ther ideas abut the military r t reflect very well-defined attitudes n the part f mst peple. The last item dealing with civil-military relatins is the index dealing with amnesty. Here we wuld like, primarily, t interpret its cnsequences fr the view f the relatinship f sciety t the military. The rejectin f amnesty in the findings seems -t us t indicate the basic legitimacy f military service even in an agnizing and divisive war. It must be remembered that ur measure f attitudes tward the war itself revealed cnsiderable belief in its imprtance fr natinal security, althugh it was seen as harmful t the cuntry generally. In summary, three pints need t be emphasized: 1. The distributins f the respnses t the varius measures we have used reveal thai, althugh Bhiilliir rdering f individuals may exist acrss the measures, there remained cnsiderable differences in the values, perceptins, and preferences f the public as a whle cncerning different aspects f "the military." 2. At this aggregate level, civilians did nt appear entirely r even primarily pr- r antimilitary. There was a generally favrable evaluatin f the perfrmance f the military rganizatin and its relatinship t sciety, but a dissatisfactin with the, present high levels f financial supprt frj the j military audita'reluctance t supprt the use f military frcejexcept in self-defense. 3. The differences in attitudes cncerning the military rganizatin and thse cncerning the use f the frce made pssible by that rganizatin led t a clarificatin f apparently discrepant prir research findings. Sme f these studies indicated that the public rejected Vietnam plicy and acts f atrcities, but thers shwed that the public psitively evaluated military leaders and the jb dne by the military fr the cuntry. VARIATION BY AGE AND EDUCATION Thus far we have lked at the distributin f respndents' values, perceptins, and preferences nly at the aggregate level. During the Vietnam era, Americans became sensitized t the cunterculture expressed amng yuth and t the quite different attitudes displayed by hard hats. These reactins t the Vietnam War and t "the military" suggest that aggregate findings d nt reflect adequately sme very imprtant differences in the public view f the military. There are ther pssible bases f cleavage within the civilian ppulatin that may als be imprtant in examining these issues. Hwever, ur ther wrk with these data reveals that the primary bases f cleavage are age and educatin. Fr simplicity in presentatin and cmparisn we have dichtmized age int "yunger" and "lder," defined as 34 r yunger and 35 r lder. In additin, we have separated ur sample int cllege graduates and nncllege graduates. Our ther wrk cited abve has shwn that these tw backgrund variables have generally cnsistent, linear relatinships t the attitudinal measures that we have been examining here. As a result, little is lst in ur understanding f these relatinships by using dichtmies. The findings fr the fur grups that result frm cmbining these 1 I

228 THE SOCIAL PSYCHOLOGY OF MILITARY SERVICE JOHN D. BLAIR and JERALD G. BACHMAN 229 THE MILITARY WORK ROLE Perceived Military Jb Opprtunities Perceived Fair Treatment in Services Perceived Discriminatin Against Vimen and Blacks MILITARY LEADERSHIP Perceived Cmpetence f Military Leders MILITARY INFLUENCE OVER NATIONAL POLICY Preference fr Higher Military Spending < ^ Rle f Military in Sciety Perceived as Negative Perceived HiWtery (Versus Civilian) Influence Preferred Military (Versus Civilian) Influence Adequacy f Military Influence (Perc. Minus Pref.) FOREIGN POLICY AND MILITARY POWER Supprt fr Military Interventin Preference fr U.S. Military Supremacy Vietnam Dissent ISSUES INVOLVED (N AN ALL-VOLUNTEER FORCE Supprt fr Amnesty Oppsitin t Unquestining Military Obedience Oppsitin t Obedience in My Lai-type incident Preference fr "Citizen Sldiers' (Vs. "Career Hen") Preference fr Wide Range f Views Amng Servicemen Yunger (<34) Older (> 35) Cl lege Grds Cl lege Grads Older(>35) Anti-Military" /Pr-Military" Nn-Grds 1/2S.D.MID I/2S.D. Yunger (<34) Nn-Grads Center line shws midpint n each measure: ther lines shw 1/2 S.D. fr il civilians n each measure. Figure 2: MEAN SCORES OF CIVILIAN GROUPS CONTRASTED WITH MIDPOINT OF EACH MEASURE tw dichtmies are shwn in Figure 2. The same measurement and data manipulatin prcedures are used as in Figure 1. Because the variance frm ne grup t the next n each item is very similar, we have again used the standard deviatin f the civilian sample as a whle, and.the results are reprted in prprtins f standard deviatin units.[4] It shuld be particularly nted that yunger cllege graduates shwed a fairly high level f "antimilitary" respnses. This is cnsistent with Schuman's finding (1972) that mral, as ppsed t pragmatic, ppsitin t the Vietnam War as discvered in his sample was fund primarily amng yung cllege graduates in tlie mass public. Hwever, even here the majr distinctin between the military rganizatin and the use f military frce applies. The yung graduates were cnsistently mre ppsed t all frms f using military frce than they were t the military rganizatin, with the exceptin f the measure f fairness within the military services. They als evaluated civil-military relatins in a very negative light. They wanted cnsiderably reduced military spending, perceived a very influential military.whse influence is excessive, and were cncerned abut all the issuesjinvjlved in an all-vlunteer frce. The lder nngraduates shwed, the \ mst cnsistently "prmilitary" views. But their cnsistency was less than that f the yunger cllege graduates. They cnsistently perceived the military rganizatin as a psitive entity (except, again, fr fair treatment). When it came t the use f military frce, hwever, they were als reluctant t supprt any actin nt related t natinal defense. In fact, their scres n the supprt fr military interventin index equal thse f the yunger cllege graduates and prbably reflect tw very different types f islatinism. In terms f civil-military relatins, the lder nngraduates saw the military as making an imprtant cntributin t sciety and were less reluctant t prvide mney and t supprt greater than present military influence. Nr did they seem greatly cncerned abut the kinds f issues raised by Ihe ;i!i-vhmteer frce. The ther tw grups fall in intermediate categries in terms f pr- r antimilitary attitudinal cnsistency. Sme aspects f the military were mre likely t be called int questin by lder cllege graduates, and ther aspects were mre critically evaluated by the yunger nngraduates. In sum, there were, in early 1973, imprtant variatins within the

230 THE SOCIAL PSYCHOLOGY OF MILITARY SERVICE, JOHN D. BLAIR and JERALD G. BACHMAN 231 Table 2. IMPACT OF PAST MILITARY SERVICE ON ATTITUDES OF CIVILIAN Crrelatins* With: MEN B. Psitive A. Feelings Past Military Abut Having Service**] Served***. i (All civilian (Veterans men-n - 719) nly-n = 349) The Military Wrk Rle Perceived military jb pprtunities.00.iu Perceived fair treatment in services.02.33 Perceived discriminatin against wmen and blacks -.06 -.22 Military Leadership Perceived cmpetence f military leaders.02.31 Military Influence Over Natinal Plicy Preference fr higher military spending and influence.07.33 Rle f military in sciety perceived as negative -.05 -.34 Perceived military (versus civilian} influence -.23 -.17 Preferred military (versus civilian) influence.07.25 Adequacy f military influence (perc. minus pref.) -.23 -.30 Freign Plicy and Military Pwer Supprt fr military influence.10.18 Preference fr U.S. military supremacy.07.18 Vietnam dissent -.07 -.32 Issues Invlved in an All-Vlunteer Frce Supprt fr amnesty -.19 -.36 Oppsitin t unquestining military bedience -.04 -.35 Oppsitin t bedience in My Lai-type incident -.09 -.29 Preference fr "citizen-sldiers" (vs. "career-men") -.16 i -.08 Preference fr wide range f views amng servicemen.06 -.10 'The crrelatins are prduct-mment; hwever, the "past military service" variable is a dichtmy, and thus the crrelatins in Clumn A are als termed pint-biserial. (The pint-biserial is a special case f the prduct-mment crrelatin; in the present instance it can be interpreted in essentially the same wbv as the mre typical prduct-mment crrelatin invlving cntinuus distributins alng bth dimensins see Nunnally, 1967, pp. 120-133.) Crrelatins in Clumn A may be cnsidered statistically significant if tlmy exceed.11, thse in Clumn B if they exceed.16 (p < randm sample)..001, tw-tailed, unsumlnij a simple "Past Military Service is based n respnses t Questin D14, but scring has been reversed s that a psitive crrelatin means that veterans have higher scres n the dimensin than d nnveterans. ""Psitive Feelings Abut Having Served is based n respnses t Questin D20, with scring reversed s that a psitive crrelatin means that the dimensin is assciated with psitive feelings abut have served. public values, perceptins, and preferences cncerning the military. The mst imprtant f these were linked t the age and educatin f the respndents. At the aggregate level, the yunger cllege graduates exhibited the greatest cnsistency f attitudes. This may well reflect the generalizatin f cncerns with the Vietnam War t all things military, althugh even in this grup the military rganizatin fared better than did the use f military frce r the state f civil-military relatins. i* \ VETERANS' VrEWS OF THE MILITARY There are a number,f,ways f frming attitudes abut the military. But t firsthand persnal cntacts may have a much greater impact and credibility. Thus it seemed useful t explre the degree t which the varius perceptins land attitudes abut the military and its missin are,linked t suchipersnal.cntacts.. \., In this sectin we will explre firsthand military; cntacts amng the civilian men in ur sample., First, we will see t what extent civilian men wh had served in the armed frces differed frm thse wh had nt. Secnd, we will explre the extent t which psitive r negative feelings abut past military service were related t military attitudes and perceptins. Is the average veteran mre supprtive f the military than the average nnveteran? Our data, summarized in clumn A f Table 2, indicate few differences between the average veteran and nnveteran alng ur dimensins, and thse differences which d appear are nt very large. When asked hw they wuld feel abut a sn's enlistment, 29% f the veterans checked "strngly psitive," cmpared with 23% f the nnveterans; thse respnding "mstly psitive" were 45% and 46%, respectively. Differences this small are neither statistically trustwrthy nr substantively imprtant (they crrespnd t a pint-biserial crrelatin f.08). Reviewing the figures in clumn A f Table 2, we find n average differences between veterans and nnveterans in their perceptins f the military wrk rle r the cmpetence f military leaders, and n substantial differences in, their _ views abut freign plicy and military pwer. In ratings f ideal levels f military versus civilian influence, there was little difference between the tw grups; bth veterans and nnveterans preferred a rughly equal sharing f influence by military and civilian leaders. On the ther hand, there

232 THE SOCIAL PSYCHOLOGY OF MILITARY SERVICE JOHN D. BLAIR and JERALD G. BACHMAN 233 was a difference in perceptins f actual levels f military influence; veterans perceived the military as smewhat less influential than the nnveterans. As a result, the tw grups differed alng ur dimensin f adequacy f military influence; veterans preferred a bit mre military influence than they felt was actually the case, whereas nnveterans preferred slightly less. ( ' ( It is interesting t nte that the veterans' scres alng the military versus civilian influence dimensins were quite similar t thse f first-term enlisted men in the Navy. (See als in this vlume Bachman and Blair's " 'Citizen Frce' r 'Career Frce'? Implicatins fr Idelgy in the All-Vlunteer Frce.") It may be that ne f the mst cnsistent results f past r present- experience in military service is a lwered assessment f the amunt f influence that military leaders actually have ver a range i f decisins affecting natinal security. Veterans and nnveterans differed little in their evaluatins f past U.S. actins in Vietnam; bth grups were split nearly equally between thse wh tended tward supprt and thse wh tended t be critical. In their feelings abut amnesty, n the ther hand, the grups differed nticeably. While 36% f the nnveterans agreed r agreed mstly that the men wh went t Canada t avid fighting in Vietnam were ding what they felt was right and shuld nt be punished, nly 18% f the veterans agreed r agreed mstly. In answer t the cmparisn questin stating that ging t Canada was wrng and thse wh did s shuld be punished, 78% f the veterans agreed r agreed mstly, cmpared with 61% f the nnveterans. It is perhaps understandable that mst civilians wh nce served in the armed frces themselves wuld have little tlerance fr thse wli avided service by ging t Canada. One ther difference is wrth nting between veterans and nnveterans. The veterans shwed greater supprt fr the ^dea f "career men" in the military rather than "citizen sldiers," whereas the ppsite was true fr nnveterans. The data are summarized in Table 3. Table 4 displays tw questins which deal with veterans' reactins t their military service and their perceptins f family reactins. Mst veterans reprted their wn feelings as psitive; nevertheless, there was rm fr variatin. This variatin in feeling abut ne's wn military experience is strngly assciated with' ther attitudes abut military matters. Table 3. PREFERENCE FOR "CITIZEN-SOLDIERS" VERSUS "CAREER MEN" Vatarans Nnveterans Scres n the Index ' ' (N = 324) (N = 386) 1.0-2.0 Prefer "career men" 46% 31% 2.5,Mixed feelings 24 26 3.0-4.0 Prefer "citizen-sldiers"- 30 43 Ttal 100% 100% Perceptins abut family reactins, shwn in the lwer part f.table 4, were mre balanced between psitive and negative views. The tw items shwn in Table 4 were nly mdestly crrelated (r = ".23); this finding, cupled with the different levels f,psitive feeling shwn in the tw items, indicates that sme veterans held psitive feelings abut their military experience in spite f a perceptin that their entry int the service was nt especially favred by family members. Veterans' wn feelings ; abut having served are strngly crrelated with feelings, abut the pssibility f a sn's enlistment (r =.53). These feelings abut having served are als related t a number f ther dimensins, as shwn in clumn B f Table 2. The pattern f crrelatins is similar t the factr ladings shwn' in Table 1; thse items which lare mst psitively, r negatively assciated with a general factr f prmilitary sentiment are als mst strngly linked t veterans' feelings abut their wn military experience. In clumn B f Table 2, veterans' psitive feelings abut their military service are crrelated with perceptins that the military rganizatin ffers fair' treatment and cmpetent leadership. i i i it 1 1 ' i i 'i i' Table 4. VETERANS'FEELINGS ABOUT HAVING V SERVED '.1 V, j *. Questin D20- Wuld yu say yur feelings abut having been in the military are: (1) Strngly psitive ' 38% (2) Mstly psitive 43 (3) Mstly negative 14 (4) Strngly negative 5 100% Questin D21. Which f the fllwing best describes the feelings f yur family when they first learned yu were ging t enter the service? (11 They were very much in favr f it 20% (2) Smewhat in favr 26 (3) Neutral r indifferent 25 (4) Smewhat dissatisfied 22 '(5) Very much dissatisfied 8 101%

234 THE SOCIAL PSYCHOLOGY OF MILITARY SERVICE Veterans with psitive feelings abut their past service were -alslikely t prefer fairly high levels f military influence, view the rle f the military in ur sciety as being predminantly psitive, supprt past U.S. actins in Vietnam, shw abve average ppsitin t amnesty, and state that servicemen shuld bey rders withut questin. We explred several ther dimensins f veterans' experiences t see if they crrelated with ur measures f military attitudes. N substantial r cnsistent differences were fund t be assciated with branch f service (Army, Navy, Air Frce, Marines). There als appeared t be n clear differences in attitudes between thse veterans wh had been drafted and thse wh had enlisted althugh mst f the veterans wh.had enlisted thught they wuld have been drafted therwise. One dimensin which did shw sme cnsistent differences is length f past military service. Mst f ur civilian respndents wh served in the military remained in the service n mre than 4 years. But thse wh had served lnger tended t shw attitudes mre favrable t the military. In particular, the. greater the length f service amng these veterans, Ihe higher their ratings f fair treatment and cmpetent leadership in tlie military. Thusu firnlillgfl are prbalby a reflectin f career rientatin amng the lngerserving veterans. In summary, the effect f persnal cntact with the military was nt fund t be as strng as is smetimes suppsed. This might well indicate the differences between studies f large and representative samples as ppsed t thse fcusing n case studies f individuals r rganized grups such as the VFW r the VVAW (see Helmer, 1974). There was, f curse, enrmus variatin in attitudes within the grup f veterans as was true fr the nnveterans as well. Sme f this variatin amng veterans can be explained in terms f differing evaluatins f their wn military experience. Hwever, a recent study f persnal cntact with gvernmental agencies (see Katz et al., 1974) fund that nly negatively evaluated cntact has any effect different frm n cntact at all. Similarly, we fund cnsistent relatinships between the evaluatin f military experience and military attitudes but little relatinship between veteran status and attitudes. Mst veterans were psitive in their evaluatin, as shwn in Table 4; s it appears that thse with negative cntact were negative in their views, but thse with psitive cntact did nt have greatly JOHN D. BLAIR and JERALD G. BACHMAN 235 i different scres frm thse with n cntact. Had the evaluatins f past military service been mre evenly distributed, the lack f differences in mean scres between veterans and nnveterans culd have resulted frm an averaging prcess, but that is nt the case here. Thse wh had negative experiences als had negative attitudes; thse wh had psitive experiences lked like thse with nne at all i.e., like nnveterans.) j : tj ' 1 1 i \! i i Summary and Cnclusins In this lk at the public view f the military, we have stressed several different themes.. First, we nted a fairly strng pattern f intercrrelatins amng the different dimensins alng which peple rated the military and its missin. In general, we fund that thse wh were mst favrable tward the military alng ne dimensin tended t be amng the mre favrable r less critical alng ther dimensins. The secnd finding that the military and its missin received "mixed reviews" at the hands f a nallfminl ttuss-iuicliun f civilians muy help shed light n sme apparently discrepant prir research findings. Several studies fund that the public rejected Vietnam plicy t an increasing degree during the late sixties and early seventies, but ther research indicated that the public fairly cnsistently gave high ratings t military leaders and t the jb that the military was ding fr the cuntry. We have argued that these tw kinds f findings are nt incmpatible, particularly if the gvernment rather than the military leadership was blamed fr getting the U.S. invlved in Vietnam. At the same time, it must be acknwledged that thse mst strngly critical f Vietnam plicy als tended t be the least supprtive f the military alng ther dimensins, thus suggesting that fr sme individuals a frustratin with Vietnam' plicy may have led t a heightened dissatisfactin with the military as a whle. Evaluatins f the military and its missin vary smewhat acrss different grups. Yunger cllege graduates as a grup nt nly called the use f military frce int questin but were als critical f virtually all aspects f the military rganizatin and existing civil-military relatins. On the ther hand, lder nngraduates as a grup shwed predminantly prmilitary attitudes, but were still

236 THE SOCIAL PSYCHOLOGY OF MILITARY SERVICE critical f the use f military frce fr interventin, high levels f military spending, and the fairness f treatment within the military. The ther tw grups examined lder cllege graduates and yunger nngraduates exhibited intermediate levels f pr- r antimilitary attitudes. The lder graduates were mre likely t call sme aspects f the military int questin, and the yunger nngraduates t evaluate thers critically. When we cmpared veterans and nnveterans, we fund few differences in attitudes abut the military. The veterans were, hwever, mre likely t rate the military as less influential, were mre strngly ppsed t amnesty, and were mre supprtive f a "career military" than were nnveterans. Within the grup f veterans there were variatins in ratings f the military: thse wh were dissatisfied with their wn service experience tended t be less favrable t the military as a whle. Ntes ' 1. The 17 measures cntain ne instance f redundancy. The indexes f perceived Military Influence and Preferred Military Influence are ingredients fr a single discrepancy measure (Perceived minus Preferred) which indicates the extent t which a respndent thinks the level f actual military influence exceeds, r falls shrt f, what he wuld cnsider ideal. In ur factr analyses the separate Perceived and Preferred measures are excluded, thus leaving a set f 15 measures in which each item appears n mre than nce (Table 1). 2. The substantive midpint is the pint n the scale which divides respnses int thse that are favrable r "prmilitary" and thse that are unfavrable r "antimilitary." The classificatins used are cnsistent with the results f the factr analyses reprted. 3. Blair (1975) has separated the item (C25) dealing with military spending frm the index cmbining preference fr higher military spending and influence. This single item is the measure we have used in Figures 1-2. The tw items used in the index are very highly crrelated. 4. Generally, a difference between the tw nngraduate grups f mre than.20 standard deviatin, between a nngraduate grup and a graduate grup f mre than.25 standard deviatin, and between the tw graduate grups f mre than.30 standard deviatin is bth substantively and statistically significant. Als, the lcatin f the mean f either nngraduate grup mte than.10 standard deviatin (and mre than.20 standard deviatin fr either cllege graduate grup) frm the midpint means that the distributin jj nt evenly divided but leans significantly in the pr- r antimilitary directin.