College and Research Libraries M A T T R O B E R T S Guards, Turnstiles, Electronic Devices, and the Illusion of Security Based on evidence indicated by a four-year study of book thefts, the author offers hypotheses to show the conditions under which books will be stolen and theories to explain why they are stolen. He proposes the theory that scholastic pressure, resulting in high potential use of a collection by persons granted access but denied borrowing privileges, will result in a high rate of loss. The paper concludes with explanations as to why current methods of exit control are ineffective, and what librarians might do to alleviate the problem. T H E P U R P O S E of this paper is to examine t h e problem of book thefts—the volume, rate, and cost of thefts in a c a d e m i c li- braries, as indicated b y a study of one library over a period of some four years. This, however, is b u t one purpose, for besides presenting t h e grim and dismal facts, the author hopes to establish a relationship between thefts and some of t h e factors which m a y influence losses, such as r a t e of growth of the collection, intensity of use, location of t h e library, and borrowing policies. T h e theories and hypotheses presented will b e t h e foun- dation upon which attempts will b e m a d e to fathom t h e motivation of book stealing and to construct a general theory of thievery. Theories, hypotheses, and even opinions, then, there will b e ; but when all is said a n d done, the funda- mental significance of t h e problem does not rest for belief or doubt on opinion b u t rather on fact, or, failing that, at least on t h e distinct probability that a c a d e m c libraries n o t only lose books b u t lose them in considerable numbers. Mr. Roberts is chief, Circulation De- partment, John M. Olin Library of Wash- ington University, St. Louis, Missouri. I t may surprise, or it may dismay, the reader that this should b e the raison d'etre of an entire essay. " E v e r y b o d y knows that books are stolen!" it may b e said, perhaps in derision. B u t in f a c t does everyone understand the true na- ture of t h e problem? Perhaps not. F o r in w h a t volume, or at what r a t e are books stolen? Are the losses increasing, decreasing, or constant over a period of time? 1 W h a t connection, if any, is t h e r e b e t w e e n losses and the number of dupli- c a t e copies available? W h a t is t h e cost of book thefts? And, finally, if everybody does understand the problem, why has so little b e e n done about it? T h e ques- tions that may b e asked are legion; t h e answers, unfortunately, seem to b e in short supply. T h e literature is not lack- ing in comments on security in general and book thefts in particular, b u t it seems to b e notably lacking in what could b e called a "control" study. Insofar as such a study of book thefts and li- 1 A t l e a s t o n e w r i t e r h a s e x p r e s s e d t h e o p i n i o n t h a t t h e f t s r e p r e s e n t s o m e t h i n g l e s s o f a p r o b l e m t o d a y ( w r i t t e n i n 1 9 5 6 ) t h a n i n y e a r s p a s t , p r i m a r i l y b e - c a u s e o f m a s s p r o d u c t i o n o f b o o k s a n d r e l a x a t i o n o f l e n d i n g r e g u l a t i o n s . S e e R o l l a n d E . S t e v e n s , " L o s s of B o o k s a n d L i b r a r y O w n e r s h i p M a r k s , " i n CRL, X V I I ( N o v e m b e r 1 9 5 6 ) , 4 9 4 . / 2 5 9 260 / College b- Research Libraries • March 1968 brary security is at all possible, this is an attempt to fill that gap. T h e need for real and factual evidence is obvious, at least to this writer. W h a t is less obvious is the apparent paucity of statistical evidence.2 T h e r e are, one might propose, three basic reasons for this. In the first place, there is a natural hesitancy to attempt to discover the true extent of losses, not only because of the possible traumatic effect, but because when the truth becomes known some- thing will have to b e done about it. Sec- ond, some librarians feel that nothing can b e done about it, that losses are an inescapable condition of open stacks and free access.3 And third, the prevail- ing opinion seems to be that inventories are too expensive, that the cost of such an undertaking is far greater than the monetary loss in books. This harks back to reason one. But is it true? I t is difficult at best to equate book losses, especially of out of print books, with anything, even time and reader inconvenience. Doubtless a year- ly inventory of an entire collection is out of the question; but research in such depth is hardly necessary to determine with a great degree of accuracy what the over-all losses are, the factors influ- encing these losses, and the direction in which one must proceed in order to con- trol them. One can discover just what is to b e expected throughout the collec- tion, and so determine rate, volume, and cost by means of a fairly small sample. After many hours conducting the pres- ent study, the author found to his amaze- 2 S t a t i s t i c a l e v i d e n c e is n o t e n t i r e l y l a c k i n g . T h e B r o o k l y n p u b l i c library h a s r e p o r t e d a p p a l l i n g losses. S e e : " B r o o k l y n T a k e s A c t i o n o n ' S t a g g e r i n g ' B o o k L o s s e s , " i n Library Journal, L X X X V I I ( J u l y 1 9 6 2 ) , 2 5 0 9 . 3 O p e n s t a c k s t o all, b u t f r e e a c c e s s t o w h a t ? T h a t is t h e q u e s t i o n . W h a t is t h e great v i r t u e of f r e e a c c e s s if t h e d e s i r e d b o o k c a n n o t b e f o u n d ? A c c e s s a n d a v a i l a b i l i t y m u s t g o h a n d i n h a n d , e l s e t h e v a l u e of o p e n s t a c k s is g r e a t l y d i m i n i s h e d . A n d b e s i d e s , w h o b e n e f i t s if t h e library m u s t s p e n d p a r t of its b o o k f u n d s f o r r e p l a c e m e n t s ? S u c h q u e r i e s b r i n g u p t h e p e r t i n e n t q u e s t i o n o f just w h e n t h e library f i n a l l y r e a c h e s t h e p o i n t w h e r e b o o k l o s s e s b e c o m e i n t o l e r a b l e . ment that h e was not dealing in hun- dreds of dollars but in tens of thousands. Disregarding for the moment the not insignificant costs involved in locating, buying, and cataloging replacements, let alone the irretrievable inconvenience to borrowers, the library's losses in mone- tary equivalent exceeded $150,000 over a three-year period. An inventory is ex- pensive in terms of time or money or both—perhaps it should be a "labor of love," to use a favorite expression—but in wandering about the stacks in the early hours of the morning, in the nas- cent glimmering of still another day one has time in which to consider how the li- brary might have enriched its collection had it not been required to spend so much in replacing missing books. It is at this time that the real tragedy of book losses becomes manifest. Obviously the library itself is not the principal sufferer in these matters. It is the public that has the most to lose in tolerating book losses; and t h e attitude of the borrowing public toward library security problems is strange and per- plexing. Indeed, it affords a major clue in unraveling the mysteries of borrower morality. A department store manager, for example, who reported to his board of directors that ten per cent of the an- nual inventory had been lost to shop- lifters might well expect those gentle- men—much in the manner related by Dostoevsky's Marmeladov—not to chase him out of their presence with a stick, but to sweep him out with a broom. Consequently, department store man- agers, as well as others in comparable occupations, have clamored long and loud about pilfering and what to do about it and have gained, to a certain extent, the support of a righteously in- dignant public. It may be that the public is more concerned with price increases than with the morality of the thing; but whatever the reason, shoplifting is be- ginning to receive the attention many Guarcls, Turnstiles, Electronic Devices / 261 think it deserves. B u t in libraries, the situation seems to b e different. Perhaps because w e are so eager to rid ourselves of the last vestiges of the custodial im- age, we have done little to overcome the problem or to educate the public in the matter of thefts. Education is desperately needed. T h e prevailing attitude seems at times to b e that it is a person's "right" to steal a book, much as it is his "right" to walk on the grass in front of city hall, if he so chooses and as long as h e is not caught. Why, a borrower might ask himself, cannot I take whatever I want from the library? W h o owns these books, anyway? I t is a curious but by no means inexplicable attitude. I t stems in large part apparently from a kind of arrogant individuality and from immaturity. That adults, too, steal books does not negate the argument, for age alone has never been a guarantee of maturity. T h e in- dividual in this country, or perhaps in any country, for t h a t matter, seems to have far greater respect for individual property than for corporate or public property. A man's property is his own; therefore to an honest man it is sacro- sanct; but public property, to that same honest man, whatever its form, belongs to all; ipso facto, stealing a library book is not really stealing at all. S U B J E C T S S E L E C T E D F O R S T U D Y T e n L C classes w e r e selected for the present study. They were: Social sciences Psychology ( B F ) Sociology ( H M - H X ) Political science ( J ) Language and literature English literature ( P R ) American literature ( P S ) Humanities English history ( D A ) French history ( D C ) American history ( E ) Other Medicine ( R ) Military and naval science ( U - V ) From a statistical aspect, a purely ran- dom sample of classes would have been preferable and might have resulted in slightly less deviation in the projections which were made to include the entire collection. T h e r e were several reasons for not taking a random sample. O n e was the desire to compare classes ex- periencing heavy, moderate, and light use. Although the entire collection might have been divided into three parts ac- cording to degree of use, and the selec- tion made from there on a random basis, certain obstacles stood in the way of doing this. T h e inclusion of class R, for example, which in the main library con- sists in the majority of books pertaining to psychiatry, made it possible to join (and compare) R with B F , to which it is closely allied. T o have been able to include B F and R in the sample by means of random selection would obvi- ously have been all but impossible. Second, another purpose of the study was to compare subject areas, such as between humanities and literature. Here, too, a random selection would not have accomplished the desired purpose. Fi- nally, certain classes had to b e arbi- trarily eliminated because most of the books in those classes were in other li- braries on the campus. In conclusion, although the principal purpose of the project was to determine over-all losses by means of a sample, to have eliminated classes not represented in the main library and to have divided the remainder by use and then b y sub- ject would have fragmented the collec- tion to such an extent that a true random selection would have been virtually im- possible, and, in addition, would prob- ably have defeated two major purposes of the study. T h e sample used, on the other hand, is large enough and covers such a broad spectrum of classes both 262 / College b- Research Libraries • March 1968 in use intensity and other important characteristics, to warrant belief that the figures gathered can be projected to en- compass the entire collection in the main library. Books in the social sciences receive much greater use than those in the humanities, and about the same as those in literature. Psychology and sociology represent high circulating classes, where- as political science falls considerably below average, at least in comparison with the other social sciences. All books in the western languages and literatures (including classical literature!) enjoy heavy use; therefore PR was selected be- cause it represents an average, and PS because it was known that American literature was entering a stage of very rapid growth. Finally, military and naval science was selected because it was vir- tually a dormant collection. But interest in military science—as in guerrilla war- fare—mushroomed shortly after the study began; and to a certain extent this was fortunate, because U-V circulation increased from less than .5 per cent to more than 4 per cent in one year, while losses, which were infinitesimal in the pilot inventory, increased as circulation increased. H Y P O T H E S E S AND T H E C O N D I T I O N S U N D E R W H I C H B O O K S W I L L D I S A P P E A R T h e hypotheses adopted for the study represent intralibrary conditions, involv- ing both borrowers and the book collec- tion, which encourage or deter book thefts. T h e r e are, in addition, other con- ditions, which alternately might b e called interlibrary factors, over which the library may or may not have control. As it turned out, these conditions have even greater influence over losses than intralibrary factors. Because they are theoretical (not in the dictionary defi- nition of a theory as "an analysis of a set of facts," but defined as "a more or less plausible general principle of- fered to explain phenomena"), and diffi- cult to verify, they can be proved or disproved only after investigation by many libraries. Hypotheses (intralibrary f a c t o r s ) : 1. A collection of large size, relative to another collection within the same li- brary building, will suffer a lower rate or percentage of loss; the larger the col- lection, the lower the rate. T h e actual number of books lost may be greater, but the rate of loss will b e lower. T h e rea- son for this is that borrowers have a greater range of selection and will b e less inclined to appropriate a particular book for their exclusive use. T h e ob- vious weakness in this hypothesis is that a large collection in one subject area may not b e comparable to a smaller collection in another subject area, par- ticularly if t h e larger collection has not been kept up-to-date. If, however, the intensity of use of the two collections is comparable, the hypothesis should b e valid. 2. In any given collection, a higher ratio of multiple copies to volumes will result in a lower rate of loss. This is be- cause borrowers will have greater op- portunity of securing titles in great de- mand, and therefore will b e less inclined to take a copy for their exclusive use. This hypothesis presumes that the library will have multiple copies of titles in de- mand at the time and not merely many copies of books used at some time in the past. I t also presumes an awareness on the part of the borrower that multi- ple copies are available. 3. T h e greater the intensity of use made of any collection, the higher the rate of loss will be, because a greater number of borrowers will be competing for a fixed number of volumes. 4. It follows, then, that given a con- stant rate of use, a collection experiencing a greater r a t e of growth will suffer a declining rate of loss. The difficulty here is that it is impossible to control the Guarcls, Turnstiles, Electronic Devices / 263 rate of use and difficult to predict the rate of growth. In addition, the state of the collection is of considerable impor- tance. A relatively undeveloped collection undergoing rapid expansion with basic titles will differ markedly from one which is already fairly well developed and being filled out with peripheral titles. This makes rate of growth a rather nebulous concept; however, in conjunc- tion with the first hypothesis above, a rapid rate of growth should promote a declining rate of loss. 5. T h e greater the number of books on reserve (i.e., closed reserve), the low- er will b e the rate of loss. T h e reason for this, presumably, is decreased access. Hypotheses one, two, four, and five represent inverse ratios, that is, the high- er or greater the first (controlling) fac- tor, the lower will b e the second (rate of loss). Hypothesis three, on the other hand, represents a direct ratio. T h e vari- ous hypotheses must b e presumed to b e interdependent—a large collection being used intensively will experience a greater loss rate than one of equal size under- going less use, but its loss rate should b e lower than that of a smaller collection being used with equal intensity, of a collection of equal size with a lower ratio of multiple copies. And so on. T h e weakness in proposing hypotheses to predict book losses is lack of control. This is especially true in the short run. T h e number of multiple copies can per- haps b e controlled over the long run, as can rate of growth to a limited extent. Size, quite naturally, is largely a product of age, unless as a matter of policy size is restricted. Intensity of use is a factor the library should not even attempt to control, except by increasing the rate of growth. But even though control over the hypothetical factors is not always possible, it does not seem unreasonable to expect the factors named to have some influence over losses. There are other intralibrary factors which will affect the rate of book loss. These factors involve the library and its patrons, but do not, except in one case, involve the book collection directly. 6. Relevancy of collection. Since books are stolen almost exclusively because people want to use them, a library that maintains a collection that is not or cannot b e used will experience a low loss rate in that collection. A library with an Oriental collection, for example, but existing in an environment that includes no one who reads Oriental languages, may expect its losses in Orientalia to b e practically nil. W h y the library would have such a collection is another matter. 7. Lending policy, including the de- gree of difficulty involved in obtaining a library card and the subsequent difficulty encountered in trying to borrow a book. T h e less red tape involved in obtaining and using a library card, the less a po- tential borrower will b e tempted to "bor- row" a book illegally. This thesis—free access without borrowing p r i v i l e g e s - will b e explored in greater detail further on. 8. T y p e of exit control. This factor will also b e considered in some detail later. It suffices to say at this time that exit controls, regardless of their type, are suc- cessful only to the extent that they keep honest men honest. No exit control can frustrate the designs of a determined thief, unless he is totally inept. Theories (interlibrary f a c t o r s ) : Interlibrary factors come into play be- tween one library and another, or be- tween the same library in two time periods, and between the library and the community it serves. 9. Open or closed stacks. It would seem beyond question that a library with closed stacks would suffer fewer losses than one with open stacks. And yet it cannot b e proved. It is virtually impossi- ble to compare an open stack library with one having closed stacks, even if they are comparable institutions in the 264 / College b- Research Libraries • March 1968 same area. Furthermore, a single library with open stacks at one time and closed at another cannot compare the difference very readily because so many conditions will have changed in the interim. 10. T y p e of borrowing public. A uni- versity library serving primarily graduate students, or a public library serving mainly post-school adults, should expe- rience lower losses, because there will be less concentration on the basic core col- lection, which exists in every library, and more on peripheral works relating to the particular interests of the individual bor- rower. T h e r e will be less competition for specific titles; therefore less likelihood of their being removed surreptitiously. 11. Urban or rural location. A library located in a rural or small town area should suffer lower losses than a com- parable institution in an urban area, because losses in an urban public or university library will vary directly with the quality of other libraries—specifically college and school—in the area. T h e ratio of potential patrons to total library re- sources may b e the same in both areas, but in a rural area the patrons may have but one (convenient) choice, whereas in an urban area, the library with the fin- est collection will find itself serving a disproportionately high number of bor- rowers. If the urban school and college libraries cannot meet the needs of their students, and apparently many cannot, their students will eventually gravitate to the well equipped public or university library; and if it is the policy of those libraries not to lend to high school or college students, some may find the temptation, a product of desperation and immaturity, to remove books irresistible. T H E S T U D Y A word about percentages. Several sets of percentages will be offered, pertaining to use, number of multiple copies, books on reserve, rate of growth, and, finally, number of volumes missing. Since the figures for losses are of the greatest im- portance and will be the only figures projected to include the entire col- lection—and in the process converted in- to a monetary equivalent—they are the only ones which will b e carried to two ( o r four, as in a decimal) places, in order to assure the greatest degree of accuracy. This will be done even though it is a known statistical principle that the results of any computation cannot b e more accurate than the least accurate figure involved. But in the use here, al- though loss figures will b e compared with other figures, they will not b e in- volved in computations with them; there- fore we can be excused for a little varia- tion from the rule. This being the case, in all calculations other than volumes missing, a figure such as 7.48 per cent, for example, will b e rounded off to the nearest significant figure, e.g., 7.5 per cent. Method of anticipating books that will be found. A certain number of books thought to b e missing in any one year will b e found the following and subse- quent years, consequently some method must b e devised to account for them, otherwise the loss figures for the last year of the inventory will b e disporpor- tionately higher than those of the first. I f we know how many of the books miss- ing in 1963 ( t h e first year of the inven- t o r y ) are found in 1964, 1965, 1966, and 1967, respectively, we should b e able to determine how many will b e found in 1968, 1969, and so on, as well as how many of those missing in 1964, 1965, and 1966 will b e found in en- suing years. W e can in this manner re- duce our loss figures accordingly. T h e experience available indicates that subsequent to the number of missing vol- umes found the first year following the inventory, roughly two-thirds as many will b e found the third year as were TABLE 1 Per Cent Rate of Growth Per Cent of Volumes Missing V O L U M E S I N C I R C U L A T I O N A N D O N R E S E R V E Year Volumes Duplicate Copies Total Volumes Per Cent Rate of Growth Volumes Missing Per Cent of Volumes Missing In Circulation Per Cent On Reserve Per Cent Class BF (Psychology) 1963 4,668 874 5,542 1964 4,918 949 5,867 5.9 102 1.74 307 5.2 565 9.6 1965 5,067 1,052 6,119 4.3 120 1.96 357 5.8 570 9.3 1966 5,498 1,225 6,723 9.9 142 2.11 443 6.6 617 9.2 Average 5,161* 1,075 6,236 6.7 121 1.95 370 5.9 584 9.4 (121.33) Ratio of duplicate copies to volumes 1,075 5,161 Class DA (History—Great Britain) = 20.8 per cent 1963 6,987 318 7,305 1964 7,156 337 7,493 2.6 25 .33 104 1.4 196 2.6 1965 7,402 379 7,781 3.8 31 .40 115 1.5 200 2.6 1966 7,695 409 8,104 4.2 33 .41 122 1.5 213 2.6 Average 7,418 375 7,793 3.5 30 ( 2 9 . 6 7 ) .38 114 1.5 203 2.6 Ratio of duplicate copies to volumes 375 7,418 Class DC (History—France) = 5.1 per cent 1963 4,816 189 5,005 1964 4,925 200 5,125 2.4 21 .41 81 1.6 108 2.1 1965 5,063 223 5,286 3.1 22 .42 88 1.7 110 2.1 1966 5,226 249 5,475 3.6 19 .35 88 1.6 117 2.1 Average 5,071 224 5,295 3.0 21 ( 2 0 . 6 7 ) .39 86 1.6 112 2.1 Ratio of duplicate 224 copies to volumes 5,071 = 4.4 per cent Class E (America [general] and United States [general]) 1963 8,037 908 8,945 1964 8,152 946 9,098 1.7 58 .64 220 2.4 469 5.2 1965 8,436 978 9,414 3.5 65 .69 248 2.6 479 5.1 1966 9,135 1,039 10,174 8.1 64 .63 330 3.2 548 5.4 Average 8,574 988 9,562 4.4 62 .65 266 2.8 499 5.2 Average ( 6 2 . 3 3 ) Ratio of duplicate copies to volumes 988 8,574 0 Averages derived from 1 9 6 4 - 1 9 6 6 only. = 11.5 per cent TABLE 1 (cont.) Year Volumes Duplicate Copies Total Volumes Per Cent Rate of Growth Volumes Missing Per Cent of Volumes Missing V O L U M E S I N C I R C U L A T I O N A N D O N R E S E R V E Year Volumes Duplicate Copies Total Volumes Per Cent Rate of Growth Volumes Missing Per Cent of Volumes Missing In | Circulation Per Cent On Reserve Per Cent Class HM-HX (Sociology) 1963 6,325 1,221 7,546 1964 6,485 1,398 7,883 4.5 140 1.78 401 5.1 750 9.5 1965 6,735 1,523 8,258 4.8 156 1.89 474 5.7 796 9.6 1966 6,994 1,709 8,703 5.4 183 2.10 519 6.0 834 9.6 Average 6,738 1,543 8,281 4.9 160 1.93 465 5.6 793 9.6 Average (159.67) Ratio of duplicate 1,543 copies to volumes 6,738 Class J (Political Science) = 22.9 per cent 1963 9,167 812 9,979 1964 9,389 880 10,269 2.9 84 .82 312 3.0 553 5.4 1965 9,684 907 10,591 3.1 91 .86 352 3.3 551 5.2 1966 10,031 950 10,981 3.7 94 .86 371 3.4 574 5.2 Average 9,701 912 10,613 3.2 90 .84 345 3.3 559 5.3 Average 10,613 (89.67) Ratio of duplicate copies to volumes 912 9,701 Class PR (English Literature) = 9.4 per cent 1963 17,410 1,986 19,396 1964 18,266 2,153 20,419 5.3 219 1.07 1,035 5.1 1,055 5.2 1965 19,286 2,253 21,539 5.5 241 1.12 1,179 5.5 1,047 4.9 1966 20,446 2,360 22,806 5.9 261 1.14 1,276 5.6 1,076 4.7 Average 19.333 2,255 21,588 5.6 240 1.11 1,163 5.4 1,059 4.9 Average 2,255 21,588 (240.33) (5.39) 1,059 Ratio of duplicate 2,255 copies to volumes 19,333 11.7 per cent Class PS (American Literature) 1963 7,546 813 8,359 1964 8,475 969 9,444 13.0 116 1.23 698 7.4 491 5.2 1965 10,346 1,127 11,473 21.5 140 1.22 768 6.7 490 4.3 1966 11,638 1,220 12,858 12.1 140 1.09 815 6.3 508 4.0 Average 10,153 1,105 11,258 15.5 132 1.17 760 6.8 496 4.4 Ratio of duplicate 1,105 copies to volumes 10,153 Class R (Medicine) 1963 2,449 391 2,840 1964 2,570 407 2,977 4.8 64 2.15 143 4.8 185 6.2 1965 2,763 431 3,194 7.3 68 2.13 186 5.8 199 6.2 1966 2,916 479 3,395 6.3 74 2.18 190 5.6 216 6.4 Average 2,750 439 3,189 6.1 69 2.15 173 5.4 200 6.3 Average 2,750 ( 6 8 . 6 7 ) ( 5 . 4 2 ) Ratio of duplicate copies to volumes 439 2,750 Class U-V (Military and Naval Science) = 16.0 per cent 1963 831 28 859 1964 892 31 923 7.5 14 1.52 40 4.3 10 1.1 1965 974 52 1,026 11.2 24 2.34 53 5.2 23 2.2 1966 1,055 65 1,120 9.2 20 1.79 61 5.4 21 1.9 Average 974 49 1,023 9.3 19 1.89 51 5.0 18 1.8 Average ( 1 9 . 3 3 ) Ratio of duplicate copies to volumes 4 9 rcn = 5.0 per cent 9 7 4 Composite 1963 68,236 7,540 75,776 1964 71,228 8,270 79,498 4.9 843 1.06 3,341 4.2 4,382 5.5 1965 75,756 8,925 84,681 6.5 958 1.13 3,820 4.5 4,465 5.3 1966 80,634 9,705 90,339 6.7 1,030 1.14 4,215 4.7 4,724 5.2 Average 75,873 8,965 84,839 6.0 944 1.11 3,792 4.5 4,524 5.3 Average ( 9 4 3 . 6 7 ) 4,524 Ratio of duplicate copies to volumes 8,965 , . Q 7 5 ^ 7 3 = 1 L 8 p e r C e n t 268 / College b- Research Libraries • March 1968 found the second, and in the fourth year about two-thirds as many will b e found as in the third year. This projection may b e carried out until we reach a point where none or only one book will be found. F o r the purpose of this study the projected loss reduction was calcu- lated to one. Thus the individual and composite figures for losses do not repre- sent the number of volumes presently missing, but a smaller number, that should b e missing at a calculated time in the future. Admittedly, this is a some- what rough method of arriving at a true figure, but it has the advantage of indi- cating losses at a minimum and is cer- tainly more accurate than simply record- ing the figures as they now stand. Method of calculating volumes out and on reserve. T h e figures given for volumes out and on reserve do not represent total circulation, or the total number of volumes on reserve in any given year. They were derived from an average of the greatest and least number within the year, as obtained from two one-week periods representing the high and low points within the year. T h e second week in May was selected for the high point, and the first week in Sep- tember for the low. This method, while not indicating total use, is quite accept- able, as the figures for all classes were derived in the same manner. The inventory. All classified mono- graphs and serials were included in the inventories, and, while separate figures were recorded for each, only the com- bined figures for monographs and serials are included herein. T h e total number of volumes and duplicate copies in each class was obtained from a shelf list count taken immediately preceding each in- ventory. In the correlation charts, the validity of the five hypotheses is examined against the losses. T h e loss figures are arranged from high (class R ) to low (class D A ) . T h e factor intensity of use is also arranged from high (class P S ) to low (class D A ) , because this hypothe- sis stated that a greater intensity of use would result in a greater rate of loss. With the other hypotheses, however, the classes are ranged low to high, in keep- ing with our prediction that the higher the percentage of multiple copies, the greater the rate of growth, the larger the collection, and the greater the number of books on reserve, the lower would be the rates of loss. In the factor of multiple copies, to give but one example, class D C (4.4 per c e n t ) ranked last, whereas class H M - H X ( 2 2 . 9 per c e n t ) ranked first; consequently, on the basis of this one hypothesis, w e should expect to find class D C first in rate of loss and H M - H X last. T h e f a c t that class DC was ninth in losses means that it was +8 positions from its predicted position, while HM- HX, being third in losses, was - 7 po- sitions from its predicted location. W e must therefore say that there seems to be little relationship between losses and the rate of multiple copies (as an isolated hypothesis). Had the ranking of percent- age of multiple copies been DA, DC, E , J, PR, PS, U - V , HM-HX, B F and R (low to high ranking), and the rate of loss the same as the study demonstrated, we would have had perfect correlation ( 0 ) , and would have been able to assert that, in any collection, losses are directly re- lated to the number of multiple copies available—i.e., the greater the latter the lower the former. Sad to say, we cannot make this assertion. T h e maximum correlation possible is 0, as we can see from the listing on the left ( T a b l e 2 ) , while the minimum is + 25 (right, Table 2 ) . T h e correlation between losses and the five measurable hypotheses is given in T a b l e 3. Lest the reader b e misled b y the fig- Guarcls, Turnstiles, Electronic Devices / 269 T A B L E 2 M A X I M U M C O R R E L A T I O N M I N I M U M C O R R E L A T I O N 1st F a c t o r 2 d F a c t o r 1st F a c t o r 2 d F a c t o r e.g., L o s s e s e.g., U s e V a r i a t i o n e.g., L o s s e s e.g., U s e V a r i a t i o n B F . . . B F 0 B F . . . U-V + 9 DA . DA 0 DA . . . R + 7 DC D C 0 DC . . . PS + 5 E . . . E 0 E . . . PR + 3 HM-HX HM-HX 0 HM-HX . . J + 1 J J 0 J . . . HM-HX - 1 PR . P R 0 PR . . . E - 3 PS . . PS 0 PS . . . DC - 5 R R 0 R . . . DA - 7 U-V . . U-V 0 U-V . . . B F - 9 Total variation 0 ± 2 5 ures for book and dollar losses, it must b e pointed out that not all lost books are actually replaced; h e n c e the expression "monetary equivalent." Obviously, since the figure 17,342 is derived from a sam- ple,4 only those books known to b e lost are considered for replacement, and, in general, only titles of which the library has but one copy a r e replaced, unless demand indicates t h e need to replace duplicate copies. Evaluation of hypotheses. W e have seen that the maximum correlation pos- sible between rate of loss and any hy- pothesis is 0, and t h e minimum ±25. A correlation of ±25 would indicate no re- lationship between a so-called control- ling factor and losses, while one of 0 would indicate perfect or absolute cor- relation. As it turned out, of all the hy- potheses, only one proved to b e closely related to losses. T h a t intensity of use is directly related to losses is unquestion- able, and hardly surprising. It is sur- prising, though, that the other hypothe- ses had so little apparent influence. At the beginning of the study, it seemed a foregone conclusion that a high per- centage of multiple copies would result in a lower rate of loss, but apparently it does not. H M - H X , B F , and R have a 4 T h e s a m p l e c o n s i s t e d o f a p p r o x i m a t e l y % ( 1 6 p e r c e n t ) o f the c o l l e c t i o n in t h e m a i n library. high ratio of duplicates, and therefore should have experienced a low loss rate, but they actually were the highest in losses! A low rate of growth, such as experienced by classes D C , J, and DA, should have resulted in a high rate of loss. I t did not. T h e same observation may b e made of volumes on reserve, with the possible exception of classes U-V. 5 T h e only other controlling factor which seemed to have any influence at all was size of collection, and even here the correlation was far from outstanding. In conclusion, then, we must say that of the five hypotheses, only intensity of use had any real direct effect, and that the others either had little or no effect, or, and this is more likely, were simply over- shadowed by the factor of use. So much for hypotheses. Theories. W e must now examine the policies and characteristics of the library itself with regard to several of the theo- ries put forth at the beginning of the study. T h e reader will recall that we theo- rized: ( a ) that a liberal lending policy will deter thievery, whereas a policy that excludes potential borrowers will 5 A l t h o u g h t h e h y p o t h e s e s m u s t b e a s s u m e d t o b e i n t e r d e p e n d e n t , a n d t h e r e f o r e t o s o m e d e g r e e s e l f - c a n c e l i n g , it d o e s s e e m o d d t h a t f o u r h y p o t h e s e s s h o u l d b e s o c o m p l e t e l y d o m i n a t e d b y o n e . 270 / College b- Research Libraries • March 1968 encourage it; 6 ( b ) that an open stack library will suffer heavier losses than one with closed stacks; ( c ) that a library serving a specialized public will lose fewer books than one serving a primarily undergraduate college, or school stu- dents; a n d ' ( d ) that a library located in an urban area will experience greater losses than a comparable institution in a rural location. F e w would deny that any attempt to fathom the motivation of thievery by means of theoretical devices is risky. But we have seen that even the most valid intralibrary factor, while serving well to show why losses are great, does not explain the behavior of the thief. T h e fact that heavy use and high losses are closely related does not explain why books are stolen. Nor could it. There must b e another factor, or factors, which provide the impetus. W e must, there- fore, by necessity, seek some plausible theory relating to use and losses. It is of the greatest importance that we do so, for unless we can establish a theory, we cannot hope to achieve a lasting solution. T h e library we have examined has the simplest of registration procedures. All full-time students, staff, and faculty are 0 By this I mean not only absolute exclusion of "nonresident" potential borrowers, but also the various obstacles which may impede the use of the library by a legitimate borrower and cause him to "borrow," so to speak, before he is a borrower. issued I D cards, which also serve as li- brary cards. Part-time students, and others, of whom there are many, need only fill out an application, whereupon they are issued a library card imme- diately. W e can say, then, that there are few impediments placed in the way of the qualified borrower obtaining and us- ing a library card. In addition, the li- brary has a very generous lending poli- cy, offering unlimited renewals, unless wanted by other borrowers, by mail, telephone, or in person. Books are loaned for two and four weeks to undergraduate and graduate students respectively, and on an indefinite basis to faculty, and any book in circulation will be reserved and held for any borrower. No charges (other than tuition) are m a d e for library cards or services. There is no limit to the num- ber of books that may be borrowed; and a complete listing of all books in circu- lation is available for use by the public. Finally, faculty and graduate students of other institutions in the area are per- mitted to borrow. T h e r e are restrictions, however; and within them we may find the answer we are seeking. Undergraduate students of the other twenty-odd colleges in the area, as well as high school students, are not permitted to borrow; but all college students are permitted to use materials in the library, and it is well known that TABLE 3 B O O K S L O S T P E R C E N T R A N K E D H I G H T O L O W I N T E N S I T Y O F U S E R A N K E D H I G H T O L O W R A N K E D L O W T O H I G H P E R C E N T M U L T I P L E C O P I E S Class Rank Lost Class Rank Variation Class Rank Variation R . 1 2.15 PS . . 1 +4 D C . . 10 + 8 B F . . 2 1.95 B F . . 2 0 U-V 9 + 2 HM-HX 3 1.93 HM-HX 3 0 D A . . 8 + 7 U-V 4 1.89 R . . 4 - 3 J 7 + 3 PS . 5 1.17 PR 5 +1 PS . . 6 0 PR . 6 1.11 U-V 6 - 2 E . . 5 + 2 J 7 .84 T 7 0 PR 4 - 1 E . 8 .65 E . . 8 0 R . . 3 - 7 D C . 9 .39 D C . . 9 0 BF . . 2 - 7 D A . . 10 .38 D A . . 10 0 HM-HX 1 - 7 Total variation ±5 ±22 Guarcls, Turnstiles, Electronic Devices / 271 T A B L E 4 R A T E O F G R O W T H S I Z E O F C O L L E C T I O N V O L U M E S O N R E S E R V E R A N K E D L O W T O H I G H R A N K E D L O W T O H I G H R A N K E D L O W T O H I G H Class Rank Variation Class Rank Variation Class Rank Variation DC . . 10 + 8 U - V . . 10 + 3 U - V . . 10 + 3 J 9 + 5 R . 9 - 1 DC . . 9 + 7 DA . 8 + 7 D C . . 8 + 6 D A . . 8 + 7 E 7 + 4 B F . . 7 - 2 PS . . 7 + 1 HM-HX . 6 - 2 DA . 6 + 5 PR . . 6 + 1 P R . . 5 0 HM-HX 5 - 3 E . . 5 + 2 R . . 4 - 6 E . . 4 + 1 T • • 4 0 B F . . 3 - 6 I • • 3 - 1 R . . 3 - 7 U - V . . 2 - 5 PS . . 2 - 4 B F . . 2 - 7 PS . . 1 - 5 PR . . 1 - 4 HM-HX 1 - 7 Variation ±24 ±15 ±21 high school students use the library despite efforts to prevent it. Thus we have access without the privilege of borrowing. But access alone does not explain losses. I f the number of books in circulation at any time is indicative of intensity of use, then it should also be indicative of potential use by non-borrowers. Con- ditions will not be the same, but they will be similar to t h e extent that broad subject areas used intensively by one group may well b e used by another group, given the opportunity, which though not exactly the same in nature, is at least comparable. Literature and t h e social sciences, which are the classes most heavily used and greatly depleted, are more likely subjects of general interest than are English and French history, or military and naval science. Admittedly, there are weaknesses in this argument, but it cannot b e denied that of the ten sub- jects considered, the five in the top half of use were with but one exception the same group which suffered the heaviest losses. And it is these subjects that are of "universal" interest and subject to the greatest demand b y library-using seg- ments of the population within the po- tential environment of the library. This would indicate very strongly that rigid interlibrary relationships, taken in con- junction with known facts, provide a solid foundation upon which to build a theory of book losses. The library and its environment. The library has open stacks, except for spe- cial collections and about five thousand books on closed reserve. I t is located in a metropolitan area of approximately two million persons, and, along with the public library, ranks as the finest general library in the area. Not a great amount of pertinent information is available con- cerning the numerous school and college libraries in the area, but what there is indicates very strongly that many are in- adequate ( a n d some grossly inadequate) in comparison to the major libraries in the area. In fact, two of the largest col- leges in the area (with a combined en- rollment of more than 10,000 students) are relatively new, and have what might at best b e described as embryonic li- brary facilities. T h e milieu in which the library op- erates would seem to fit our theory very well: a large urban complex, open stacks, free access, and a very large group of potential borrowers. It is impossible to overlook the basic fact that the library is convenient to two groups: a small group that is permitted to borrow, and a much larger group that is not. I t would b e less than wise to b e en- tirely unyielding in attributing losses to those who are not permitted to bor- 272 / College b- Research Libraries • March 1968 T A B L E 5 L o s s D E R I V E D F R O M I N V E N T O R Y I N C L U D E S E N T I R E M A I N L I B R A R Y C O L L E C T I O N L o s s AS A P E R C E N T A G E O F T O T A L B O O K B U D G E T Y e a r T o t a l V o l u m e s i n C o l l e c t i o n " P e r C e n t o f V o l u m e s M i s s i n g T o t a l V o l u m e s M i s s i n g C o s t p e r V o l u m e f L o s s e s in M o n e t a r y E q u i v a l e n t T o t a l A l l o c a t e d f o r B o o k s a n d C o n t i n u a t i o n s P e r C e n t 1964 . . 1965 . . 1966 . . 495,250 516,950 548,418 1.06 1.13 1.14 5,249 5,841 6,252 $9.00 9.00 9.00 $ 47,241 52,569 56,268 $322,041 361,547 478,700 14.7 14.5 11.8 Total 17,342 $156,078 ° M e d i a n f i g u r e f o r t w e l v e m o n t h p e r i o d , m a i n library o n l y . f E s t i m a t e d a v e r a g e p r i c e , b o o k s , a n d c o n t i n u a t i o n s . T h i s figure, w h i l e p r o b a b l y l o w , i s a d e q u a t e f o r t h e p u r - p o s e i n t e n d e d . row. A theory must b e proved before it can be stated as an axiom, and a theory such as ours is most difficult to prove. But we do have a beginning. It would seem that there must b e some connec- tion between losses and inability to bor- row. Taking into account the inade- quate library facilities of many institu- tions in the vicinity, both secondary and higher, and considering the fact that the faculties of those institutions cannot and will not scale down their own standards because of inadequate library resources in their own institutions, and, finally, recognizing the tremendous pressure put upon students to achieve high scholastic ranking, it is not surprising that a li- brary which permits and even encour- ages free access, and possesses one of the finest book collections in the area, would suffer heavy losses. W e can, there- fore, offer as a tentative theory that the pressure of scholastic achievement in association with freedom of access to those persons who are not permitted to borrow will result in a high rate of loss. Methods of preventing thefts. Given that libraries lose books in sufficient numbers to warrant remedial action, we must consider the current methods em- ployed to prevent thefts, their shortcom- ings, and the possible alternatives. I t would b e best to begin by admit- ting that there is no known method of preventing a determined thief from mak- ing off with a book. This is so because the effectiveness of all systems, regard- less of their type, is contingent on the cooperation of all borrowers, and since there are borrowers who obviously do not cooperate, all systems must fail to the exact extent to which that coopera- tion is withheld. I t is ironic that exit control systems, all of which are expen- sive (some more so than others), are effective only to the extent that they remind cooperative borrowers to check out books, whereas they can b e rendered ineffective b y a determined thief.7 T h e two major methods employed to- day are guards and turnstiles, and a method employing magnetic influences —electronic detection. Guards and turnstiles. Somewhere in this land there may b e a library that has but one exit. Most have two or more, al- though only one may be legal; and since we are dealing with the minority who steal books, either we guard all possible exits or the battle is lost, for honest men always walk out the front door, but thieves do so only when they are confident or desperate. Obviously, even in a modern building, not all exits can b e guarded. Windows that open, unguarded doors, fire escapes, crash doors, delivery entrances and exits, and the like all provide ready roads of egress for the clever thief. There is no way 7 F o r c o m m e n t s on this s e e W i l l i a m L . E m e r s o n , " T o G u a r d o r N o t to G u a r d , " i n Library Journal, L X X X I V ( J a n u a r y 1 5 , 1 9 5 9 ) , 1 4 5 - 4 6 . Guarcls, Turnstiles, Electronic Devices / 273 of guarding all of them. As for the rare (and probably nonexistent) library that actually has but one exit, the book thief is confronted by a more formidable but by no means insoluble problem. T h e methods he may employ are numerous— by concealing books under coats and belts, in innocuous looking packages and typewriter cases, in laundry bags and handbags, in dust jackets and news- papers, or brazenly walking past dis- tracted guards and preoccupied attend- ants, and on and on ad infinitum. It is pointless to dwell on the many methods; it is enough to remark that they are all too effective. Electronic devices. T h e writer recalls having visited the display booth of one of the electronic detecting companies at a recent ALA conference. Included in the propaganda handed about was a short story, one that reminded him no little bit of the famous dime novels so prevalent in the post-Jesse James era. In this thriller, one John was to b e seen skulking about the stacks of a library carrying what could b e described either as a large briefcase or a small suitcase. John was looking for valuable books; and, disdaining the dealers' catalogs, he was looking for them in the dark and dingy stacks of a dark and dingy library. He wanted only valuable books because he intended to peddle them—to whom was not disclosed. B e that as it may, John crept from range to range, and with many a surreptitious look up and down the aisle ways, selected his books. Finally, case bulging with loot, John headed for the front door. All was quiet in those halcyon halls. But lo! Barely had John stepped through the front door, when the long arm of the law claimed its hapless victim. John had been caught by the electronic detector, with the as- sistance of an alert librarian who, upon hearing the warning bong, had signaled the library guard who just happened to b e at the front door waiting in breathless anticipation. And so there was a happy ending. T h e library's books were spared an ignominious fate, the theory of elec- tronic detection by means of the sinus- oidal propagation of the magnetic influ- ences was vindicated, and John was carted off to jail, there on "the torture of the mind to lie in restless ecstasy." One more book crook consigned to limbo. I t was a good story. I t warmed the heart and emphasized three salient fea- tures of electronic detection. First, the system presumes not only that valuable books alone are stolen, but that thieves are interested in books only as artifacts. Second, even though the system pur- portedly does away with exit attendants, inherent in its operation is the continu- ous presence of someone in authority at the front door to apprehend the would-be thief. And third, the system is predicated upon the belief that electron- ic devices are infallible. Therein lie the fallacies of electronic detection. W e have contended that books are stolen because people want to use them; that is, they want the intellectual con- tent of the book, not necessarily the whole book itself. If this is true, and there is no reason to believe otherwise, a system which in effect protects the cover or spine of a book, but not the contents, can b e circumvented simply by removing the contents and leaving the protected part behind. In truth, such a system might even b e said to make book stealing easier and safer! One can readily envision a library floor littered with book covers. It is true that some books are stolen for their intrinsic value and thus can be protected by the elec- tronic device; but they represent only a small percentage of the vast number of books pilfered from libraries every year. T h e reader may wonder at this scorn- ful and invidious attitude toward elec- tronic detection. I t stems in part from the brash manner in which the promot- 274 / College b- Research Libraries • March 1968 ers of the devices seem to rejoice in their conviction that conventional turnstiles are antediluvian and must be replaced at the first opportunity by the miracle of magnetism. And all this with little or no foresight or thought about the reasons people may have for stealing books. Here, once again, is a case of nonlibrari- ans telling librarians what is best for them. It stems also from the happy aban- don with which they recommend con- verting to a new method, while letting the library itself worry about how much it will cost. How much, indeed, would it cost to "protect" a collection of five hundred thousand volumes? At that same conference I was told—in a whisper, as though the speaker feared that some guardian of the living wage might be hovering about—that it could b e done for about one cent a volume, not includ- ing the cost of the discs and equip- ment.8 An experiment along these lines indicated that at one cent per volume the person performing such labor would b e receiving approximately $ .09 an hour! Furthermore, the discs are usually attached to the board paper and then covered by the fly leaf, which is then glued to the board paper.9 Unfortu- nately, the misuse of the endpaper would defeat the purpose of the hinge, which is designed to prevent the covers from being torn away from the book. It is unfortunate that electronic detec- tion is not the answer to the theft prob- lem. I t is a reasonably good idea in theory; however, its cost makes it im- practical, while its inherent weaknesses make it unworkable. C O N C L U S I O N Is there no way in which losses may be reduced? Before we can answer this question we must ask another: what level of losses are we willing to tolerate? Is one per cent of the collection per year too high a price to pay for open stacks and so-called free access? Many would say it is. What, then, is an ac- ceptable level? If zero per cent is the goal then the library may as well close. T h e individual library must recognize that it is going to lose some books, the number being directly related to the environmental and other conditions un- der which it operates; the essential idea is to reduce losses to the lowest possible level within those conditions. There are several possible approaches, only a few of which can b e considered in brief at this time. First, the library might simply accept losses as they are and request of its governing board an in- crease in book funds to compensate for them. This, however, is no solution and, in addition, it might b e much easier to reduce losses than to convince a board of the merit of such a program. As a second approach, it is entirely possible that losses could b e reduced to an acceptable level by reducing access. This need not necessarily involve closing the stacks to all; the library could simply turn its exit control about and screen all who come in rather than those who go out. Bar the door to all who do not have library cards! But such a solution might well turn into a hydra of enormous pro- portions. Not only would a "no-access to outsiders" policy bring forth strong pro- tests, and doubtless reciprocal treat- ment, but it would b e inimical to the spirit of librarianship. A college or uni- versity library, to be sure, owes its first obligation to its own faculty and stu- dents; but all libraries, university as well as public, are obligated to a greater or lesser extent to serve the community. It is a sticky problem. 8 To get around this embarrassing predicament, the promoters sometimes recommend that only certain books be protected. If by this they mean an entire reference collection, or any other complete group of books, that is one thing; but if they mean every fifth or tenth volume in the general stack collection, they are merely encouraging the disappearance of the other four or nine books. 0 Books sent for binding or rebinding can have the discs attached directly to the board. This would be much better for the book, at least until someone tore off the cover. Guarcls, Turnstiles, Electronic Devices / 275 Third, since we cannot hope to pre- vent people from stealing books if such is their intent, we must attempt to con- vince them that it is not in their own best interest to do so. This is not so naive as it may at first seem. Another writer once said, perhaps with tongue in cheek, that "education is usually prescribed for every ailment of democra- cy, and to some extent it can b e used to develop a more healthful attitude toward publicly owned book collections."1 0 This was written in 1935 and, unless we can prove it invalid, it is still worth serious consideration. However mushy it may sound, education must b e one of our basic approaches. T h o s e of us who are committed to the educational process, and presumably all of us are, and who believe in its great virtues, should at least attempt to reach out to those who stand so painfully in need of education. W e need not preach that stealing books is morally wrong—the function of librarianship is not to preach morality, and, in any event, the attempt would fail, for morality is not something that adults or near adults learn easily. W e need not preach at all. T h e would-be book thief must b e convinced by irrefutable logic that stealing books inevitably be- comes a reciprocal curse; that the theft of a book injures all; that stealing books represents the ultimate in folly. W e must persist in our attempt to edu- cate the book thief. I t is not enough to put up an occasional poster, or print a few notices in newspapers; it must b e a personal approach, and it must b e 1 0 Ralph Munn, "The Problems of Theft and Mutila- tion," in Library Journal, L X (August 1 9 3 5 ) , p. 590. hammered at over and over again. F o r we must assume that library users are rational beings, and rational beings do not set out deliberately to hurt them- selves, which is what they do when they steal books. If they are rational they will see the truth; if they are not, then there is little point in worrying about the prob- lem at all. Finally, it is possible that the ultimate solution lies in a combination of educa- tion and cooperation. Instead of reduc- ing access, we might expand it. A com- prehensive interlibrary metropolitan co- operative lending plan could prove to b e less costly in the long run than thefts. No doubt such a program would place heavy burdens on the well-endowed li- braries and might cost them more than they are willing to pay. But that is some- thing that cannot b e answered at this time. Ultimately the most unfortunate conse- quence of book thefts is not the mone- tary loss, but the irreplaceable loss in human effort. T o the casual reader a lost book may be an inconvenience; but to the serious reader it can b e a minor tragedy. And it is little comfort to the scholar to be told that a missing title will b e reordered and may b e available some time in the future. I t is to b e hoped that there will b e further exploration of the problem, with the idea of attempting to discover the relationship, if any there is, between de- termining factors and the types of books stolen, as well as of devising an effective program of education and cooperation to counteract losses. Book stealing is an intriguing problem; its cost warrants more than passing attention. • •