PROFESSIONAL PAPER 356 / April 1982

ESTIMATION OF THEPERSONAL DISCOUNTRATE: EVIDENCE FROMMILITARY REENLISTMENTDECISIONS

Steven CylkeMathew S. GoldbergPaul HoganLee Mairs

, DTICEL*

DEC5 -3iVCENTER FOR NAVAL ANALYSES

for public releamm trod L(a'3 t ]

I d8" 21 042

PROFESSIONAL PAPER 356 / April 1982

ESTIMATION OF THEPERSONAL DISCOUNTRATE: EVIDENCE FROMMILITARY REENLISTMENTDECISIONS

Steven CylkeOffice of the Chief of Naval Operations (Op-16)Mathew S. GoldbergCenter for Naval Analyses

Paul HoganOffice of the Assistant Secretary of Defense (MRA&L)Lee MairsOffice of the Chief of Naval Operations (Op- 16)

Accorssion For

NITIS GRA&IDThIC TAB EUzannounncedJustificat ion__...

Dis~tribution/

Avaldhbtlity Codes

Dit special

Naval Studies Group

CrNTER FOR NAVAL ANALYSES2000 North Beauregard Street, Alexandria, Virginia 223 11

ESTIMATION OF THE PERSONAL DISCOUNT RATE

I. INTRODUCTION

-,)This paper presents some estimates of the personal

discount rate of Navy enlisted personnel. Many of these

individuals can receive a bonus payment of/ up to $20,0'00 if

they choose to reenlist after completing 'heir initial term

of service.'>½rior to April 1979, the reenlisiment bonus was

paid in annual installments at the beginning of each year of

reenlistment. In fiscal year 1979 (FY79), the Department of

Defense requested sufficient funds to begin making lump sum

payments and, starting on 1 April 1979, the entire bonus has

been paid as a lump-sum at the date of reenlistment 1 The jrationale .or this- change is a familiar one. •ecause

individuals discount deferred payments, lump-sum /onuses

should induce more reenlistment than instal lmentibonus~s.'

comparing the effects of bonuses before nd after the

policy change, we hre able to0stimatelthe differential

impact of lump-sum and installment bonuses. We then use

these estimates to compute the implied discount rate.

We find that Navy enlisted personnel have a real

discount rate of about 17 percent. This estimate may seem

high, but it is comparable to Heckma 's (1976) estimates of

18 to 20 percent and Landsberger's (1971) estimates of 9 to

27 percent, and lower than Friedman's (1957 taslat:-r" .. e of 30

per-" -t. -

~If N~O1 M

-9 Pa

/ '-2-

There would seem to be a clear policy implication of

our findings. As long as the real discount rate relevant to

the Government is less than 17 percent, the payment of lump-

sum bonuses would appear to b~e more efficient than the

payment of installment bonuses. .There is another factor

which must be considered, however.AAlthough payment of

lump-sum bonuses increases the incentive to reenlist it also

decreases the incentive to complete the reenlistment

contract, thus leading to higher costs of contract

enforcement and collection of unfulfilled obligations.

'.' Recognizing the difficulties in recoupment of lump-sum

bonuses, Congress decided to partially return to the payment

of bonuses in installments"--h Fiscal year 1982, 50% of

the reenlistment bonus will e paid as a lump-sum while the

remainder will be paid i annual installments over the

reenlistment period. The economic efficiency of such a

policy depends upon the net effects of the two offsetting

costs -- the increased costs to the government of paying

deferred bonuses versus the decreased costs of contract

enforcement under the installment policy. Using our

discount rate estimates we are able to compute the implicit

tradeoffs made by Congress and the values assigned to these

r'espective costs.

iAnother Congressional rationale for returning toins Itall ment payments is referred to as the "what have youdone for me lately" syndrome -- a result of individualsserving extended obligations long after receiving theirbonus payments. Although somewhat misstated, the economicimplication is also one of difficulty in encouraging andenforcing contract completion.

*

"-3-

II. EFFECTS OF BONUSES ON REENLISTMENT RATES

Individuals enlisting in the military services sign an

employment or enlistment contract with an initial commitment

to serve between three and six years. The duration of the

contract depends upon the skill or military occupation they

enter and the initial training they receive. After

completing their initial contract, these individuals may

choose to reenlist for another period again ranging between

three and six years.

Since military pay and allowances are based only on

rank and experience and not on occupational skill, the

services have experienced shortages in certain skills with

high private sector demand (such as'nuclear p.3wer plant

operators and data system technicians) or in skills with

extrer.ely arduous working conditions (such as boiler

technicians). In order to offer more competitive wages in

these skills, the military 3ervices have been authorized to

pay bonuses to enlistees or reenlistees at various points in

their careers.

Pricr to April 1979, reenlistment bonuses were paid in

annual installments. At the beginning of each year of the

reenlistment period, the individual received a bonus payment

equal to the product of:

(1) monthly basic pay at the date of reenlistment, and

(2) a "bonus multiple" ranging between 0 and 6.

In the Navy, for example, approximately 100 skills are 'I

_ I1

-4-

assigned a bonus multiple at the beginning of each fiscal

year. These bonus multiples are determined on the basis of

the manning shortage or surplus anticipated in each skill.

Starting on April 1, 1979, the entire bonus has been

paid as a lump-sum at the date of reenlistment. Hence, the

bonus payment is equal to the product of:

(1) monthly basic pay at the date of reenlistment

(2) the bonus multiple

(3) length of reenlistment measured in years.

The undiscounted sum of the reenlistment bonus remains the

same for the two methods of payment, only the timing of

payments differs.

The reenlistment rate is defined as the fraction of

individuals who choose to reenlist (remain in the service)

among those whose initial enlistments expire within a given

fiscal year. To estimate the effect of bonuses on the

reenlistment rate, we pooled annual data over three fiscal

years (FY78-FYB0) for the 87 Navy skills having complete

data. Since the fiscal year begins on 1 October, the policy

change on 1 April 1979 occurred exactly half-way into FY79.

Hence our data set contains one year under the old policy,

one transition year, and one year under the new policy.

We expressed the reenlistment rate as a logistic

function of the bonus multiple and other variables; i.e.,

(1) log [!/(l-R)] - Xb + u

where R is a vector of reenlistment rates for each skill and

b is the vector of coefficients of the independent K

J -

-5-

variables. Because the disturbances in equation (1), u, are

heteroskedastic, we estimated equation (1) using weighted

least squares.1

We used two specifications for equation (1). Under the

first specification, we introduced a set of dummy variables

for the FY79 and FY80 observations. These dummy variables

were intended to capture the effects of time-dependent

variables other than bonus multiple which influenced

reenlistment decisions in our sample period. Under the

second specification, we replaced the dummy variables by the

unemployment rate among males aged 25-34. It should be

noted that this variable only took on three distinct values

in our data set, one for each fiscal year. Hence the two

specifications are quite similar, except that the second

constrainr the time-dependence to occur through the

unemployment rate. 2

Under both specifications, we included interaction

variables between the bonus multiple and fiscal year dummy

variables. These interaction variables enable us to

estimate the effects of bonuses on reenlistment rates A

iOur observations are not on individuals, but ratherare cell averages within each skill. Hence R is the sample Aproportion who reenlist within a skill. Cox (pp. 104-107)demonstrates that Var(ui) [NiRi(l-Ri)]-

where N1 is the cell size. Thus the weighted least squaresprocedure places more weight on the skills having greaternumbers Ni or having Ri closer to 0.5.

2t would be impossible to include both the fiscal year

dummy variables and the unemployment rate into the Wregression, since they would be perfectly collinear.

- t

separately under the two policy periods and the transition

period. Hence we may test the hypothesis that the policy

change in FY79 led to an increase in the effect of bonuses.

Our regression results are reported in Table I. .0

two specifications yield almost identical estimates of the

effects of bonuses. The interaction coefficients are both

positiver and the FY80 interaction coefficient is larger

than the FY79 interaction coefficient.

Specification 1 Specification 2

Intercept -. 7941 -. 8569

(5.98) (3.23)Bonus .1840 .1914

(3.98) (5.96)FY79 -. 0410 --(0.25) IFY80 -. 0166.-

(0.11) --Unemployment -. 0081

-- (0.16)Bonus x FY79 .0427 .0329

(0.69) (0.76)Bonus x FY80 .0747 .0664

(1.23) (1.38).343 .342

261 261

(t-ratios in parentheses)

Table 1 - Regression estimates of equation (1)..

Table 2 reports the partial deri'zatives of the

reenlistment rate with respect to the bonus multiple for

each of the three fiscal years. 1 Again, the two V

1Differentiating equation (1), we find that the partialderivative of R with respect to the F*Y78 bonus multipleequals the bonus coefficient times the factor R(1-R). Weevaluated R at its sample mean of .40 in computing this Hfactor. For FY79 and FY80 the result is similar, exceptthat the sum of the bonus coefficient and the appropriateinteraction coefficient must be multiplied by R(I-R).

A

4 - ----- ____ ___ _ _ ____ ___ ___ ___ ____ ___ ___ ____ ___ __ ____ ______ __

4"

-7-

Pcifications are nearly identical. We find that the

3installment bonuses paid in FY78 were only 71.0 percent as

Specification 1 Specification 2

FY78 .044 .046FY79 .054 .054FY;10 .062 .062

Table 2 - Partial derivatives of R (dR/dB).

effective as the lump-s~ira bonuses paid in FY80 under the

first specification, and 74.2 percent as effective under the

second specification. Since the first specification

provided a slightly better fit to the data, we will employ

71.0 percent as the relative effectiveness of installment

bonuses.

III. ESTIMATION OF THE IMPLIED DISCOUNT RATE

In comparing the effects of bonuses on the reenlistoient

rate in rvY78 and FY80, we have found that the installment

bonuses paid in FY79 were only 71.0 percent as effective as2

the lump-sum bonuses paid in FY80. In this section we

compute the discount rate of the "marginal" reenlistee

implied by this comparison. We assume that the nominal

discount rate is equal to the real discount rate plus the

expected rate of price inflation. We will first solve for

the nominal discount rate and then compute the real discount

rate by subtracting the observed rate of inflation, thusa, using the observed inflation rate as our estimate of the

expected rate. _34%

-8-

Consider an installment bonus of $1 per year. Assuming

a four-year length of reenlistment fthe majority of

reenlistments are for four year terms) 1 , the discounted

present value of the installment oonus at a nominal discount

rate of r is equal to:

4 1 l+r ( 1 1

(l+r) -r (+r) 4

By contrast, the discounted present value of a $4 lump-sum

bonus is simply $4. Hence at a nominal discount rate of r,

the effectiveness of an installment bonus relative to a

lump-sum bonus is equal to the ratio of the discounted

present values:

1+r 1

(3) 0.25 1- -

r (l+r) 4

To estimate the nominal discount rate, we set

equation (3) equal to .710 (the i.easured ratio of relative

£ bonus effectiveness) and solve for r. 2 We find that, on the

miargin, Navy enlisted personnel have a nominal discount rate

of 29.1 percent 3 . Since the Consumer Price Index grew at an

1 The length of reenlistment is chosen by the reenlisteeand will not be independent of the method of bonus payment.A paper is available from the authors discussing this issue.

2 This is a third-degree polynomial equation in r.However, applying Descartes' rule of signs, the solution isunique since the installment payments are all positive.

It is only a coincidence that the estimated nominaldiscount rate is approximately equal to (1-.71).

-7 -!

F . .. -- --- 1 I II --- -- --- -I '• -- - -lmi•-", ",,-- -I

"-9-

average annual rate of i0.6 percent over our sample period,

this implies a real marginal discount rate of 18.5 percent. 1

IV. THE EFFECTS OF PROGRESSIVE INCOME TAXATION

Our regression results indicate that installment

bonuses are about 71.0 percent as effective as lump-sum

bonuses, and from this we have inferred a real discount rate

of 18.5 percent. The situation is altered when we allow for

progrr:-sive income taxation. 2 The individual may still

prefe a lump-sum bonus since deferred income must be

discounted. However, progressive income taxation implies

that the individual's total tax payments are higher when all

of his income is earned in a single year. Hence the

individual's preference for a lump-sum bonus, and the

corresponding relative effectiveness of lump-sum bonuses in

inducing reenlistments, may be mitigated or even reversed.

Moreover, our estimates of the discount rate may be biased.

To explore this matter analytically, let t[I] denote

taxes as a function of income, where 0 < t'[I]<l andt"[I] > 0, so that taxation is progressive. Let B denote .

A the annual bonus installment, so that the lump-sum bonus is

equal to 4B. It is easily shown that the installment bonus

1This estimate assumes that reenlistees on the margin

perfectly anticipated future inflation. If in fact they2W goete osblt oricmaveaigutelunderestimated (overestimated) the actual inflation rate,

then the real discount rate is higher (lower) than ourestimate.

24We ignore the possibility of income averaging until

the next section.

S. . . .. . . . .. . I

-10-

increases the discounted present value of the individual's

after-tax income stream by

4 B (1 - t' [Ii + B]

(4) Y.i --i (1+r ) i- 1.

where I4 is taxable income excluding the bonus. By

contrast, the lump-sum bonus increases the individual's

present value by

(5) 4B (1-t' [I1 + 4B])

The individual will prefer the lump-sum bonus if r is

very large or if t" is very small (i.e., taxation is not

very progressive), or if income in the absense of a bonus

grows at a high rate. However, the individual will prefer

the installment bonus if r is very small or if t" is very

large or if income in the absense of a bonus grows slowly.

The relative effectiveness of installment bonuses is no

longer given by equation (3), but. is now equal to the ratio

of equations (4) and (5) or,

4 1- t'[Ii + B] 1

(6) .25 -- '

i=l 1 -t'[I + 4B] (l+r) i-l K

Equation (6) reduces to equation (3) for proportional

taxation (t" = 0), but exceeds equation (3) for t" > 0 and

may even exceed unity. Moreover, the discount rate implied

by setting equation (6) equal to the empirical ratio of 71.0percent will exceed the discount rate computed earlier using

equation (3). Hence our earlier estimate of 18.5 percent

----------- *- A- --- ----

I -l

represents a lower bound to the real discount rate. The

implied discount rate is larger when we consider progressive

income taxation because discounting now must be sufficiently

important not only to yield the empirical effectiveness of

lump-sum bonuses, but to do so in spite of the tax

disadvantage of lump-sum bonuses.

V. THE EFFECT OF INCOME AVERAGING

In the previous section, we demonstrated that our

estimate of 18.5 percent as the real discount rate

represents a lower bound under progressive taxation. In

this section, we consider the further effects of income

averaging. We will see that our estimate of 18.5 percent in

fact represents an upper bound when income averaging is

allowed.

The individual receiving a lump-sum bonus may choose to

income average in order to reduce his total tax payments.i:

Let 10 denote the sum of taxable income during the four

"base years" prior to the year in which the bonus is

received. Let I1 again denote taxable income excluding the

bonus during the year in which the bonus is received. Then

the tax liability under income averaging, ta, according to

Form 1040 Schedule G for 1980 is equal to

S(7) ta - t[.2410 + .20(1 1 +4.)3)1

+ 4(t[.24I0 + .20(II+4B)] - t[.3010)"

ýIII

I

-12-

Without the bonus, the individual 4ould have earned 1

- t[II] after tax. With the bonus, he earns I1 + 4B - ta"

Hence the net value of the bonus equals:

(8) 4B -(ta - t[Ii]

In this expression, the term 4B measures the lump-sum bonus

payment, while the term ta-t[Il] measures the additional tax

payments attributable to the bonus.

The relative effectiveness of installment bonuses is

now given by the ratio of equations (4) and (8) as

B 4 1 - t'[Ii + B](9 ) .-,1 ,

4B - (ta-t[ll]) i=l (l+r)i-

The form of equations (7) and (9) make it difficult to

solve analytically for the implied discount rate that sets

equation (9) equal to the empirical ratio of 71.0 percent.

Therefore, we have numerically analyzed equations (7) and

(9) to solve for the discount rate under a reasonable set of

assumptions.

Consider a married individual in the Navy who files

jointly 'with his spouse, who claims himself and his spouse

as exemptions, and who has no inco:.ie outside of military

pay. We assume this individual sailor faces the FY80 mean

bonus multiple of 1.57 if he reenlists in 1980. Table 3

lists the income stream (excluding any bonus) that such an

individual could have expected had he entered the Navy at

+ 7 7-7 7.. ..-

-13-

the start of 1976 and reenlisted during his fifth year in

1980. 1

Year Military Pay

76 $4,83077 5,20078 5,94079 6,77080 7,920

Table 3 - Income Stream of Typical Indiv'idual

Without the bonus, our individual would earn $7,920 in

1980 and pay federal income tax of $362.2 The lump-sum

bonus at a multiple of 1.57 equals $4,145. Without income

averaging, he would pay income tax of $1,076 on his total

income of $12,065. With income averaging, he would pay only

$972, a saving of $104.

Evaluating equation (8), the net value of the bonus

equals the bonus payment minus the additional tax payments

of $972 - $362 = $610. Hence the net value equals $4,145-

2F $610 = $3,535.

To evaluate the remaining elements in equation (9), we

must estimate Ii and hence t'[Ii + B] for i=1,2,3,4. We

have already calculated I1 = 7920. Since B = 1036, we may

compute from the 1980 tax table t'[I1 + B] - 0.16.

IWe assume a median promotion path. Hence the

individual i3 promoted to rank E-2 early in his first year,E-3 at the beginning of his second year, E-4 halfway. intohis third year, and E-5 halfway intQ his fifth year.

2We anore state and local income taxation. There areseveral states that Co not tax the military income ofservicemen. Resident.' is simple to change, and thisSinformation is well .•,cw A by service members.

-14-

We estimated Ii over the remaining three years of the

reenlistment term. For this purpose, we tracked the

individual through the rows of the FY81 military pay table

corresponding to the sixth, seventh, and eight years of

service. Our estimates of Ii and t'[Ii + B] from the tax

tables are noted in Table 4.

Year Military Pay Marginal Tax Rate

81 $ 8,970 0.1682 9r560 0.1683 10,540 0.18

Table 4 - Income Stream and Marginal Tax Rates.

Given these assumptions, equation (9) implies a nominal

discount rate of 27.4 percent and hence a real discount rate

of 16.8 percent. This estimate is very close to the

estimate 18.5 percent that ignored progressive taxation. It

appears that tax considerations make little difference in

the calculations as long as income averaging is allowed.

VI. POLICY IMPLICATIONS

We have estimated a real personal discount rate of

approximately 17 percent. It is unlikely that the

government uses a real discount rate as high as this and, in

fact, most previous estimates have been in the area of 10

percentl, Therefore, in the simple case, the net present

cost to the government is higher than the net present value

"to the bonus recipient and installment bonuses are

inefficient relative to lump-sum bonuses.

1 See Baumol (1968), Feldstein (1964), and Marglin(1963) for a detailed discussion of this issue.

-7.- - -

A -- -". - - w -

-15-.

The situation becomes considerably more complex,

however, with the introduction of contract enforcement

costs. A combination of lump-sum and deferred payments may

now be most efficient, depending not only on the relative

size of government and personal discount rates but also on

the relationship of enforcement costs to various bonus

payment schemes. Also, up front payment in a lump-sum may

have an adverse impact on incentives for performance.

To illustrate the magnitudes of the factors involved,

we compare a lump-sum policy to the current policy whereby

50% of the bonus is awarded up front and the remaining 50%

paid in equal annual installments. A lump-sum bonus of B1

will have cost to the government of B1 plus the present

value of the expected default costs denoted R. We define R

as the discounted expected value of the unamortized portion

of the bonus payment when the individual defaults prior to

the end of the contracted reenlistment period. 1

To keep the quantity and quality of reenlistments

°• cnstat, he sm, 2 Of the installment payments must have

a discounted present value equal to B1 when evaluated at the

sailor's discount rate of rs. This implies the relationship

+ I 125B2(10) B1 .5B2 2i-l (i+ rs~

.125B2 A

-. 5B2 + - 4r s (1 + rs)4

iWe have deliberately biased this analysis against thelump-sum bonus by assuming a zero default rate in the caseof the partial installment bonus.

-16-

The cost of the installment bonus must be evaluated at

the government discount rate of rg, and presumably rg < rs"

We may express this cost as

(11) .5B2 + .125B (1- 21rg (1 + rg) 4

Substituting (10) into (11), the cost of the installment

bonus equals

.125 1B1 [ .5 + - (1 - )]

(12) rq (I + r.)4

.125 1(.5 + - (1- )

rs (+ rs) 4

Recall that we have chosen B2 to keep the number of

reenlistments constant between the two payment schemes.

Hence, the combination of lump-sum and installment bonuses

will be socially efficient if and only if the cost to the

government in equation (12) is lower than the cost of the

lump-sum bonus, B1 + R. Therefore, the partial installment

bonus will be socially efficient if and only if

.125 1.5+ - (1- 4 )

r ( + rc)4 R( ) . .... < 1 + -

.125 1 B1.5 + - (1- ,)

r (1 + rs) 4

The left side of this expression equals 1.063 when

evaluated at r s - 17 percent and rg - 10 percent; therefore

the ratio R/Bl must be at least 0.063 to justify the partial

installment bonus plan. In other words, the cost of default

N-l

N'

-17-

must exceed 6.3% of total bonus program costs. Since we

have implicitly assumed a zero default rate for the partial

installment bonus, this value represents a lower bound to

the true value required to justify installment bonuses.

Actual data are available to analyze the optimality of

the installment bonus program. The total amount of new

bonus payments in FY79 was $174.2 million. A combination of

actual data (FY79 and FY80) and Defense Audit: Service

projections estimate that the undiscounted (and therefore

overest'mated) default costs will be approximately $4.0

million, or only 2.3% of lump-sum bonus cost. This is

barely one-third the default costs required to justify

installment bonus program enacted by Congress.

VII. CONCLUSIONS

We have compared the differential effects of lump-sum

and installment bonuses paid to Navy enlisted personnel.

From this comparison we estimate that, on the margin,

individuals have a real discount rate of about 17 percent.

This estimate takes account of both inflation and

progressive income taxation. We also find that tax

considerations make little difference in the calculations as

long 1-q individuals income average when they receive lump-

sum bonus payments.

In the simple case, as long as the government has a A.

real discount rate of less than 17 percent, it is more

efficient to pay lump-sum rather than installment bonuses to

increase the supply of military reenlistments. Returning to

_-_---__-_- . . .. . . . .. __..__= • _-,jj-- -. . . -- -- • •

-18-

a system of installment bonuses would reduce the number of

reenlistments for a given level of the bonus program budget

or, equivalently, increase the budget required to achieve a

given number of reenlistments.

When the costs of contract default are considered,

however, this preference for "up-front" payments may not be

optimal. The current bonus policy attempts to take this

into account somewhat arbitrarily by providing an equal mix

of current and deferred payments. Using our estimate of the

personal discount rate and a commonly used estimate of the

social discount rate we showed the minimum contract

enforcement costs necessary to justify this policy on

efficiency grounds. Recent evidence on contract

nonfulfillment shows that this policy places far more

emphasis on deferred payments than is warranted on

efficiency grounds. A return to lump-sum bonuses would

achieve the same number of reenlistments at a lower cost to

taxpayers.

I.

I !SI I• t 1

' * i- -

IL --~ -~---~ i li i - - - - -~

-19-

References

Baumol,, William,, "The Social Rate of Discount and theOptimal Rate of Investment,," Quarterly Journal of Economicsz,Vol. 77, Feb 1963.

Cox, David, The Analysis of Binary Data, London: Methuen,1970.

Feldstein, Martin, "The Social Time Preference Discount. Ratein Cost Benefit Analysis," Economic Journal,, Vol 74, June1964.

Friedman,, Milton, A Theory of the Conu tion Function, NewYork:NBER, 1957.

Heckman, James,, "A Life Cycle Model of Earnings, L.earning,and Consumption," Journal of Political Economy, Vol 84, Aug1976.-

Landsberger, Michael, "Consumer Diascount R-ates and theHorizon: New Evidence," Journal of Political Ecnoy Vol179, Nov 1971.

Marglin, Stephen., "The Social Rate of Discount avid theOptimal Rate of Investment," Qurel Journal of Economics,IVol. 77# Feb 1963.

,.:argling, Stephen, "The Opportunity Cost of PublicInvestment," Quarterly Journal of Econom~ics, Vol. 77, May1963.-

414IZia

PP 222PP 222MI~abi, Mql ., -a I" lsleting "we Cirotilar Ceefage "Iiziunl. NWbOl N., esraondence Mlels and Pe?11

frvseiofl,* 14 pp.. Feb 1978. MD A054 423 lrr.grals,' 30 pp., Jun 2978 tinvited pepwn prevsarts at tileaNS meting an %athemiicai Pi~Mbl In Feynmew's Pam

PP 212 Integrals,- Morsalle. Franc%~ Ner 22-M., 1978) (Ptb1111eftlge., MAIM 9001 Singular hOSCarIctaaic Iintial ValUe Mi Winger Velet; Lectu.- N~otes In "*Icaa, 106, (1979),Probeme W" Unique SaotIlon.6 20 pp.. Jun 1978. 234-233) AD AM'. 536AO AVAXM

PP 213 Oaspife. Naro, q9todweati "Kaciaics of Obleoaielon Iftlecil.Mmgai, Mmr-, 1qiU~Intvai in wite 1th bitipl* steady Awbectiots," 21 pp., Jun 1976. AD) A056 227States). Apliln-Aw.oR ft Landlest Equlo.tis 22 pp.,

Feb (Foecviei 4f 'Owe first Anawal Nokshqu an met PP 224lMwomilon Liftege Q.W..E wgid MMA-teW.Ic and Meanger. Marc. 4Ag~aetioa. Sit~ralo, dad Extinctionl Ininaatr,, 31a -1 M M Feb 7-42!If 197), MAD 071 472 Ext~leil" Mimk PopulatlcnsO,' 48 pp., Var 1976,

AO A058 53611P, Z14 Ofortiaf of "his lam ware ft.rted at the. Institute of

WePleled, Raliat Q., "rA 3~.t Oittwrtwt Vims of The Applied PNatl.O.tift and StatiStics, Unleralty of IrItlal.Optiftl k4"vI PMTXWq,. 37 .5p., ~. )O 97M L~foaeet at 1116 Coliebia, Vnwm.e 41.C., ZeAgia1973 C~arcwtlr of 1%e Afti-a POSITI.~ai Scloaa Aspeacta-ties (APW2A~'I Pooel an 00"Ii;g~ StrateIc Ilawiira~notsan ~4 PP 223l4111"~ Pm easha I, Wcaf. .11., Sept~e 2. 1I76) Mange,, Marc, Vciflatkisa, Roctubtiona, and tti HapiIV AM5 226 SII f~190.*, 03 pp., Jun 1M.7to ADM05 233

Prt~loas of "his so aea Xwptw" at The Institute oPP 219 1Ieiib Nather.'ie 04~ Staistics, nmsl ffiif

Coils. %%&aell C-, -CZLnta or: Principles of lnftroetlsan Cu00i8, VOOMMOe. CaONadaUnvrit.fD'ts

Motrievel by IMmfre Kelts4J,* 10 pp., OW 76 (Pubih~sid asWletsr to t.. Edltar, J~awael of Dowwrwtiatl , Vol. 31, AI 226ba 4, ri 210-,3011, Doomar tit"), AD AM5 436 Relaton, J. N. and P. W. Mass,- -r~rtr W4 Current

Oesgandane of OWe Iadan le Wadsitting Wd LWs.0 34 pp..PP 216 Joen 1973 MPbiAC td in Jaiwael of fvgl led Pbpale, 50, 630,

Celi*. flas"ll C.. "Lomeas tlfpsnc Olstwlriaetlo at Kry 1979) AD 4MN 531Scientif ic Produtivity., Is pp., FO~ 1976 CP016fled In the Oft I Tal *Owns Loratw Ias, Inc.JawIrl of Ine I vroun Saclsty forn I formal os 34latm.Vol. 26, w%. 6, ps.. 346-3M0, Noeasr 1977, ; MADP4 423 pp 22

-hi NWpl. OMm *Wlfwo ement of Fiatiwtlanat tO~ lwiPPD 217 PolntS,' 50 pp., Ney 13711, At) AM6639

Coll#,. %40.1a~ C.., I~fiIw! IC studies of SclentifIcvtadwct1vI-,y,* 17 pW. MA' 73 sPnawte at ITh Mental pp 229Blin San Frik~lcloo, California, Oct-aber 1976), j AD 09 442 and Stodletl Theerry U4 pp.. Jun 11178, AD MM5 54

PfI 223 - Clawalyled pp 229i~tin rtwi

pp 210Forwilstlen and El"aInt-SMIWn~eaMMK ApproslmtfftlR,Huftl Inger. ft. Later, 'bftet Meis[Vsa .4". Nattoaml tdE%.- 50 pp., Ps. 1976, AD AOS 542tatIORa: Theory and W£atlmin, 60 pp.- P 76. AO A054 422

PP 230pp22 M anal, Marc. "Diffuaion Thew'p of Rewtacon Pata", It

"Mrim 00ne10 G. -vidpulasati.o by 'am" wbri2ac, 0mnsatn-thlftlbedt MAp etln, 34 pp., Feb 1976,X p.Apr 70, V A04 43 A M5

mit"Vt . 1" pw wa aaupmw alaqminp the ;: Li eradP., Jr..*"@ actaf Fo sTe Cs

In mpoer s f* fdt~g 2Political wiltytt and1111"y Cnitrant, ikel~ftlifonPa"er "o 41, faw It p PP23Hillsme Lom; W PAl~iate,2973) MD Mel 964 J41000601 Lowls, '*Plo ~stsIeNf t~~*t

and Lcwr Rtetlatos ftal..) ADI AM1 52

~ mefeeelona Ft.et ani X ftnaber Oft :s obtained freb mas Naflen TOINSlCal Inkfwtlam Sarvict U.S. Deapulmt of3-- ýwe. ringf lad, Virginia 22152, Otter papers areI belsf irw . ma Iboamet tnformtIgm Ofties, Center far "ev*l

Me~lyase, 'M f 1 Ne1rohn I Street, Alemneria, Virginia 2WI1. An 2*66 of Selecte rolicatlet ns i 19 M9il3610 anreppue. TAO ao 2n--k~ 1n1de Listing of Professional Papert withu wtsrcftg ss 5ued iroa 1909 to Jims 2941.

- ~04

a LiE

PP 233 PP 249Jacobison, Loils, *M Altsrnolý* E£9lenaon of fthe Cyclical Gleesei, Kenneth S.. 'A 2acretary Problem willi a ftan~Pella of Qults023 pp., Sep 1978 NimbwrofOwlces,* 23 pp., Mor 979

PP 234 - Nov I ad pp 290Jcnaw. Jones mad Levy, Robot IA., Mmoe Federal Expendl- M~engel. Mac, %bftiltg Fluctuations Mn Macrocopic Sys-tars Displaeo Staft end Local Expendlturer The Cow. of teen,0 26 pp., Jun 1979Construction Grents,* 29 pp., Oct 1979. AD AD01 52 P 5

pp 239 Trost. floort P., 'The Ws laiftton and Int arprett in ofMlwafl. Mourlom V 'The Semicleassical CExpnsion of the Seisrl Selectivity Mfodels,' 37 pu., Jun 1979, AD AM7 941Amberoic-09cillauw. I~ropapgatrs' 41 pp., Oct I9M (Pub-haled Io Jownsl of NettomIen tie PmysIcs 20 (1919) pp. 644- PP 232633), AC AOGI 336 10iAM, Walter ft., 4os0tlon Finding EIltI Prior Kawtledga of

Coverlenom Parawlerss, 5 pp*, Jun Mil, (Published In IEEEPP 237 Treectlaume an Aerospace & Eloctroni c Sy.t , Val. AIES-13

Hearer, Donald. *A Maotrix Criterion for no I5 lategral No. 3, Nor 1979geemse,' 10 pp., ian I9M (Puiblelod In the. Il~litla, Journalof Notienotlics, Vol. 22 (197M) pp. 972-4141 pp 233

Cleaerw, Kmee'dl S..*Th '7i. -ho~ca Secreary P~onies,'PP 236 32 pp..ion 1979.,AD A075225

Utgf.f, Kathleen Classes, 'Unemployment Ineareece and ThMEmpioymnt Rate,* 20 mp, Oct IM7 (Presented at the COm*- PP 294foreane on Economic Indicators sad Porforience The Dorsent 1116,0l, Moc en QWenbe. David 16. -Integration of aOl1mm Facing government end Maltness Leeders, pieeented by Alierla% N I~u Ove so Offset Clrcl*.* 14 pp., Jun 1979,India" n. v I yrIty Gadveate School of balositna). AC 3061 927 AC 3096 471

pp 239 PP 259 - Classified. AC 1101 "I1LTroe. ft. P. end Marie, J. 1.. *The Effect of MI IIts-,Occupationel Training on Civilian Earnings: An Incom pp 296Selectivity 4goem, X pp., Nov 1979k 1 AC n &0731 Meano, Donald E.. -felng Personnel Distribution Models.1

27 pp.. Feb I.16, 0 A0112 216

or 247Paosit, growe, 'Onto of Iftu 0ntor for Nevel Analyse,' PP 297I13 pp., Dec 1976, AD A063 739 Thaler. Rt.. 1019ow"~Ing and Fiscal Ocnatralnts: thy Dis-

c"!tin, Is Always RItght,' 10 pp., Avg 1979. AD A073 224pp 241

NOWge, Morc. 'Yloctuaetlaft at Cheem~al lP*?bllltI*ai, PP 23324 pp., Doe 197 tPblasbod na i2omai of Cneadcal Physics, Mengel. Marc S. end Thoes, Jaes A., Jr.. w*naytiCalVol. 6p, No. 8. w is, 1978). /C AO0 737 IftheIn Sewarhl~heoy,* 6 pp., fowu1979, AD Aon 32

pp 242 ep 23Stqepson Willive Rt., MOh Analysis at Cynasiceily Inter- Glaes, David V.. Mae. lh-OhIngz wann, Walter f. end Perln,scl~iw Wates t~ Air Cawt by tMu Nomers),* 160 pý., David A., 6A Clees o4 Coomirtstin Mwhrov Matrices,' 17 pp.,Dec 1976. AD A063 760 Nov 197), AC AD"? 333

pp 243 pp 260Statism, Willian ft. 'A Probailistic Formulation of Marhy Ifogel, More S. and 0*%. Davls K., 'Dotectlon Re" andOynanlcs as Applaed io Iftu Analysis of Cporwtlomal PnaewwV Sweep width lin Vlsual Search.E 14 p;6, ?--w 1979, AC Am7 a34Problems.' 16 pp., Doec 1976, AC A063 761

PP 261

PP244 AlnadHooit talyA. 1mnoefe ot Vile. Carlos L.; Z-ljoc, David J. end Maes. Jam., 'Franr-90e15eAlln ad ltos~i, taney ., IfoltannosOcas Condom Theory of Oua04cal Dynamics, Vt. Angular Olstrlb,'

Of COMPaU IEAMIlpeset, 20 pp., DeC 197S (P~a~llaild Sly The ttome of Reaction Proeects,* 14 pp., lov 1979 (AmprintedAmeican Society of Newel EngIneers. kwa Engineers from Jomal Chemical Phly. 70(12), 13 Jus 1979),Jw ne1. Val. 91, No- 4, 0.0ITO 1979 AC071 473 AC 3076 237

pp 26S PP 262Simpson, William ft., 'Th Acelargager Net11 of Obtaining Petervsx Owelee C.. 'ThIrd world Mlhitary Elitesm In SciletAlricall Panris a trea I'liit Test Dafta (OyuianI: Per- Precve'90 pp., May 1971r, AC AC?? 339lammes TentIng), 403 pp., Jo Mg.9 AC A073 226

MY263pp 244 Atoblean Ka"~ I.. 'UsIng Comrcial Tommers end Conta new-

3rlnFrees 'Layoffs end thomloysmny I w,.raece,* 35 ships for Navy Unsorway liepiealshmnt.0 25 pp., Nov 1979,p. , Fab 97M (Preesented at Iftu Now Chntam on 9.oM AC A077 SMlft Labo Miets,' chlcep, Ja 1070), 1 AC " 30629

pp 24Thoma, Sinms A-. Jr. "tha Trsaepov-. Prerw!too of DIltuft

eSem tn Applied Vilald," 143 PP., Mr 1979 AM09 444

PP 264 PP 277WeInland, Flrt 0.. 'Th. U.S. Navy In thle Pacifict Post, Mengel. Marc, NSeall Fluctuations in System wifth MultiplePresent, and GlIi.ps of thle Future.' 31 pe., Noy 1979 Limit Cycles,* It pp., Hae* 1IM (Publiseild In SIMG J. Appl.(Deliwaed at tile International Symposium an til See, Meth., Vol. 36, No. 1. Feb 10K~) AD A016 229sponsored by thle lnteraloal Institute for StrategicStudies, The GroakIngs Institution and the Yomiuri Siilabun. PP 278Tokyo, i6-20 Oct i97) ADto" a0637 MI~raIh, Mmurice, 'A Targeting Probiems Exact we. Expeted-

Value Approach* ,' 23 pp.. Apr 1960, AD A063 09rp 263

Weiniald, ebert Q., 'War and Peace In Ifil Nor t: Soam pp 279ftlitical liplcaionfea of tile Changing Millt"r Situation In wait, Steph~en N., 'Cousal inferences end tile Us. of Force:t ANorthern Eurape, 1S pp., Mow 1979 (Prepared for Critique of corcm Without War.* 50 pp., Noy 1960,preeentation to tile Conference of tile Nordic Balance In AD ADII$ 097Perapeatiwe: The Cteeing M41litary and Pol itical Situation#*Center far Strategic ead internationai Sttoles, Oarorgecn PP 280Unlwaraity. Jun IS-IS, 1978) AD A077 836 Goldberg, Leerance, 'Estimation of tile Effects of A Ship's

Stewning on tile Failure Amta of Its Equlpmnt: An Appl ice-pp 266 tion of Econointric Analysis#* 23 pp., Apr I96, AD ADO5 09

Utgaff, Ualiy Cleassm, end grenling, frc*.i 'Texas andinflation,* 23 pp., NGew IiS, to A0111 194 pp 201

Nizrahlu, Maurice M.. 'Co~nwt on 'OlacretIzatlom Problem ofpp 267 Funtctionl Integrals In Phase Space','1 2 PP.. May IM6,

Trost, .. tert P., and Vogel, Vobert C., 'lbe tespons Of publaished In 'P"ilylal Revlew D', Vol. 22 £196),State Gaveamwtý Receipts fto Economic Fluctuetione and tle' AD AO94 994Allocation of Pna-Ccl a v ewln Sharing iGranta,'12 pp., Doc I97 (Reprinted fro, tile Rlewiem of Economics act PP 283Statistics. Vol. LXl, INa 3. Aeguet IM7) Oemimhes, Bradford, 'Expected '2amend for thle U.3. Navy to

Serv'e as An Instrument of U.S Foreign Policy: ThinkingPP 263 About Political and Military Environmental Factors,' 30 pp.,

Thowson, Jews S., 'Seaport Dependence and Inter-State Apr IM6, AD ADIS 059Cooperation: The Caee of S46-Saiueran Africa,* 141 pp..

Jan 19110, AD A61 193 PP9234J. Keleson.* W. iftnn, and U. Sumite." '11, Laguare Trano-

Veils. Kewneti 0., 'lb. Sowiet Involvement In tile 094Me. *Ths Gradiaft Scilcol of MNtaflgmant, Univwslaty of ReoneeterWmr,' 42 pp.. Joe 1960 (Presented at ftie Southern Conferenca and thle Canter for Navl Aealyseseom Slavic Studies In October. 1979). AD 0902 211 "'The gadute $,-cIco of Nanegenent, Uaniversity, of fechestmur

pp 270 PP9289

Ilaeaeft Richar, ' Soviet Policy In tile Norn of Africa: The Rlaamk. lichard 3., 'Suparpome Security interests In tile

'The Soviet Union In tile Third World: $soees" or Failure.*"a. W .110110t R. Coneldeon, NostyldW Pyles, 11o4lder, C4.. PP 266Sumer 1960). 10 A061 195 Mirrahi, Naurice 34. G Itil 0S Appoximation to tile

Propagator for Arbitrary Namlitoniana,' 23 pp*, Aug 196PP 271 (Publiseftd In Journal of Na". Pilys, 220I) Je. 1961),

Mc~onnell, Jamn, 'S~wiet and Aesrican Strategic Does, Inest .12 A09i 307One Wire TIm,' 43 pp.. jam I96, AD ADS91192 p 6

PP 272 Cope, COwls, 'Limit Cycle Souiu$lone of iReaction-OIffuslonWells, Kenneth 0.. *TeAzores In Diplomacy and Strategy, Eguatlana,' 39 pp.. Jun 1960. AD A067 1141940-1945, 46 pp., Xer 1960, AD AM6 094

pp 20PPPm 27 olums aNlter. 'Can't Let your SI Ides FIp IVYou: A Painless

lhadet Michael K., 'Labor Sopply of Wlvwe win, Ra~bende (Wlde to Vlauaie That Really Aid,' 26 pp., (revlsAqd-Employed Eltiler Full Ties or Pert Tine,' 39 pp., Nar 1iW0, Aug I96), Ao AM9 732ADA11102220

pp 299pp 1.4 PReb49neon Jack, 'Adagufte Classification OeldancI - A

Mean. Olte Rt., 'A Rteewt In thle Theory of Solral Sopci, Solvtlon and a Problem#* 7 pp** Aug 1940p AD A091 212

*~~p p2,9r0w

pp 273 Matson, Grgory 14., 'Eveluatlon of Comte Softwar In eneoldberg, Lawrtsca, "locultera Advertiaing weg Navy Enliat- Opwv~tlonoi InVirowmnt, 17 pp.. Aug 19110, AD A091 213

mets.4 34 pp. War 1911A, AD AM3 221pp9 291

PP 2741 Naddela .G65.' and Trost. Rt. P.. 'Sawm Extensions of tileeildberfl. Lowent*, 'Celaying; en Overhaasi end Shipts Equip- Marloft Proe" Nadal,' 17 pp.. Oct iwo,) AD A091 946 41

mast.' 40 pm. Nay 1960,10.A AM6 093 OUalwersIty of Florida

PP 292 PP 305Thorns, Jon* A., Jr., MTe Transport Properties of Binary Nunr, Laura N., 'A~i Introdiction to the Literature of SearchGas Mixtures In Appie~d Wagnetic Fields," 10 pp., Sept 1980 Theary," 32 pp., Jun 1961(PubI Ished I n .Jounaie of Coiastis Physics 72(10).15 miry 1960 PP 306

Anger, thoses E., 'ldhat Good Are Warftar Padams?" 7 pp.,PP 293 Ny& 1961

Thorns. Jacas A., Jr., "Evaluation of Kinetic Theory Colli-sion Integrals timing the Bmnelraiid Rhaso hifft tppromch," PP 30712 pp., Sept 1990 CPrlntad In Journal of Cheaical Physics Thowmon Jawr ipendence, Risk, and 'Ailnearbl~lty,"72610), 15 Pay 1980 43 pp., Jun 1981

PP 294 pp 106Rotwets. Stephen S., 'French Nvew Pliiq Oil-side of Mizrahi, MN.., "Correspondence lules and Path Integrals,-Europe," 30 pp., Sept 1980 (Praesented at the (onfcrencc of Jul 1981. Publ ished In *t~uov* Cmanto 8-, Vol. 61 (1981)tl~e Section an Military Srudles, International 9ýudletsAssociation Klatsh Island, S.C.), AD 1091 306 PP 309

Walniend. Robert G.. "An (The?) Explanation of the SovietPP 293 Invaeion of Afg~anlstan,' " pp.. "a 1961

Rober-ts, Stephe~n S., "An Indicator of Informal Empire:Peatterns of U.S. Navy Ouuisi ng on Over**"s Stations, 1869-. PP 3101897, " 40 pp. , Sept 1980 (Presentel at Fm rtu NavalI H1 stary Stanford, Janette M. and Tel To Wu,* *A Predictive MethodSyummoahm, US Naval Acaraf,. 26 October 1979, AD 1091 316 for Determining Popsible Three-dianloo onsaI Foldings of

lmunoglotulln Backbones Around Antibodl Comblning Sites,*PP 296 19 pp., Jkun 1981 (Published In 1. theor. elol. (198') 68,

Olsaules, Stadford and Pal-arson, Charles C., "Meritimms 421-439i'aci'ort Affecting iberlan %arlty.l (Fact-or** 1marltimos Cue ONortfwestorn University, Evanston, ILAfectan La Seairidad Ibeica) 14 pp., Oct 1980, AD A092 733

PP 311PP 297 - Classifled Bawer. Iftrlanna, Brhl ing, Frank P. R., and Uti-fl,

Kathleen P. Classan. *An Evealuation of Ut Funds." 13 pp.,pp 298 May 1901 (Published In National Coamission on Uneiwpatent

Nlzrahi, Maurice N., "A Marlov Ppproach to Uarge Missleo Compensation's IUeeplmmitz nt Coopensarion: Stud~ies andAt-tacks." 31 pp., Jan 1961, AZ IM.6,59 Rosearch, Volt.", 2, July 1980

PP 299 PP 312Jondr., iames N. and Levy, ikabert A., "Wage Leadership In Jondro. James; Boees. Marianne and Levy. R~ert, MTeConstruution. 19 pp., Jan 1981, AD0/094 797 Cýtiuw Speed Limit.- 23 pp., May 1961

PP 300 PP 313Jondroo, Jams and Scheldt. Peter,* INv the Estlantion of Roberts. 4tophon S., MTe U.S. Navy In the 19ft,* 36 pp.,Technical Inefficiency In til Shochastic R ontifar Production Jul 1981Function Mo~del.- I I pp., Jan 1.961, AD A.096 160"mMichigan State UnivesIty PP 3t4

PP 301Jehn, Oiristcpher; Horwitz, 9tanley A. end Loduman,

PP do.Jw .;Lv.Fbr A n uhs Clare %bbet F., "Exainlng the Draft Ostate," 20 pp., Jul 1981

*Technical Change and Expi oymnt I n Steel,. Attos, Aliualnum, pp 315and !ron 0-a, 17 pp. . Mar 1981. AD A099 394 Suck, Ralph V., Cqpt., %Lq Catastrophe y r other

nomw-.." 4 pp., Jul 19F1pp 302

Jondrat. .imse N. and Levy, Roubert A., MTe Effect of PP 316Impw I on E~jyiomnt (hider Fational Eupectatlons," 19 pp., Aober-ts, 9rephen S.. "'dastern Eurupeasn and NATO Navies,Apr 1981. P0 PA099 392 1980," 20 pp., Aug 1981

PP 303 PP 317Thomson, James, MTe Rlerest C~o~dlty In thue Goeilig Roberts. Stephen S.. 'Suporposer Naval Crisis Management InResource Wars," 3 pp., Aug 1981 (Pubiiahed In the Washington the Oaditerreanean" 33 pp., Aug 1981Star, April 13, 1981)

PP 318PP 304 Veg% MI Ion N., M~goelevia and the Soviet Pollq, of Force

Duffy. M1icheel K.; Greenuocod Michael J. * and Ne~owel 11, John In the Maulitarranean Since 1961," 1117 pp.. Aug 1961M.."t "A 0,oss-Soctional tbdei of Annual intart-*gions$'111ation end Eplq~eent Orol-lu Intel 1 oral Evidence of PP 339Strucwuai Change, 1"86-1975," 31 pp., Apr 1981, AD P099 393 Sailh, Michael W.. "Antlair Wierfant Defense of Ships at*Iuniverslly of Colorado Sea." 46 pp.. Sep 1981 (1his talk was delivered at the Navai"**rizuua State University Warfare System snd Tochnologr Conferencem of the Ammrican

Institute of Aeronautics anid Patronamrtics In Waehingtonl onDecember 12. 1980; In Boston on January 20. 1981; and In LosAngeleson Jue12. 1961.)

-4

PP 320 pp 335Trost, R. P.; LurlI% Ph Illp end Berger, Eebarq, 'A Note on Iee., UrngFol, G.S. Maddals. and R. P. Trost, *A*e)totlcEstimating Cotinuous Tica Decision lbdols:* 15 pp., (lovrlance Matrices of Two-Saoge Problt and Two-tags Tobitsvp 1981 Motho&. for Simultaneos Equations Models wlith Selectivity."

13 pp.. Jan 1982. (Pthllshed In Eccnontrica, Vol. 48, No. 2PP 31ff~I a " (Morch. 1980)

lbxlo R,,,dr RgionEcoomis,* 4 p.. Sp 181 ONeilThorns, ftbillty Fuels for the Navy.* 13 pp.,QAasoclate Professor of Economics. Arlsns *at* Iwiversity, Jan 1982. (Accepted for pubilcatlon In Novel Institute

Tooe AZ. PI-ocesdl ngs)

PP 322 PP 337Werner, John T., "Issues Iii Nsvy Idanpaver Researdi and Werner. John T. jn1 Goldberg, Matthew S., 11he Influence of

ol Icyt An Ecnomist's I~rspectlvo,8 66 pp,., Dec 1981 ?ton-Peacmiar', rectorsa on Loaor Sipply,* 23 pp., Dec 1981

PP 323 PP 339Bass%~ Frdorldt M., 'Gneration of (Llrolated Log-fNorab w bIlo#, Dexmond P., 'The Persian Gulf end the NationalSequences tI" the SIoujlstion of Clutter A ,es.' 33 pr., Dec interest,' 11 pp., Feb I9M1981

PP 341PP 326 ul hlp Tront, R.r P.. an d Voe RBerger., Edwr~dl'Alanwtho

Horcots, Stephey h.. "Olestiyng Suren end NR eadiness, Pfor nled in Muls- tilen $ell Duration Deta. Tic pp .~6pp.," 27 pp., Jpublise I9 n Deudlnse 6ia n pp.,al Feb190

PP 327PP 342

PP ou 326 l Trost, Rober an au ei ,~Le ug P.l lhddia V.gS, RoenTrotC, O.P., 'TetIon witRobets. tephe S ., oMelsl f en European adln PrOogras, foore 0tr estwalocione by D1Thsinw Sw tding Data lTw Maneo19is0r 7 pt ., Coss, 192 Sptlr190Euations.' al, 6 pp., Feb 198 2

PP 328 PP 343

*rsemlncn a Ibb rt ls of, Nrt avy tptuarw heir trotgl foder Stru turl (hng Sby DPrtojectIng Swth hin Na imuEltaedorceDsruptifcnCost,' 127 pp. Mac 1980 Leveal,' oel, 9 pp.. Feb 1982

PP 329 P~P 344Wobngei, lbrt rpp , ledNorthetrins vetd Naalr Sfr"tos, Foldtoer, Mee t S., 'Projectdingy ofLhe Navd REnlistednt."40nfc~ .pp. pp., 1982 (e 0FLove 13 p. No 198p. Fb19

PP 330 PP 344

LodussN Flbbert F.. *Alt ernative Approaches to Attrition Utgoff, Kaethy and Thaler. ODick "The Economics of Multi YearIbnegemont.0 30 pp.. Jan 1982 ontracting," 47 pp., NIbr 1982. (Aresonted at the 1982

Annual Keating of the Puol Ic Chioce Society. Sen Antonil6PP 331 Texas. Morch 5-7. 1982)

obberts, Stephen S., 'The lurkish Straits and rhe SovietNavy In the Meditorranean,* 15 pp.. Iftr 1982 (published In PP 346Navy International) Ibstiker, Bernard, *Sel ct lve ServI ce and the Al I-Voluateer

Force.' 23 pp.. MaIr 1982PP 332

Jell; Chrisiqcher, "The ROF and Amphibious Warta,%* 36 pp., PP 347Mar 1982 MrConrall. Jams. N.. *A Possible Counterlorce Role for the

Typhoon,* 24 pp., Nor 1982PP 333

Lee. Lzing.Fai and Trost. Robert P., "Ettiestlon of Sam PP 348Limited Dependent variable Models with Application to Jondrow. Jess. Trost, Robert. *M LEpirlosI Stud~y ofHousing Deeend, 26 pp., Jan 1982. (Published In Journal of Production lPeefflclancy In the Presence of C-u-era-In-The-Eronomatl. cs 8 (1978) 357-38121 Variable$,* 14 pp., Feb 1982

PP 334 PP 349(enry. Lowrence W.. Leo, Lung-Fal, Iaddeliý G. S.. and Trost W. N. Bteduenridge. 0. KIm MalIrn, 'CiIslaional intra-R. P., 'blturns to GaIlege Eduction: An Investigation of multlplet ftlaxstion of Cd(5S~p3p 1I 2) by Alkaen; HydraSelf-SelectIon Blas Based en the Project Talent Oat%" 15 carbons,' 7 pp., Jul 19181. (Publistlei (n Jojrnal a ofCusicalpp.. Jan 1992. (I'Wllshod In International Economic Review. Prytlcs, 76(4), 15 Feb 1982)Vol. 20, No. 3. October 1979)

tI

-r ------ ~ ii

PP 350Lavin, ware, "A Method for Increasing the Firepmer ofVirginia Class Cruistrs,* 10 pp,. Apr 1982. (To "l publishedIn U.S. Navel Institute Prnceedings)

PP 351Coutra, S. E.; Sti~nford. J. M.; Novis, J. 0.6; Stevens.P. W.; Wi, T. T., "Possible Three-Oimonsional BectboneFoiling Around Aitibodv Combning Site of ImmunoglcbaiinlWWC 167.0 1S pp.. Apr 1982. (Published In Journal ofTheoretical BioogiO)

PP 352Barfoot, C. Bernard, -Aggregation of Conditional AbsorbingMorkov Chains,* 7 pp., Jun* 1982 (P eeted to thi Six3'hEuropean Meeting an Cybernetics and $Qtfts Pleeerch. heldat "a' University of Vienna, Apr 1982.)

PP 353Berfoot, C. Bernard. -Son WAthemetlcai Methods for Model ingthe Pertoruenc of a Distributed Doat I*e " tt, Is pp/.June 1982. (Presented to the International Working Con-fernece on Model P401111a, hold at Bad Honheh,, Nast Gervany,Apr 1982.)

PP 354Hall, John y., 'I~y the Short-4lar Scener~o Is Woog forNavel Planning," 6 pp., Jun 1982.

PP 356Cylke, Steven; Goldberg, Motthow S.; Hogen. Paul; Nairs,Lee; 'Estimntion of the Personal Discount Wat: Evidencofrom Mi litary Reenlistment Decisions,- 19 pp., Apr 1982.

PP 357Goldberg. Vatthew S.. CscWiminatloW4 "epotism, endLong-flaa Wae" Differentials,- 13 gp.. Soo 1982. (Pt~lisnedIn Quarterly Journal of Econoilcs, Way 1982)

PP 358Alkst, Ceorge, "Evaluating Tactical Comnd And Control"jltem-A Three-Tiered Approach," 12 pp.. Sop 1942.

PP 361Quenbe*, David B.. -W-hods for Senertln9 AlircraftTrojqct~jrles,- 51 tp., Sep 1982.

PP 362Horooltz, Stanil~ A., -'is te Military Budget LUT ofBalance?," 10 pp., Seo 1982.

PP 363Warcui,. A.' J., 'Personnel2 Substitution end NayAviationleedines'. 35 pp., Oct 1982

-8V*