Towards a U.S. Army Officer Corps Strategy for Success: Retaining Talent

8/6/2019 Towards a U.S. Army Officer Corps Strategy for Success: Retaining Talent

1/71


2/71


3/71

iii

FOREWORD

In Towards a U.S. Army Ofcer Corps Strategy forSuccess: Retaining Talent, Colonel Casey Wardynski,Major David S. Lyle, and Lieutenant Colonel (Ret.)Michael J. Colarusso continue their examination ofhow the U.S. Army accesses, develops, retains, and em-ploys ofcer talent. In this third of six monographs,the authors focus upon the signicant decline in juniorofcer retention rates since the 1980s and the long-termimplications for the Ofcer Corps. More importantly,they identify failed responses to the challenge,provide a theoretical framework upon which to buildsuccessful talent retention programs, and make specicrecommendations for restoring rates to previouslyhealthy levels.

As the ight of talented young ofcers engenders

signicant risk to both the Army and to U.S. nationalsecurity, studies of this kind are critical to the creationof a successful Ofcer Corps strategy.

DOUGLAS C. LOVELACE, JR.DirectorStrategic Studies Institute


4/71

iv

ABOUT THE AUTHORS

CASEY WARDYNSKI is Director of the Ofce ofEconomic and Manpower Analysis and an AssociateProfessor of Economics at the United States MilitaryAcademy, West Point, NY. In addition to creatingthe concepts for the Americas Army Game andthe Armys pre-commissioning retention incentives,Colonel Wardynski has published in the area ofmilitary compensation policy and manpower. ColonelWardynski earned a B.S. at West Point, a Masters inPublic Policy at Harvard, and a Ph.D. in Policy Analysisfrom the Rand Graduate School.

DAVID S. LYLE is an Associate Professor of Economicsand Deputy Director of the Ofce of Economic andManpower Analysis at the United States Military

Academy, West Point, NY. He has publications inthe Journal of Political Economy, the Journal of LaborEconomics, the Review of Economics and Statistics,the American Economic Journal: Applied, and theEconomics of Education Review. Major Lyle earned aB.S. at West Point and a Ph.D. in Economics from theMassachusetts Institute of Technology.

MICHAEL J. COLARUSSO is a research analyst inthe Ofce of Economic and Manpower Analysis atthe United States Military Academy, West Point, NY.He is a retired U.S. Army lieutenant colonel and hasserved in a variety of military positions, to include asan Assistant Professor of History at West Point. Mr.Colarusso earned a B.A. in History from Saint Johns

University and a M.A. in History from the PennsylvaniaState University.


5/71

v

SUMMARY

The U.S. Army has made signicant investmentsin its future, especially in its leadership. In particular,the Army has devoted billions of dollars to ofcerundergraduate-level education, world class training,and developmental experiences. Since the late 1980s,however, prospects for the Ofcer Corps future havebeen darkened by an ever-diminishing return on thisinvestment, as evidenced by plummeting company-grade ofcer retention rates. Signicantly, this leakageincludes a large share of high-performing ofcers,many of them developed via a fully-funded under-graduate education.

In the last few years, the Army has responded to thischallenge with unprecedented retention incentives, toinclude broadly offered cash payments. The objective

has been to retain as many junior ofcers on activeduty as possible. However, such quantity-focusedincentive programs run counter to a talent-focusedOfcer Corps strategy. The objective should not bemerely to retain all ofcers, but to retain talentedofcers while simultaneously culling out those lackingdistributions of skills, knowledge, and behaviors indemand across the force.

Retaining sufcient rather than optimally perform-ing ofcers may have dire consequences for the Armysfuture. New ofcer cohorts of high-potential talentmay be driven away by the prospects of serving underlackluster leadership, while those continuing theirservice may experience stunted development due to adearth of talented mentors.

Low junior-ofcer retention increases risks to thewell-being and capabilities of the Ofcer Corps in otherways as well. It strips away the Armys ability to screen,


6/71

vi

vet, and cull for talent, forcing it instead to over-access,increase promotion rates and compress promotion

timing. It degrades the developmental experiencesof junior ofcers and undercuts the Armys ability todiscern which ofcers possess the talent it needs. Leftunchecked, such developments could signicantlyundermine the Ofcer Corps performance levels,taking perhaps a generation to rectify.

Given that the Army is competing in the Americanlabor market for its ofcers, its retention strategymust be built upon sound theoretical concepts. Itmust focus upon talent, guard against systematicdecisionmaking errors, redress market failures, andcreate an employment climate that powerfully meetsthe expectations of ofcers with talents that are indemand. It must also be continuously resourced,executed, measured, and adjusted across several years

and budget cycles. Absent this, systemic policy anddecisionmaking failures will continue to confoundArmy efforts to create a talent-focused Ofcer Corpsstrategy for success. With mutually supporting prac-tices in the realm of accessions, development, andemployment, however, a sound ofcer retention strat-egy can forestall a talent crisis, allowing the Army toselect its leaders rather than settle for them.


7/71

1

TOWARDS A U.S. ARMY OFFICER CORPSSTRATEGY FOR SUCCESS:

RETAINING TALENT

INTRODUCTION

The latest global economic downturn has destroyedAmerican wealth to an alarming extent. Declining realestate values have reduced home equity by $5.1 trillionnationally, and millions of people have lost trillions ofdollars in the stock market.1 This grim news holds ourattention because we expect our investments to yieldhealthy returns, not daunting losses. Inadequate orfailed investments curtail our prospects for a successfulfuture.

Much like the citizens it serves and protects, theU.S. Army has also made signicant investments in

its future, especially in its leadership. In particular,the Army has devoted billions of dollars to ofcerundergraduate-level education, world class training,and developmental experiences. Since the late 1980s,however, prospects for the Ofcer Corps future havebeen darkened by an ever-diminishing return on thisinvestment, as company-grade ofcer retention rateshave plummeted. Signicantly, this leakage includesa large share of high-performing ofcers, many ofthem developed via a fully-funded undergraduateeducation.

In the last few years, the Army has responded to thischallenge with unprecedented retention incentives, toinclude broadly offered cash payments. The objectivehas been to retain as many junior ofcers on active duty

as possible. However, such quantity-focused incentiveprograms run counter to a talent-focused OfcerCorps strategy. The objective should not be merely to


8/71

2

retain all ofcers, but to retain talented ofcers whilesimultaneously culling out those lacking distributions

of skills, knowledge, and behaviors in demand acrossthe force.

Given the hierarchical nature of the Armysorganizations, retaining sufcient rather thanoptimally performing ofcers could have adverseconsequences for the Armys future. New ofcercohorts of high-potential talent may be driven away bythe prospects of serving under lackluster leadership,while those continuing their service may experiencestunted development due to a dearth of talentedmentors. Left unchecked, such developments couldcascade across all ranks, requiring a generation torectify and meanwhile signicantly undermining theOfcer Corps performance levels. With mutuallysupporting practices in the realm of accessions,

development, and employment, however, a soundofcer retention strategy can forestall this talent crisis,allowing the Army to select its leaders rather than settlefor them.

TALENT RETENTION GENERATES BENEFITSAND MITIGATES RISKS

In previous works, we have argued that everyperson has talent that can be liberated and extended ifthey are properly employed. This is not to say that allpeople can or should be retained, however. What kindof ofcer should the Army seek to keep? The answeris those ofcers whose individual talent sets best alignwith current and future requirements.

This is easy to say but tough to deliver, particularlyas todays operating environment is increasinglycharacterized by high levels of task interdependence,


9/71

3

skill specicity, and uncertainty. It is made even tougherby the fact that, in its core warghting competencies,

the Army cannot buy talent from outside. Theprofession of arms is indeed a demanding profession,requiring a distribution of skills, knowledge, andbehaviors that takes years to assemble.

A 35-year-old project manager at Microsoft, forexample, may possess an abundance of the generalskills demanded by the Army in its core talentsegment of eld grade ofcers. He or she will not,however, command the specic knowledge andbehaviors required to plan a battalion hasty defense,effectively represent the Army to the news media,predict enemy courses of action, or care for the familyof a fallen comrade. Nor will he or she immediatelyacculturate to a profession unlike any in the privatesector, one that employs deadly force within a moral-

ethical framework as sanctioned by responsible civilauthorities. The ofcer ethos is honed across a seriesof progressive entry-level experiences, allowing theArmy to observe the degree to which its junior leadersembody it while the scope of their authority is stillrelatively narrow.

Therefore, whether the Army seeks to expandlateral entry in some areas or not, it is clear that therewill always be signicant limits on its ability to buytalent from outside.2 New accessions and internaldevelopment processes will continue to generate anoutsized portion of the Ofcer Corps talent pool. Thisentails a signicant investment that can yield enhancedforce capability and national security, provided theArmy retains the talent it needs.

Given that the Army is competing in the Americanlabor market for its ofcers, its retention strategymust focus upon talent, guard against systematic


10/71

4

decisionmaking errors, redress market failures, andcreate an employment climate that powerfully meets

the expectations of ofcers with talents that are indemand. Figure 1 demonstrates the consequences offailing to balance service expectations against externalopportunities in a limited lateral entry organization:

Figure 1. Talent Flight Reduces WorkforceProductivity.

As we see in Figure 1, talent ight occurs, leadingto employment mismatches. This not only reducesproductivity, but also lowers morale, raises costs,increases personnel turbulence, and results in quantity-

focused rather than talent-focused practices. It runscounter to good talent management.

In contrast, Figure 2 highlights the benets to anorganization of meeting the expectations of its talentedworkers. The ability to screen, vet, and cull for talent is

restored, and optimal productivity ensues.


11/71

5

Figure 2. Talent Retention Optimizes WorkforceProductivity.

In addition to optimizing productivity, talentretention dramatically lowers the costs of internaltalent development. The longer talented ofcerscontinue their service, the more time the Army has torecoup the costs of their development. Relative to thevalue of their performance, developmental costs areparticularly steep in the rst 3 to 5 years of ofcerscareers, when they receive signicant education andtraining, as well as indirect benets that are generallyon par with those of more experienced (and thusmore productive) ofcers. Retaining talented ofcersbeyond the 5-year mark (seasoned captains) offsetsdevelopment costs via increased productivity. It alsoreduces retraining costs, the administrative costsassociated with higher personnel turnover, and the

costs of increased accessions to make up for seasonedcaptain shortfalls.


12/71

6

This last point is particularly important. As theArmy has increased lieutenant production to replace

the talented captains lost to the private sector, thenumber of new ofcers waiting to ll a nite number ofplatoon leader and company executive ofcer positionshas increased.3 As job queues have grown, idealdevelopmental experiences have declined, and morelieutenants are given make-work duties that deatetheir career enthusiasm. Furthermore, as the Armytries to cycle its new ofcers through a nite numberof developmental opportunities, the average numberof months served in key positions is being signicantlycompressed (see Figure 3). This trend compounds thechallenge, reducing opportunities for young ofcersto benet from experiential learning, mentorship, anddevelopment. Reduced developmental opportunitiesalso mean fewer evaluative opportunities for the

Army, making it increasingly difcult to screen, vet,and cull for talent.

Figure 3. Over-Accessing Ofcers Is UndercuttingDevelopmental Opportunities for Lieutenants.


13/71

7

A closely related development has been compressed

promotion timing. As the Army commissionedthousands of excess lieutenants to replace the captainsit failed to retain, it simultaneously shortened time-in-grade requirements for promotion of these lieutenantsto captain. As a result, between 1992 and 2004, theshare of captains with less than 4 years of active federalcommissioned service rose from 8 percent to 30 percent,and fewer than half of all captains had over 6 yearsof commissioned service (see Figure 4).4 As captainexperience levels declined, the Army simultaneouslyredesignated hundreds of former captains duties asmajors duties, perhaps in part because a captain wasnow increasingly unlikely to possess the experienceneeded in certain jobs. Increased losses among highpotential junior ofcers has thus signicantly shifted

the distribution of captains in the direction of lessexperience.

Figure 4. Changing Experience Levels of Captains(in Terms of Years of Service).


14/71

8

In short, low junior-ofcer retention increases

risks to the well-being and capabilities of the OfcerCorps. It strips away the Armys ability to screen, vet,and cull for talent, forcing it instead to over-access,increase promotion rates, and compress time-in-grade requirements. It degrades the developmentalexperiences of junior ofcers and undercuts theArmys ability to discern which ofcers possess thetalent it needs. In part, these challenges are due to thecontinuation of human capital management practicesfrom a bygone era.

COMPANY MAN EMPLOYMENT PRACTICESCANNOT COMPETE IN TODAYS LABORMARKET

The TV showMad Men is a pop culture phenomenon.Set in the Kennedy era, it chronicles life inside a ctionalMadison Avenue advertising rm. In 3 years, the showhas won several Emmy awards and critical acclaimfor its historical authenticity. While audiences areenthralled with the shows accurate depiction of socialmores in the 1960s, it does equally well in capturingthe corporate culture of the time. This culture includesan ethos of lifetime service to the rm by its employees,part of the organization or company man systemthat held sway in America into the 1980s.5

Under that system, companies sought to employthe same workers throughout their entire careers in aneffort to recapture training costs and preserve loyaltyand continuity. Internal managerial development and

advancement were key elements of the system, aswere rotational assignments designed to broaden thecorporations highest-potential members, who served


15/71

9

as feedstock for its future leadership. Employmentdecisions were made largely by employers, not

employees. Intercompany movement was not unheardof, but it was rare compared to todays uid labormarket. When it did occur, it was often the result of abusiness failure, merger, takeover, or perhaps a seniormanagement acquisition from an arch competitorto capture business intelligence and clients from theenemy.

As a rule, however, poaching junior or mid-leveltalent from competitors was the exception rather thanthe rule. The DuPont man who showed up at IBM orPzer would be viewed skeptically, his loyalty underquestion for having left the rm that had investedso much in his initial development. Even if thenewcomer had understandable reasons for seekingnew employment, there was always the question of

whether he could surrender the cultural baggage of hislast rm to t in at a new one.

Given such cultural realities, young executivesgenerally sought continuing professional opportunitieswith their initial corporate employer rather thanelsewhere. Healthy pension plans and the generallyexcellent promotion opportunities of the post-WorldWar II boom period were additional disincentives toight. Industrial era rms were highly specialized,creating additional barriers to intercompany talentmigration. Because of the low personnel turnoverinherent in this business climate, employers biggestpersonnel concern was whether they had a sufcientsupply of talented employees, and how much internaldevelopmental effort should be expended.

In sum, the company man system embodied humancapital management practices far different from thosedemanded by the information-age economy which


16/71

10

emerged in the 1980s. As a result, todays employmentmarket is characterized by high levels of intercompany

and innercompany mobility. Talented employees havefar greater control over their career options than everbefore, a situation made possible by the overwhelmingdemand for highly educated employees with talentsfor conceptualization and knowledge creation.

Because the Army must necessarily limit lateralentry, it will always retain some of the hierarchical andbureaucratic elements of the company man era. Aslabor market conditions began to change in the 1970s,however, the Army could have jettisoned many ofits inefcient industrial era practices and introducedelements of an internal talent market (see our discussionof the Ofcer Career Satisfaction Program later in thismonograph). Giving ofcers greater voice in theirassignments increases both employment longevity and

productivity. The Armys failure to do so, however,in large part accounts for declining retention amongofcers commissioned since 1983.

For example, about 60 percent of ofcerscommissioned in the late 1970s via Reserve OfcerTraining Corps (ROTC) and West Point scholarshipprograms remained on active duty through 8 years ofservice. As a result, the Army enjoyed an ample supplyof seasoned captains to ll key staff positions andcould be highly selective as it considered captains forpromotion to major. By the mid 1980s, however, only40 percent of ofcers being commissioned from thesescholarship sources remained on active duty through8 years of service. As a result, seasoned captains werein increasingly short supply.

Why did the Armys talent management practicesremain trapped in the past? How did it move from asenior captain surplus, then to shortage, then to crisis


17/71

11

in the decade following the end of the Cold War? Inpart, it may be because some of the Armys personnel

managers missed the epochal innovation embodiedby the rise of information technology in the 1980s.Having come of age in the industrial era, perhapsthese ofcers had imbibed too deeply from thecompany man system. Regardless, as they directed theArmys restructuring in the late 1980s and early 1990s,personnel managers continued to manage talent viaoutmoded techniques, to include generic forecastingmodels and indiscriminate quality control tools.

For example, ofcer strength forecasting modelsfailed to account for the economys increased appetitefor highly-educated workers. Army undergraduatescholarship programs had created talented youngofcers who were in greater demand than ever before,and corporate America undertook an aggressive talent

recruitment campaign to poach them (a practice whichcontinues today). In particular, this demand for highlyeducated talent drew increasing numbers of West Pointand ROTC scholarship ofcers out of the Army, andby 2001 the captain retention situation was becominguntenable.

The Army had always been mindful of ofcerretention rates as a function of commissioning sourcei.e., West Point, ROTC, or Ofcers Candidate School(OCS). Such analysis indicated that West Point ofcersremained in the Army at the lowest rates; ROTCofcers remained at middling rates, and OCS ofcersremained at high rates (see the grey-shaded panel inFigure 5).


18/71

12

Figure 5. Ofcer Retention Rates:Commissioning Source vs. Procurement Program.

However, when ofcer retention rates were analyz-ed according to procurement programi.e., particularcommissioning programs with distinct directives,resourcing, and contractual obligationsa very dif-ferent picture emerged (see the white-shaded panel inFigure 5). Four-year scholarship ofcers from ROTCand West Point remained in the Army at the lowestrates, followed in order by 3-year and 2-year ROTCscholarship ofcers, nonscholarship ROTC ofcers(NS ROTC), and OCS ofcers drawn from the enlistedranks (OCS-IS).6 By failing to anticipate the effect thatthe information age would have on scholarship ofcerretention, Army forecasts grossly underestimatedthe downturn in junior-ofcer continuation rates that

would begin with those commissioned in the late1980s.

Use of such personnel management practices asvoluntary separation further exacerbated the challenge.


19/71

13

In the industrial era, voluntary separation policiesusually engendered a self-culling by employees who

were poor talent matches for their organization.This softened their separation from the company,saved them the embarrassment of eventual removal,relieved them from existing contractual obligations,and often provided a modest nancial cushion to easetheir transition. In the context of the information age,however, such incentives had a much different effectwhen offered to Army ofcers. They opened the doorfor an exodus of highly educated, high-performingleaders, those the Army had invested the most in andwhose talents aligned well with critical employmentrequirements.

Consider. Beginning with those commissioned inthe mid-1980s and continuing through today, WestPoint and ROTCs 3- and 4-year scholarship of-

cers have remained in the Army at about two-thirds tohalf the rate of OCS ofcers from the ranks and ROTCofcers without scholarships. Years of peacetime andwartime performance data, however, clearly demon-strate that, once commissioned, the scholarship ofcersare disproportionately likely to possess the conceptualand problem-solving talent demanded by jobs such ascommander, executive ofcer, or operations ofcer.Because high-quality education amplies experientiallearning capacity, this talent advantage grows as theseofcers move from company grade to eld gradeassignments of increasing scope and complexity (seeFigure 6).


20/71

14

Figure 6. Ofcer Procurement Program.

In other words, the diligent screening inherent in the

Armys scholarship programs, coupled with the qualityof education generally embodied by those programs,produces ofcers whose talents align extremely wellwith complex jobs at the senior company and eldgrade levels.

Let us be clearwe are not arguing that scholarshipofcers are more talented than others, nor are weinterpreting these data to say that individual OCSand ROTC nonscholarship ofcers cannot performoptimally in these jobs. What we are saying is that asa population, the performance data for scholarshipofcers is signicant enough to predict their successin jobs the Army deems critical. They are not beingretained in sufcient numbers, however, creatingtalent gaps that simply cannot be lled with just-in-

time increases in accessions or changes in theaccessions mix.


21/71

15

To this day, the root causes of the current ofcershortage are still misunderstood by some. To be sure,

reduced ofcer accessions in the mid-1990s and ofcerstructure growth beginning in 2004 did not helpmatters, but the Army continues to leak ofcer talentat rates commensurate with those cohorts affectedby the 1990s drawdown.7 Low talent retention is theactual root cause of the challenge. In fact, by 2004,the retention challenge was already well-entrenched,as demonstrated by increased promotion rates,compressed time in grade, increased accessions, andshifts in the accessions mix.8

Each of these developments had undesirableripple effects. Rising promotion rates reduced Armyopportunities to vet ofcers for advancement.Accelerating promotions limited the time available forjunior ofcers to develop at each rank. Rising accessions

against a xed number of entry-level ofcer positionsreduced the likelihood that job opportunities availableto lieutenants would match their developmental needsor expectations (recall declining platoon leader time inFigure 3). Finally, the shift in the accessions mix awayfrom scholarship ofcers and towards OCS epitomizedtime-inconsistent behavior (pursuing short-termbenets in the face of serious long-term risk, a conceptwe will elaborate upon shortly). This all but ensuredan enduring mid-ranks talent gap, as OCS ofcerstypically retire from the Army after serving 10 to 15years of active federal commissioned service.

In retrospect, an effective retention strategy wouldhave provided the Army with a hedge against thedual risks of an increasingly competitive labor market

and the vagaries of wartime demand. For example, ifsuch a strategy had maintained ofcer retention ratesat industrial-era levels, the Army would enjoy full


22/71

16

manning in its eld grade ranks and could reducenew ofcer accessions by 20 percent.9 The potential to

generate such positive outcomes exists, but to do sothe retention component of any Ofcer Corps strategymust rest upon sound theoretical underpinnings.

THE ARMY MUST BUILD TALENT RETENTIONPOLICIES UPON SOUND THEORY

Both employers and employees face critical deci-sions bearing upon talent retention in an organization.For employers to successfully retain talent over the longterm, they must avoid time-inconsistentdecisionmaking.For employees to make sound career decisions, theymust assess the value of the next best alternative totheir current employment, also known as opportunitycost. Lastly, both employers and employees need a

mechanism for efciently exchanging commodities,but market failures often confound their ability to do so.Discussing each of these theoretical concepts withinthe context of the Armys ofcer retention challengesshould help clarify them.

The Armys Time-Inconsistent Behavior.

As opposed to the rank-stratied representationsof the Ofcer Corps used by Army personnel strengthmanagers, we view ofcers as talented people movingacross time through a funnel-shaped pipeline (seeFigure 7). Time is the critical component of thismodel, the unifying aspect of a successful long-termofcer strategy.Accessions decisions made today affect

development efforts over a 30-year horizon, are closelyconnected to retention rates, and ultimately shape theemployment of talent in the senior leader ranks some 25years later. The length of time between ofcer strategy


23/71

17

decisions and their outcomes may at rst blush seemirrelevant, but the implications are often far-reaching.

In the 1970s, economists and Nobel Laureates FinnKydland and Edward Prescott characterized theseimplications as the time inconsistency problem.10

Figure 7. Army Ofcer Human Capital Model.

Time inconsistency refers to the irrational reorder-ing of preferences as the consequences of our choicesbecome more proximate in time. For example, smokersmay plan to enjoy smoking today but quit tomorrow toimprove their health. The next day, however, their planis the same; enjoy smoking today and quit tomorrow.This goes on, and they may never quit even thoughthey want to, hence the inconsistency. The risks of this

behavior are tremendous because while benets accrueimmediately (the pleasure of smoking), costs accruewell into the future (lung or heart disease, death).


24/71

18

Similarly, because it unfolds across decades, thebusiness of building an Ofcer Corps is ripe for time-

inconsistent behavior, and the Army has fallen victimto it. In fact, for the last several years, the Army hasimplicitly accepted near-term benets in exchange forlong-term risks to the Ofcer Corps. One example wasthe end of forced distribution ratings for lieutenantsand captains, which occurred in 2004. Eliminatingforced distribution ratings made it extremely difcultto distinguish high-potential ofcers from the others,the same challenge the Army faced on the eve ofWorld War II.11

Another example of time-inconsistent behavior wasa signicant increase in ofcer promotion rates. Whenthese increases were briefed in the Pentagon in 2004,a senior Army leader responded, Its a great time tobe a captain. In his estimation, the Armys mounting

near-term ofcer shortage clearly trumped the need tovet and cull talent for the future.

By promoting and advancing ofcers whopreviously would have been culled from service,however, the Army only accelerated talent ight.Ofcers forced to serve under lackluster leaders willseek opportunities elsewhere, preferably where talentmatters. As retention rates continue to fall, short-term demands will force the Army into additionaltime-inconsistent behavior, further exacerbating theretention challenge. Eventually the Army could reacha tipping point where the downward spiral accelerates,and its talent core collapses. Much like an individualstime-inconsistent behavior of smoking, the true costshit unexpectedly in the form of a total breakdown. By

then it is too late.Guarding against time-inconsistent behavior

requires signicant discipline. In the currentenvironment, most Army strategic leaders direct


25/71

19

manpower policy for fewer than three years.12 Thisplaces an inherent emphasis on the now, creating an

ideal breeding ground for manpower challenges thatwill emerge 5, 10, or 20 years into the future.

An Ofcers Opportunity Cost.

While it is important to understand how theArmys decisions shape ofcer retention, it is equallyimportant to understand how individual ofcersmake the decision to stay or to leave. Althougheconomic decision theory has many dimensions, itreally boils down to a very simple principle: peoplechoose the option they believe will provide the highestsatisfaction. Each of us does this daily: Coke or Pepsi,cream or sugar, stairs or elevator?

The same is true of far weightier decisions. Each

ofcer, whether they realize it or not, routinely weighsthe opportunity cost of his or her service in the Army.In the context of our discussion, opportunity costis the value of an ofcers next best employmentalternative outside of the Army, an opportunity thatis forfeited by the decision to continue commissionedservice (see Figure 8).

Factors that may affect an ofcers opportunitycost include unemployment rates in the civilian sector,educational opportunities, potential civilian com-pensation, job satisfaction, and spousal employmentopportunities. For the most part, the Army can do verylittle to inuence an ofcers opportunity costeachpersons is different, governed by the intersection ofhis or her talent set with current market conditions.

Those with the highest opportunity costs are the oneswith the most to gain by leaving the Army. Generallyspeaking, these ofcers possess the talent needed to


26/71

20

perform well at the Armys highest levels because, aswe have seen, there is a high correlation between thetalents sought by the Army and those sought by themarketplace.

Just as ofcers unique talent sets shape theiropportunity costs, so too do they shape their expec-tations of military service. As Figure 8 illustrates, anofcer weighs his or her opportunity cost against theseexpectations, which the Army can shape via soundpolicies. Expectations run the gamut from current oranticipated job satisfaction and promotion potentialto the value of retirement and insurance benets,commissary privileges, the scope and quality of familymedical care, fully-funded educational opportunities

for oneself and ones family, etc. For some, jobsatisfaction may trump any earnings differential. Forothers, education benets may matter most.

Figure 8. Individual Retention Decision.


27/71

21

While weighing service expectations againstopportunity cost seems a fairly straightforward affair,

the element of time complicates matters. Marketconditions are dynamic. Information is imperfect.Family needs change. For the most part, however, therelative stability of Army policies allows ofcers tovisualize their career trajectories with some accuracy,whereas forecasting civilian sector opportunities ismuch more difcult. That very predictability givescommissioned Army service a slight advantage inhead-to-head competition with potential alternatives.This is why the Army must thoughtfully considerall ofcer personnel policiesif it unthinkinglyintroduces career uncertainty, it may forfeit one of itskey advantages in todays labor market.

Market Failures and Talent Retention.

A market failure exists when there is an inefcientuse of goods or services and a better outcome ispossible. Correcting market failures via thoughtfulpolicies often yields tremendous efciencies, withgains far outweighing losses. One example of a marketfailure is a missing market, the lack of an efcient wayto exchange a service. Bureaucratic organizations suchas the Army are often riddled with missing markets,but the one most germane to our discussion is themissing ofcer talent market.

Most ofcers desire an assignment that leveragestheir unique talent set. At the same time, the Armywould benet tremendously if it could successfullymatch individual ofcer talents against requirements.

Productivity would soar. Satisfaction would improve,leading to higher retention. Currently, however, thereis no talent matching market mechanism, no way for


28/71

22

Army strength managers and ofcers to make efcienttalent transactions. As a result, the ofcer talent market

fails to function optimallyin other words, assignmenttransactions still occur, but there is a signicantmismatch in talent supply and demand.

Markets can also fail from asymmetric informationchallenges, where one party has more or betterinformation than the other.13 This is true of the ofcertalent market. All ofcers have more informationthan the Army regarding both their opportunity costand their expectations of military service. Becausethe Army knows relatively little about each ofcersparticular desires and capabilities, and because it treatsindividuals as interchangeable parts, it can do littlemore than offer generic retention incentives. Whenit does so, this information imbalance ensures thatofcers who intend to stay in the Army are more likely

to opt for retention bonuses than those who intend toleave.

Another form of market failures is externalitiesimpacts upon people outside of the transaction. Thesecan be positive or negative. For example, when atalented ofcer decides to stay in the Army, that actionproduces a positive externality that may inuenceothers to continue their service. Conversely, when theArmy mismatches an ofcer with a requirement, thatmismatch creates a negative externality that may causeseveral peers or subordinates to leave the service.

FROM THEORY TO PRACTICE:BUILDING SUCCESSFUL RETENTION POLICIES

Improving ofcer talent retention requires far morethan dramatic pay raises or other nancial incentives.First, it calls for a mutually reinforcing mix of soundaccession, retention, development, and employment


29/71

23

policies. By employing all individuals in the right placeand time, and by providing them with the type and

amount of developmental opportunities best suited totheir needs, the Army can engender a virtuous cyclethat ensures the highest possible retention of the ofcertalent it requires.

Second, these policies must acknowledge the dis-tinct career phases which comprise an ofcerscareer. This is critical because each phase is associatedwith different opportunity costs and service expec-tations. As a result, it takes differentiated policies topositively affect ofcer continuation rates across acareer. There are four career phases to consider (referback to Figure 7).

Phase I: Receipt of Commission to End Active DutyService Obligation (ADSO).

Contingent upon their commissioning source, allnewly commissioned ofcers incur an ADSO of 3 to5 years. More than half of each year groups 20-yearattrition rate occurs within 6 months of completingan ADSO. Since the mid-1990s, for example, only 55percent of West Point graduates, who incur a 5-yearservice obligation, remain on active duty to 5 1/2years of service.14 Therefore, retention strategies in thisphase must focus on creating positive company gradeexperiences, as well as positive expectations for futureeld grade service.

Phase II: End of ADSO to 10 Years of Service.

As they approach 10 years of service, the proba-bility that ofcers will remain on active duty untilretirement eligibility climbs to more than 80 percent. In


30/71

24

this phase, an ofcers career calculations often includehigher education goals and whether those goals can be

met in the Army. Service to this point often mitigatesthe effect of having earlier served under a lacklusterleader, increasing the odds that ofcers will encountertalented professionals who can instill in them a desirefor continued service. To get them here, however, theArmy must create positive expectations regardingcontinued employment within their talent set, selectionto eld grade rank, and rewarding service to the 20-year point (see Figure 9).

Figure 9. Growth in the Value of Family Benetsby Years of Service.

Phase III: 10 Years to 20 Years of Service.

At this point, most ofcers are committed to a 20-

year or longer career. They understand their profession,they have a strong sense of what they can accomplishas an ofcer, they have a growing need for family


31/71

25

medical and other benets as indicated in Figure 9,and they are more focused upon possible retirement

benets. The Armys dened benet pension plan isnothing to dismiss lightlya 20-year retirement isworth approximately $1.4 million.15

Phase IV: 20 Years of Service to Mandatory Retirement.

This is when ofcers typically enter the Armysstrategic talent segment. At this point, they haveheavily leveraged most material and scal benets ofactive duty. They are already vested in their retirementplan and incur little additional nancial advantagefor each day they serve beyond the 20-year mark. Asa group, their service expectations shift markedlytoward a desire to inuence signicant outcomes andto enjoy their work. Since lateral entry into the Army

at this point is, of course, impossible, the Army mustdiligently guard against talent leakage. Opportunitycosts for these ofcers tend to rise due to theirexperiences and accomplishments, which are valuedin the marketplace.

EVALUATING EXISTING RETENTIONPROGRAMS

Evaluating ofcer retention programs withinthe context of the theories outlined to this point isilluminating. It demonstrates the perils of ignoringmarket principles as well as the benets of heedingthem. Two recent retention programs that lendthemselves to comparison are the Critical Skills

Retention Bonus (CSRB) and the Ofcer CareerSatisfaction Program (OCSP).


32/71

26

In scal year 2007-2008, the Army faced a substantialshortage of seasoned captains. As a remedy, it offered

the CSRB to all competitive category (and MedicalService) active duty ofcers commissioned between1999 and 2005. The key elements of this program rancounter to the sound market principles that shouldunderpin any retention policy. As a result, CSRBmay actually have done more harm than good. TheCSRB offered $25,000, $30,000, and $35,000 lumpsum payments to ofcers in exchange for 3 years ofservice.16 With a cost to taxpayers of $500 million, thereis no evidence that it improved retention.17 In fact, theincentive was capitalized upon by a population thatdid not require it77 percent of those captains whorequested the incentive in the autumn of 2007 hadpreviously indicated an intention to stay on activeduty beyond their initial service obligation.18

The programs aws were many. First, CSRB madeno effort to retain talented ofcersits focus was onquantity. Second, programs such as CSRB can causesome to forgo other retention incentives in the beliefthat they will eventually be offered a second bite of theretention incentive apple. This effect is counter to thatdesired, epitomizing time-inconsistent behavior. Third,the bonus reected no consideration of career phaseeffects upon ofcer continuation ratesby offering theincentive so broadly (from ADSO completion all theway to 8 years of service), the Army exacerbated itsretention challenge for ofcers between 5 and 10 yearsof service.19

In terms of a lifetime earnings comparison, eventhe high-end CSRB benet of $35,000 was not enough

to forestall service departure by ofcers alreadyplanning to leave due to high opportunity costs. Atbest, the bonus would retain ofcers with much lower


33/71

27

opportunity costs, thus producing talent mismatchesfor the increasingly complex jobs awaiting them. At

worst, it would pay enormous economic rent to ofcerswho were planning to stay in the Army anyway.20 Itis a textbook example of a lagging or reactive policy,triggered because a disproportionate share of high-performing junior ofcers had already left the Army.21

Instead of throwing money at its challenges andhoping for some benet, the Army should insteadbuild its ofcer retention programs upon the sameprinciples governing the labor market in which itcompetes. These programs must be forward-looking,expending resources where they will create thehighest talent return on investment. They shouldrecognize the linkage between accessions, retention,employment, and development policies. Perhaps mostimportantly, the Armys ofcer retention programs

should specically target ofcers possessing talentthat is actually in demand across its formations andinstitutions. There is no need for the Army to accepttalent mismatches.

The Ofcer Career Satisfaction Program (OCSP) isa retention initiative designed with these principles inmind. For year groups 2006 and beyond, OCSP is offer-ed to ROTC and USMA cadets prior to commissioning.Cadets can obtain their branch of choice, post ofchoice, or a guaranteed option to attend graduateschool in exchange for extending their commissioningADSO by an additional 3 years. Once commissioned,participating ROTC scholarship ofcers will serve7 years of their 8-year Military Service Obligation(MSO) on active duty, while participating West Point

graduates will serve all 8. The graduate school optionallows these ofcers to attend the school of their choicewith study in the discipline of their choice. Because


34/71

28

it is an option, ofcers may attend graduate school,leave the Army upon completion of their 3-year ADSO

extension, or forgo exercising the option and remain inservice.

Unlike the CSRB, the OCSP is not a reactive policydesigned to entice everyone to stay. Instead, it is squarelyfocused upon a large, poorly retaining populationwith talents the Army deems critical.22 Recall thatthese ofcers are more likely to possess the conceptualand problem-solving talents demanded by jobs suchas commander, executive ofcer, or operations ofcer,and that their talent advantage grows as they movefrom company grade to eld grade assignments ofincreasing scope and complexity (refer back to Figure5). By offering this program to ROTC and West Pointcadets, the Army aims a signicant portion of theretention incentive at ofcers who would otherwise

leave active duty prior to year 8. As an additionalbenet, the Army avoids any issues of fairness becausethe offer is made at the source of commission, for whichany aspiring applicants can compete.

OCSP generates signicant benets preciselybecause it heeds market principles. For example, itavoids a time inconsistency problem by committingthe Army and the individual to a service contractwhich is executed 4 to 8 years into the future. OCSPalso addresses market failures by providing marketsthat had been missing. Previously, many cadets wereunable to secure their branch or post of choice becausebranching and posting algorithms are based primarilyon academic standing. Over the past 4 years, however,more than 4,000 cadets participated in OCSP to secure

their branch or post of choice, guaranteeing the Armymore than 12,000 obligated man-years of service at nocost to the Army. Quite clearly, giving new ofcers


35/71

29

some voice in their assignment process immediatelyincreases their satisfaction and helps meet their

expectations of service.23

Another missing market was for graduate school.In light of the 175 percent wage premium the typicalgraduate degree holder garners over a college graduate,it is not surprising that scholarship ofcers viewgraduate education as an important career objective.Indeed, a majority of ofcers who remain in the Armybeyond 10 years of service but do not participate inthe Armys existing Advanced Civil Schooling (ACS)program obtain graduate degrees on their own.24 Inaddition, many ofcers who earned a graduate degreevia ACS report that, absent this opportunity, theywould have departed the Army.25 This demonstrateshow powerfully graduate-level educational opportun-ities can affect service expectations.

Until the OCSP was instituted, the odds of attendinggraduate school under ACS auspices were less than 1in 10 for the thousands of new ofcers commissionedannually. The program made only 415 graduateschool slots available per year. More than half of theserequired ofcers to immediately follow school withan instructor tour at West Point, pulling them out ofthe operational force for a total of 5 or more years.Many others required a post-graduation functionalarea utilization tour. OCSPs graduate school incentivehas no post-graduation teaching or utilization tourrequirement, however, allowing many more ofcersto attend graduate school for up to 2 years and thenimmediately return to an operational assignment.26For the Army, its return on investment is 3 days of

operating force service per ofcer for each day spent ingraduate school.

OCSPs graduate school incentive also takes intoaccount the way ofcer career phases affect retention


36/71

30

behavior and is designed accordingly. Once selected,the incentive moves ofcers through their rst career

phase by extending their Active Duty Service Obliga-tion to 7 or 8 years of service.27 Their attendance atgraduate school takes them through their second careerphase to approximately 10 years of commissionedservice. The average additional service obligationincurred in graduate school brings them squarely intotheir third career phase, to 15-16 years of service. Basedupon historical retention patterns, 96percent of ofcerswho reach this level of longevity continue to 20 yearsof service. There would likely be some retention lift inthe fourth career phase as well (20 years to mandatoryretirement), because graduate-level education not onlyenhances career satisfaction but also extends the talentadvantage critical to strategic-level leadership.

The power of the OCSP incentives to secure

thousands of years of obligated service whilesimultaneously creating a more agile, satised, andeducated Ofcer Corps is inarguable. For example, asshown in Figure 10, extending the branch, post, andgraduate education option to ofcers in year groups2006-09 stands to increase 8-year continuation ratesfrom 47 percent to above 69 percentlevels akinto those in the industrial era. By offering the OCSPprior to commissioning, the Army also eliminatesan information asymmetry, as cadets are unable topredict at commissioning whether or not they willstay on duty past their ADSO.28

The post-September 11, 2001 (9-11) GI Bill only addsto the OCSPs appeal. By electing OCSPs graduateschool for service option and extending their service

obligation by 3 additional years, cadets are guaranteedfully funded graduate school whether they use the


37/71

31

Figure 10. OCSP Raises Ofcer Retention Rates by50 Percent.

option or walk away from it. If they depart the Armyafter completing their extended ADSO, they can usetheir GI Bill benets to further their own education.If they stay in the Army, however, they not only cancontinue on to graduate school via OCSP, but they gaintransferability of their GI Bill benets to a dependent.29

The differences between programs such as theCareer Service Retention Bonus and the OCSP arefairly stark, but perhaps the most important differenceis that OCSP represents an investment in humancapital. As we know, education has value. It increasesworker productivity. It expands knowledge and thusextends the talent advantage of an individual. Becauseofcers who participate in the OCSP are much morelikely to reach 20 or more years of service, the Armys

return on its educational investment is therefore quitesignicant. Even when this incentive is made availableto ofcers who would have remained in the Army


38/71

32

without it, it does not amount to economic rent, asadditional education still yields the benet of increased

productivity.Of course, such is not the case with cash retention

incentives. These entail no investment in human capitaland therefore yield no productivity gains. Buyingservice with cash payments simply garners a windfallfor ofcers who would have remained in serviceanyway and saddles taxpayers with an unnecessaryexpense.

CONCLUSION

Over the last 3 decades, dramatic labor marketchanges and well-intentioned but unsound policieshave created signicant ofcer talent ight, engen-dering signicant risk for the Army. Poor retention

impedes the Armys ability to screen, vet, and cullofcers, undermining its ability to properly access,develop, and employ talent. Therefore, the Army can-not undertake thoughtful policy decisions in theseareas if its ofcer talent pipeline continues to leak atcurrent rates. High talent retention is a necessary pre-condition to creating the most capable Ofcer Corpspossible.

The Army cannot insulate itself from labor marketforces as it tries to retain talent. Therefore, the retentioncomponent of its ofcer strategy must rest uponsound market principles. It must also be continuouslyresourced, executed, measured, and adjusted acrossseveral years and budget cycles. Absent this, systemicpolicy and decisionmaking failures will continue

to confound Army efforts to create a talent-focusedOfcer Corps strategy for success.


39/71

33

ENDNOTES

1. Figures are from the National Bureau of Economic Researchas reported in the Wall Street Journal, July 28, 2009, available fromfinance.yahoo.com/career-work/article/107419/the-great-recession-a-downturn-sized-up.html?mod=career-salary_negotiation .

2. In a few cases such as the legal and medical elds, the Armymakes exceptions to its limited lateral entry policy. As ofcerbranches and Army missions evolve, the Army may be able toexpand its reliance on lateral entry. However, for branches that

focus on leading Soldiers, lateral entry runs counter to importantArmy culture.

3. We calculate the number of excess lieutenants accessed bya year group in Figure 3 as follows. First we calculate the totalnumber of lieutenant requirements from the PMAD for each of the3 years that a cohort serves at the rank of lieutenant and divide thatnumber by 3. This gives us the number of lieutenant requirementsthat a year group faces each year they serve as a lieutenant. Next,

we subtract the total number of lieutenant requirements for a yeargroup for each of the 3 years from the actual number accessed.This gives us three values for the excess accessions. We averagethose three values to get the average number of excess lieutenantsaccessed for a year group across the 3 years a year group servesas a lieutenant. We use Ofcer Evaluation Report (OER) data toestimate average platoon leader time. We validate this OER trendanalysis by checking it against TAPDB data. There is a similartrend, but the TAPDB shows a slightly higher level by about a

month or two. We rely upon OER data because there are manyinconsistencies with duty titles in the TAPDB.

4. Shifts in experience levels of ofcers are a result of multiplepolicy changes. Some causes of decreases in average captainexperience include early promotion of lieutenants to captain, earlypromotion of captains to majors, increases in accession cohort sizefor ofcers who reach the rank of captain, shifts in accession mixtowards sources that continue at low rates, and declining ofcer

retention.

5. For a thorough discussion of this system, see William H.Whytes classic exploration of the American corporate ethos, The


40/71

34

Organization Man, New York: Doubleday, 1956. For a contemporarydiscussion of the same subject matter, see Peter Cappellis Talenton Demand, Boston, MA: Harvard Business Press, 2008.

6. The analysis by procurement program successfullyreframed the retention discussion among senior leaders, but theterrorist attacks on September 11, 2001 (9/11) drew their attentionelsewhere and forestalled substantive efforts to raise ofcerretention until 2004. By then, the captain retention crisis wasfully entrenched. Among Year Group 1995 to 2001 ofcers, it washollowing out the ranks of junior ofcers and leaving inadequatenumbers of seasoned captains available for advancement tomajor. See Appendices A and B for a further discussion of ofcerretention forecasting and analysis challenges.

7. Beginning in 2004, the Army increased structuralrequirements for majors by 2,802 billets. This growth in eldgrade structure exacerbated ofcer shortages accumulated duringa decade or more of low captain retention. By adding thousandsof new eld grade ofcer requirements to its structure, the Armybrought its shortage of seasoned ofcers into such stark reliefthat in some quarters, growth rather than retention became thedominant construct for addressing ofcer shortages. Adherents tothis view argue that to accommodate ofcer structure growth, theArmy naturally turned to OCS accessions as it had when growthwas required during earlier conicts. However, such comparisonsare misleading. Unlike prior conicts, the Army now incorporatedall OCS growth into the Armys corps of regular, tenured ofcers.Given current promotion rates, these OCS ofcers can beexpected to serve at will until retirement. In prior conicts OCSgrowth came in the form of reserve ofcer commissions. As such,following hostilities, the majority of these ofcers were typicallyreleased from ofcer ranks during postwar demobilization anddownsizing. Finally, while prior events can cause subsequentreactions, the reverse cannot be true. Specically, increases inofcer accessions, promotion rates, shifts in the accession mix,and reduced time to promotion preceded ofcer structure growth.Therefore, the latter cannot have been the cause of the former.

8. See Appendix C for an extensive root-cause analysis of theretention challenge.


41/71

35

9. See Appendix D for the methods and calculations thatsupport a 20 percent reduction in accessions.

10. See F. E. Kydland and E. C. Prescott, Rules Rather thanDiscretion: The Inconsistency of Optimal Plans, The Journal ofPolitical Economy, Vol. 85, No. 3, 1977, pp. 473-492.

11. At the beginning of the war, rapid expansion of the Armyrequired promotion of large numbers of ofcers to senior grades.When the Army reviewed existing ofcer efciency ratings (itsForm 67 report, which lacked a forced distribution component),it discovered that of 4,000 ground ofcers of suitable generalofcer age, [over] 2,000 were [rated] superior and best. As sucha showing was perfectly worthless for the purpose [of screeningand vetting], the selecting authorities reluctantly fell back onpersonal knowledge, which is exactly what the Army thought itwas getting away from when . . . it inaugurated the [Form 67]. . . .It seems that raters typically used only superlatives in describingtheir men or damned them with faint praise. See E. DonaldSisson, Forced Choice: The New Army Ratings, PersonnelPsychology, Vol. 1, No. 3, Autumn 1948, pp. 365-382.

12. By comparison, in 2008 the average tenure of CEOs inNorth American rms was almost 8 years. Booz & Co. , availablefrom www.booz.com/global/home/press/article/45711808.

13. Wikipedia, available from en.wikipedia.org/wiki/Information_asymmetry. Examples of asymmetries include moral hazard,adverse selection, and principal-agent problems. In all cases anindividual has better information than the organization, whichleads to changes in behavior, poor screening and signaling, andmisaligned incentives for optimal performance.

14. U.S. Army Ofce of Economic and Manpower Analysis(OEMA) analysis of data contained in the Total Army PersonnelDatabase (TAPBD).

15. This net present value calculation assumes a 4 percentdiscount rate, 3 percent ination rate, is valued at 10 years ofservice, and assumes a life expectancy of 75 years of age.

16. The cash size of the incentive sorted by Army basicbranches, with more money being offered to some shortage


42/71

36

branch ofcers than others. With the right bonus levels, it ispossible to induce the required number of ofcers to extend theirservice with such just-in-time retention tools. However, bonusand incentive pay strategies entail substantial inefciencies andadverse second and third order effects. Note: Graduate schooland professional military schools were also offered as part of theprogram but had low acceptance rates.

17. Army G1 analysis of CSRB program indicates that thereis insufcient evidence to prove we have changed retentionbehavior. At best, the CSRB program placed a oor underhistorical retention rates.

18. U.S. Army Research Institute (ARI) analysis dated March25, 2008.

19. Offering the CSRB to ofcers up to 8 years of commissionedservice (YG 1999) only deepened the Armys asymmetricdisadvantage. By this point in their careers, ofcers have servedas platoon leaders, company commanders, and staff ofcers. Datashows that ofcers with 8 years of service have above a 80 percentprobability of continuing their careers to at least the 20-year mark.Therefore, of the money paid to YG 1999, at best 20 percent of itwould go to retaining ofcers. In contrast, YG 2005 had only 3years of service. As a result, they had much greater uncertaintyregarding their Army and private sector career options. For theseyounger ofcers, committing them to 3 more years of service mayhave had some benet.

20. Economic rent is a distribution in excess of the amountrequired to sustain a production process.

21. In an ideal world, one might hope to distribute incentivesonly to desirable ofcers who exhibit intentions to leave theArmy. However, once ofcers have explored the external labormarket, the cost of buying them back rises dramaticallyinother words, it is too late for an efcient incentive. Additionally,ofcers exhibiting intentions of leaving may stop making thetypes of investments in their career necessary to maintain theircompetitiveness for key assignments or advancement. Lastly,offering incentives to those who exhibit intentions of leaving theArmy can create perverse incentives for gaming the system.In other words, ofcers might explore outside opportunities, or


43/71

37

create the appearance of doing so, to trigger more Army retentionincentives.

22. Since the Armys 3- and 4-year scholarship programs alsocomprise about 40 percent of ofcer accessions, they afford theArmy its greatest scope to both raise ofcer retention and deepenits bench of ofcer talent.

23. Analysis indicates that not receiving a branch or postof choice has little impact upon post-commissioning retentionbehavior. Source: OEMA.

24. For example, 65 percent of competitive category USMAsource ofcers, Year Groups 1980 to 1993, not attending ACS,obtained a graduate degree on their own time.

25. OEMA survey of USMA faculty, September 15, 2004.

26. Of course, use of an educational incentive engendersboth budgetary and overhead (TTHS) costs. In steady statetheestimated cost of this program is $90 million. However, as opposedto other strategies, where payments would be made concurrentwith extended ofcer service, the educational incentive calls forpayments to begin, on average, 8 years into the future.

27. At year 8, participants in such a career education optioncould allow their option to expire and depart the Army, or theycould exercise their educational option by remaining on activeduty.

28. Regardless of stated intentions prior to commissioning,cadets are unable to predict their eventual service length. Some23 percent of cadets plan to serve beyond ADSO, yet half of thoseleave. Some 34 percent of cadets plan to leave at the completion oftheir ADSO, yet half of those end up staying. Of the remaining 43percent who are unsure, half of them end up leaving.

29. See Appendix E for a discussion of Ofcer Career Satisfac-tion Program implementation challenges. Participation ratesin the OCSP prior to the Webb GI Bill were high as shown inAppendix E.


44/71

38

APPENDIX A

FORECASTINGTHE CHALLENGE OFUNSTABLE STRUCTURESSUCH AS SOURCE OF COMMISSION PROGRAM

CHANGES

During the 1990s, the Army disaggregated ofcerstrength forecasts by commissioning programs.1 Theseinclude West Point (USMA), ROTC DistinguishedMilitary Graduate, ROTC Non-Distinguished MilitaryGraduate, OCS Distinguished Military Graduate, andOCS Non-Distinguished Military Graduate. Prior tothe mid-1990s, the distinction between Distinguishedand Non-Distinguished Military Graduate had been animportant commissioning consideration. West Pointofcers and Distinguished Military Graduates from

ROTC and OCS received a Regular Army commission,while ofcers who were not Distinguished MilitaryGraduates received an Other than Regular Army(OTRA) active duty commission. In other words, allWest Point ofcers were considered DistinguishedMilitary Graduates, whereas only a small xed shareof each ROTC and OCS cohort received the samedesignation.2

Within these groupings, the Army linked accessionmissions with expected loss rates to estimate the futurestrength of ofcer cohorts. They used these gures toestablish the length of time ofcers should remain in agiven grade, to establish the rate at which they shouldbe promoted, and to estimate accessions required insubsequent cohorts to backll entry level vacancies.

So long as ofcer retention relationships within thesecommissioning program groupings remained constant,accurate forecasting was possible. However, the


45/71

39

problem with predictive forecasts is that their accuracydepends upon the stability of key structures and

continuation rate relationships which are derived fromhistorical data. In the presence of shocks, these factorscan vary widely from historical trends before the lapseof time allows sufcient new data to accumulate andreveal new structures and relationships.

When personnel managers began to take note offalling ofcer retention in the early 2000s, they did notreturn to rst principles and evaluate the need to act.Rather, they saw this challenge through the structuresand relationships available from historical Armymanpower data. The Army saw low retentions of WestPoint ofcers rather than low retentions among 3- and4-year scholarship ofcers from West Point and ROTC.

Unfortunately, the actual stability of ofcerretention rates within and across ofcer groups is

a retrospective issue that can be judged only in thefullness of time. Since all ofcers enter the Army witha minimum active duty service obligation (ADSO) of3 years, and scholarship ofcers from ROTC and WestPoint enter with 4- and 5-year ADSOs respectively, thelag in detecting a change from historical retention ratescan be 3 to 5 years or longer. Thereafter, compensatoryadjustments to ofcer accession programs can entailan added lag of as short as a few months in the caseof OCS, to 5 years in the case of West Point, and 2 to 5years in the case of ROTC.

Additional lags in gauging the severity andpersistence of changes in retention patterns, and intaking action to redress these changes, can entailfurther years of delay. We estimate the effective sum

of these lags to be about 7 years. Due to the effects ofcompounding, small variations in ofcer retentionrates during this lag period can lead to widely disparate


46/71

40

outcomes. For example, a persistent 1 percentage pointdecrease in year-over-year junior ofcer retention

rates for an initial cohort of 6,000 ofcers can accrue acumulative decrease of over 1,400 ofcers available foradvancement to major.3 In commerce, when vagariesof market turbulence present such downside risk,prudent managers purchase insurance. Unfortunately,as it restructured, the Army did not insure against therisk of an anticipated decline in junior ofcer retentionrates. As a result, the Army is currently confrontedwith a signicant ofcer shortage.

During the 1990s and into the early 2000s, theArmy increasingly relied upon commissioning sourceswhich were associated with relatively high ofcerretention rates through 10 years of service. OCSaccessions increased from 9 percent to 40 percent oftotal commissions during this period. However, over

this period the mix of procurement programs withinthese commissioning sources changed dramaticallyand in ways that required new frames of reference todetect.

A generic model that uses average retention ratesand accessions numbers for each source of commissionillustrates how this situation unfolded across YearGroup 1991 to 2002 ofcer cohorts. Figure A.1 containsapproximate accession levels and 7-year ofcerretention rates by source of commission for Year Group1991 and 2002 ofcers.

Using the product of accession levels from the leftcolumn and continuation rates from the middle columnone can estimate the number of ofcers continuing to 7years of service. As indicated in the right column,

the size of continuing cohorts between 1991 and 2002would have been expected to increase by 530 ofcers,given the 630 ofcer increase in accessions over thisperiod.


47/71

41

Figure A.1. Expected Ofcer Year Group Strength7 Years after Accessioning

Based upon Commissioning Source Retention

Behavior(Army Competitive Category and Medical ServiceCorps ofcers [ACC+MSC]).

However, as illustrated in Figure A.2, a muchdifferent picture emerges when viewed through thelenses of ofcer procurement programs in lieu ofsources of commission. Rather than retaining 2,730Year Group 2002 ofcers as indicated in Figure A.1,retained ofcers declined to 2,450. This reducedestimate is due to the low retention rates and theinuence of structural accession program changesthat ensued between 1991 and 2002. Specically, whilethe number of DMG ofcers commissioned remained

constant, the mix of procurement programs fromwhich these ofcers entered the ofcer corps changeddramatically. For example, within ROTC, the number


48/71

42

of scholarship cadets grew by about 25 percent whilethe group of nonscholarship cadets fell by about 40

percent.

Figure A.2. Expected Ofcer Year Group Strength7 Years After Accessioning as Based UponProcurement Program Retention Behavior

(Army Competitive Categoryand Medical Service Corps Ofcers [ACC+MSC]).

Since ofcers from ROTC scholarship programscontinue at about 70 percent of the rate typical ofnonscholarship ofcers, the new mix of scholarshipand nonscholarship cadets yields an estimate of ROTCcontinuations that is 120 per year lower than arrivedat using the source of commission framework. Thatframework assumed the underlying mix of ROTCscholarship and nonscholarship ofcers would remain

xed. A similar situation occurs when estimatingcontinuations for OCS source ofcers. In this case,a divergence of 160 fewer retained ofcers ensues


49/71

43

between the two methods. This is due to the greatlyincreased share of Enlistment-Option ofcers as a

share of OCS accessions. As revealed over time, OCSEnlistment-Option ofcers have departed the Army atmuch higher rates than OCS-In Service ofcers. As aresult, OCS ofcers reaching 7 years of service wouldnot increase by 570 ofcers between Year Group 1991and 2002 as predicted in Figure A.1. Rather, since OCSEnlistment-Option ofcers separated at higher-than-expected rates, the increase in expected year groupstrength was only 435 additional ofcers as illustratedin Figure A.2.

Although accessions increased by about 630 ofcersper year between 1991 and 2002, the number of ofcerscompleting 7 years of active federal commissionedservice grew by far fewer ofcers per year group thanthe Army expected. Accumulated over 7 year groups

of ofcers comprising the Armys corps of majors, thisfeature of Army forecasting methods would result inabout 1,400 fewer ofcers than predicted by the timethe Army increased its ofcer structure in 2004. Asaddressed above, the time lag engendered in detectingand acting upon this situation was also about 7 years.Due to the length of precommissioning programs andpost- commissioning ADSOs, this lag is an unavoidableaspect of the Armys ofcer accession pipeline forwhich an effective strategy must account.

ENDNOTES - APPENDIX A

1. In the Total Army Personnel Database, this data is capturedin an ofcers source of commission (SOC). Key levels of thisvariable are USMA, ROTC DMG, ROTC, OCS DMG, OCS.

2. Typically, the top 15 percent of each ROTC and OCS cohortearn the DMG distinction based upon their standing on ROTCand OCS order of merit lists.


50/71

44

3. Reducing year-over-year retention rates in Column A by1 percentage point, we obtain the year-over-year rates indicatedin Column C. Multiplying the rates contained in Columns A andC by the cumulative retention rate at 36 months of service, 93.3percent, we obtain the base and adjusted cumulative retentionrates indicated in Columns B and D. At the end of 120 months ofservice, 46.8 percent of the starting population of 6,000 ofcers,or 2,811 ofcers, would remain on active duty using retentionrates exhibited by Year Group 1999 ofcers. Using the adjustedretention rates, the continuing population would fall to 2,600ofcers. Over seven ofcer-year groups comprising the Armyspopulation of majors, this 211 ofcer difference accumulates to1,472 fewer ofcers available for advancement to the grade ofmajor.

Monthsof

Service

Column A

Base Year-Over-Year

Retention Rate

Column B

CumulativeRetention

Rate

Column C

Adjusted Year-Over-Year

Retention Rate

Column D

AdjustedCumulativeRetention

Rate

36 93.3% 93.3%

48 91.3% 85.2% 90.3% 84.3%

60 83.5% 71.2% 82.5% 69.6%

72 89.7% 63.9% 88.7% 61.7%

84 88.9% 56.8% 87.9% 54.3%

96 91.3% 51.9% 90.3% 49.0%

108 93.5% 48.5% 92.5% 45.3%120 96.6% 46.8% 95.6% 43.3%

Number of Ofcers Remaining from a Starting Accession Population of 6,000 Ofcers

2,811 2,600


51/71

45

APPENDIX B

DEFECTIVE MEASURES AND OTHERCONFOUNDING CHALLENGES

One of the reasons the Army has trouble trackingand understanding its captains retention challenge isthat its retention metrics are deeply awed. Withinthe Army, the most frequently cited ofcer retentionmetric is the company grade attrition rate (see FigureB.1).

Figure B.1. Standard Company Grade AttritionRates.

This rate is calculated by dividing the number ofcompany grade ofcers who leave the Army in a givenyear by the number of company grade ofcers in theArmy that year. The reason this method is problematic

is that the denominator (the number of company gradeofcers in the Army in a year) is not a consistent frameof reference. Rather, as shown in Figure B.2, it uctuates


52/71

46

with promotion timelines, variations in commissioningsources and seasonality, and changes in accession

cohort sizes. While the number of captains who departthe Army could be exactly the same from month tomonth, changes in any one of these dimensions resultsin a completely different company grade attrition rate.For example, if the Army decides to promote ofcers tothe rank of major a year earlier than normal (as it did in2004), the attributes of ofcers at separation risk will befundamentally different than in prior years, and thusnot directly comparable. Moreover, the populationof ofcers at risk will be smaller relative to the totalcompany grade ofcer population. This is because thegroup of ofcers still under a commissioning serviceobligation will remain xed, while the total populationof company grade ofcers will shrink. The rate willremain high in steady state as long as the Army

continues early promotions of company grade ofcersto eld grade rank.

Figure B.2. Factors that Affect Junior OfcerContinuation Rates

and Operating Strength at 10 Years of Service.


53/71

47

Alternatively, an increase in the number of

accessions will drive down company grade attritionrates. As shown in Figure B.3, an increase from 100 to200 ofcers accessed, all else being equal, results in a.8 percentage point reduction (7.0 percent minus 6.2percent) in the company grade attrition rate. But whenaccessions reach a steady state of 200 ofcers per year,attrition resumes its former rate of 7.0 percent, a riseof .8 percent. While this may seem inconsequential,compounded across a decade the annual shortfall ofofcers to be advanced to major is considerable.

Figure B.3. Increased Ofcer Accessions Yield aTransitory Reduction

in Company Grade Attrition Rates that DisappearsWhen the

Ofcer Pipeline Returns to Steady State.

With regard to variations in time in grade, FigureB.4 demonstrates how changing promotion points canalso affect company grade attrition rates.


54/71

48

Figure B.4. A 2-Year Reduction in Time in Service toMajor Yields a

Permanent Increase in Company Grade AttritionRateswhile Leaving Operating Strength Unchanged.


55/71

49

APPENDIX C

AN ACCOUNT OF THE ARMYS FAILURETO UNDERSTAND THE ROOT CAUSESOF ITS RETENTION CHALLENGES

In the mid 1990s, Army personnel managersidentied West Point graduates as central to the juniorofcer retention problem. They found that a highnumber of these ofcers departed the Army as soonas they fullled their active duty service obligation(ADSO). In contrast, their analysis revealed that ROTCand OCS ofcers stayed in the Army at higher rates.Given the substantial costs to educate and train eachWest Point graduate, this raised questions about thedevelopmental environment at West Point, the servicepropensity of cadets entering the Academy, the size of

the Corps of Cadets, the academic program, the qualityof cadets entering West Point, and the preferenceafforded to West Point graduates in selecting theirbranch of service upon graduation.

Troubled by this situation, some West Point alumniidentied what they saw to be the crux of the low-retention problem. Having offered long service, andhaving entered the Army prior to the doubling of thesize of the Corps during Vietnam, they recommendedhalving the Corps of Cadets to increase cohesion andnarrow admission to those with a high propensity fora lifetime of service in the Army. Some of these retiredofcers also felt that the West Point Association ofGraduates (AOG) had run amok in helping graduatesnd civilian careers during the drawdown of the mid

1990s. Still other West Point alumni suggested that theArmy created an expectation of short service amongcadets during the 1990s by offering ofcers earlyseparation benets during the drawdown.


56/71

50

This focus on West Point led one high rankingofcer to suggest that the Army should reduce its

investment in an institution that taught its cadets toget out of the Army. In this same vein, some leadershipdevelopment experts argued that the problem of lowretention was an artifact of toxic leaders and a zerodefect culture in the Army. Finally, perhaps moreclosely approaching the likely nub of the problem, onesenior leader jokingly suggested that expanding thefootball team would help retention. His expectationwas that by lowering cadet academic quality the Armycould moderate ofcer attrition. That is, cadet qualitywas perhaps too high for Army needs and it confrontedWest Point graduates with substantial opportunitiesoutside the Army.

Each of the foregoing hypotheses was speculativerather than grounded in hard data, and none offered

a satisfactory explanation for what is, in fact, low junior ofcer retention extending well beyond WestPoint graduates. To get to the root of the problem, it isnecessary to analyze in depth the incentive structuresthat bring new ofcers into the Army. Specically,ofcer accession programs entail two general categoriesof incentives. These are a career as an ofcer for OCSand nonscholarship ROTC graduates, and, for WestPoint and ROTC scholarship ofcers, the additionalincentive of a fully funded undergraduate education.As shown in the bottom panel of Figure 5, whenexamining ofcer retention along these dimensions, aclear pattern emerges.

Nonscholarship ROTC and OCS ofcers remain inthe Army through 8 years of service at relatively high

rates. Two-year scholarship ofcers continue at the nexthighest rate, followed by 3-year scholarship ofcers,West Point graduates, and then 4-year scholarshipofcers. Observed in this light, the locus of low ofcer


57/71

51

retention can properly be seen to lie with the 50-pluspercent of ofcers who enter the Army on the offer of

an education and a career. Consequently, low ofcerretention to 8 years of service aficts a much largerofcer population than just West Point graduates.Those who stay in the longest came to the Army on thepromise of a job. Those who came into the Army on thepromise of a job and education stay at lower rates.

In light of this, it is clear that West Points programis not uniquely linked to low ofcer retention. Rather,low retention rates extend to ROTC scholarshipgraduates from a wide variety of schools. The samelogic applies to notions that USMA as an institution insome way conditions its graduates to leave the Armyat high rates. Such conditioning could hardly extend to3- and 4-year ROTC scholarship ofcers.

Reference to historical West Point continuation rates

also counter notions that West Point graduate retentionrates are linked to the size of the Corps. Due to the needto scale class size to gradually increasing new barracksavailability and other Academy infrastructure, thedoubling of the size of the Corps was an evolutionaryprocess rather than a sudden consummation, a growthrather than a creation. This process extended from1964 to 1975, embracing the Classes of 1968 through1975. Ten-year retention rates began to declinepriortothe start of the transition to a larger Corps, bottomingat 35 percent in 1968. Thereafter, 10-year retentionrates recovered to their pre-Vietnam War averages (inthe 60 to 65 percent range) in the period during whichthe Corps grew to its new higher strength. Ten-yearretention rates then stabilized at these high levels until

the end of the Cold War and the rise of the informationeconomy, peaking at 67 percent in 1979.

Todays low retention rates are a recent


58/71

52

phenomenon, aficting those classes reaching 10 yearsof service since the rise of the information age economy

in the mid to late 1980s. Moreover, while the WestPoint AOG may facilitate out-placement of Academygraduates departing active service, perceived AOGmischief in this regard cannot be the basis of the lowretention exhibited by ROTC 3- and 4-year scholarshipofcers. Similarly, Army separation policies during the1990s drawdown could not have engendered enduringexpectations of short service among West Point andROTC graduates because such expectations do nothave appeared to have shaped the behavior of 2-yearscholarship and nonscholarship ofcers.

Absent a broad anti-West Point or anti-intellectualbias, the suggestion that low retentions among WestPoint graduates is uniquely attributable to toxic leadersis counterintuitive. All else equal, such a situation

would require that by some enigmatic process, WestPoint graduates are disproportionately likely to fallunder the tutelage of toxic leaders. Otherwise, onemust inquire why such leaders would induce USMAgraduates to remain in the Army at half the rate ofnonscholarship ofcers.

Although West Point cadets exhibit very high andhomogeneous potential for service, recruited athletesdo fall disproportionately into the lower half of thecadet academic order of merit. This reality is the likelygenesis of tongue-in-cheek suggestions that increasingthe size of the football team would yield higher ofcerretention. In other words, cadets high in order of meritare presumed to exit the Army at disproportionatelyhigh rates after their ADSO expiration. However, for

a variety or reasons, the opposite is, in fact, the case.Based upon College Board scores and cadet order-of-merit standing, those USMA cadets with the highest


59/71

53

potential and performance as an undergraduateremain in the Army to their 10th year of service at

higher rates than cadets exhibiting lower potentialand performance. In particular, they remain at higherrates than recruited athletes. In part, this situation isan outgrowth of physical commissioning standardsrequired. Specically, after graduation, West Pointcadets who participate in intercollegiate athleticsexhibit higher than normal separation rates from theArmy for disability. Intercollegiate athletes are alsoless likely to meet USMA graduation requirements.Consequently, leavening the Corps by expandingthe football team would not only lower averagecadet academic quality, it would also lower USMAsgraduate yield and reduce average USMA graduateretention in the Army.

We thus return to the one reason for the recent

retention challenges. The nub of the problem lies withthe fact that high-potential ROTC scholarship ofcersand USMA graduates have a great deal to offer potentialemployers, be that e

Towards a U.S. Army Officer Corps Strategy for Success: Retaining Talent

Documents

Transcript of Towards a U.S. Army Officer Corps Strategy for Success: Retaining Talent