7/28/2019 j.1467-7660.1971.tb00326.x

1/17

YEHEZKEL DROR

Strategies for administrative reform *

The term strategy has a variety of meanings, such as: a detailed set ofplanned responses to all possible contingencies in theory of games; fun-damental policies and overall postures in strategic analysis dealing withforeign relations and defense; and main goals and principles for oper-ations which serve as a framework for tactics in military planning.My usage of the term strategy in this paper is related to its meaningsin the analysis of foreign relations and defense and in military planning.But relying on recent work in policy sciences, I am using the conceptstrategy in a somewhat more structured way- as referring to a definedseries of master-policies or mega-policies. Strategies establish theframeworkof guidelines and the boundaries of policy space within whichoperational and detailed policies are to be established and decisions areto be made. Strategies for administrative reform must therefore deal withissues such as: overall goals of administrative reforms; the boundaries ofadministrative reforms; preferences in respect to time; risk acceptability;choice between more incremental or more innovative reform; preferencefor more balanced vs. more shock-directed reforms; relevant assump-tions on the future; theoretic (tacit or explicit) assumptions on which thereform is to be based; resources available for the administrative reform;and the range of feasible reform instruments.

* An y views expressed in this paper are those of the author. They shou ld not beinterpreted as reflecting the views of the RAND Corporation or the official opinionor policy of any of its governmentalor private research sponsors. The author is verygrateful to Professor A. F. Leemans for his very helpful comments on an earlierdraft of this paper.1 See Yehezkel Dror, Design for Policy Sciences (New York: American Elsevier,forthcoming, 1971).

7/28/2019 j.1467-7660.1971.tb00326.x

2/17

20 YEHEZKELDRORThese and s imilar dimensions of strategy can serve as a conceptualframework for behavioral study of actual administrative reforms, both

historic and contemporary . Indeed, the absence of any com prehensiveconceptua l framework is an important reason for the shallowness of mostwritings on administrative reforms an d their failure to penetrate beyondtechnicalities into bas ic features.* From a prescriptive point of view,systematic com parative study of adm inistrative reform s with the help ofthe proposed dimensions of strategy is, therefore , urgently needed toprovide feedback from experience as an aid to better administrative re-form strategy decisions. But even in the absence of com parative empiricstudies, the proposed strategy dimensions can and should be used forimprovem ent of adm inistrative reforms, at least by providing a policyanalysis networks fo r explicit consideration and decision of crucial reformfeatures.Th e preferable mix of adm inistrative reform strategies is in the m aina function of the concrete circum stances of each particular reformsituation. Specific local needs, availability of different resources, historictraditions and indigenous political culture - hese illustrate the uniquevariables which make impossible the existence of universal optimal ad-ministrative reform strategies. Some generalized recommendations inrespect to a num ber of strategy dimensions may be useful in respec t todefined groups of situations, such as developm ent coun tries, pos tindus trial societies, threa tened countries, etc. A few recomm endationsin respect to single strategy dimensions may even be of universal validity,resulting more from the very characteristics of adm inistrative institutionsthan from their particular forms in different countries. But, in general,identification of preferab le adm inistrative reform strategies depends onunique situations and therefore requires careful an d detailed analysis onthe basis of elabora te data.

Nevertheless, the concept of reform strategies and an explicated setof main dimensions of reform strategies can be of much applied useful-ness as a framework for administrative reform policymaking. My mainargument here is closely related to my use of the term administrativereform. By adm inistrative reform I mean directed change of main fea-* Distinguished exceptions include for instance, Dwight W aldo, The Adminis-trative State (New York: Ronald Press, 1948); S. N. Eisenstadt, Th e Political S ys-tems of Empires (N ew York: Free Press, 1963); and Gerald Caiden, AdministrativeReform (Chicago: Aldine Publishing Co. , 1 969).8 A policy analysis network is a morphological decomposition of an issue intocompon ents structured as a coherent program for analysis and decision.

7/28/2019 j.1467-7660.1971.tb00326.x

3/17

STRATEGIES FOR ADM INISTR ATIVE REFORM 21tures of an administrative system . This definition recognizes the existenceof a continuum between administrative improvements and administrativereforms, the exact borderline between these two depending on onesperception of what are main features of an administrative system. But,more important, this definition makes clear my concern with adminis-trative reforms as a directive and conscious activity, rather than a postfactum classification of administrative-change phenomena, which may b ethe aggregative result of incremental changes or the undirected side effectsof sweeping social movements and societal transformations (which, ofcourse, often are also causes of directed main administrative systemchanges, i.e., administrative reforms).Given this conception of administrative reform as consciously directedchange of the main features of an administrative system, the general re-comm endation that explicit and carefully considered strategy determina-tion will improve the quality of administrative reforms is a special caseof a general theory of preferable policymaking. Full exposition of thisgeneral theory would be ou t of place here.4 But the general principle thatfuller and more explicit consideration of the main dimensions of a de-cision issue will usually contribu te to better decisions is so strongly sup-ported both by decision theory and by comm on sense as not to requiremuch elaboration.Let me nevertheless point out some caveats, so as to avoid the over-simple impression that more explicit strategy determination is alwaysconducive to better administrative reform, and at the same time to in-dicate some difficulties of explicit administrative reform strategy deter-mination:1. Administrative reform strategy determination requires a high-quality staff. Unless good strategies are decided upon, it may be better tomuddle through, so that at least serious mistakes can be avoided. Th e

same is not necessarily true for administrative reforms as a whole - nsome situations nearly any change is better than perpetuation of the pre-sent situation; also, strategy determination requires more sophisticationthan many other aspects of adminstrative reform.2. Explicit administrative reform strategy determination involves clearchoice between a range of alternatives that a re favored by different values,interests, organizations and personalities. Therefore, strategy determin-ation often involves significant political costs in term s of coalition main-

4 For such a general theory, see Yehezkel Dror, Public Policymaking Reexamined(San Francisco: Chandler Publishing Comp any, 1968).

7/28/2019 j.1467-7660.1971.tb00326.x

4/17

22 YEHEZKEL DRORtenance, support recruitment, implementation possibilities, participation ,etc.3. Establishment of adm inistrative reform strategies may easily leadto over-rigidity anless the strategies themselves are elastic, are related toclearly defined contingencies, and are subjected to periodic reexamina-tion. This again assumes not only a highy qualified staff, but also quiteadvanced administrative reform policymaking and monitoring.In short, determination of good administrative reform strategies pre-supposes to some extent a high-quality policymaking system - which isusually absent exactly when administrative reform is most urgently need-ed. Therefore, it is important to emphasize that my recommendation toexplicitly decide adm inistrative reform strategies should not be regardedas a procrustean bed which provides a rigid framework to be followedequally in all cases. But I think that in practice the danger of too muchadministrative reform strategy decisionmaking is unreal: resistance toabstract thinking and to systematic policymaking is usually very strong;therefore, I think that adm inistrative reform strategy decisionmaking willnot be overdone. Rather, the real danger is in the opposite direction, withconcrete reform issues driving out of cons ideration the underlying strat-egy dimensions, resulting in ill-considered, inconsistent, over-atomized

and too short-range administrative reform attempts. Therefore, the over-all recommendation to pay more attention to administrative reform strat-egies is justified and therefore a set of adm inistrative reform strategydimensions should be of practical help, at least by serving as a check listand as a sensitizer, and often by serving as a framework for a t least som eexplicit cons ideration of the strategic dimensions of a contemplated ad-ministrative reform.Having explored the significance of adm inistrative reform strategies,let me pass on to a discussion of eleven main strategy dimensions.

I. OVERALL GOALSA standard answer to the question, administrative reform fo r what? isto increase efficiency. But this is largely a meaningless answer, be-cause the real question is efficiency, or effectiveness, in doing what?Different administrative system characteristics are more effective andefficient for fulfilling different functions and achieving different goals.Therefore, clarification of the overall goals of an administrative reformis a fundamental requisite for success. Usually, an administrative reformwill be a multi-goal-oriented endeavor. This makes it all the more neces-

7/28/2019 j.1467-7660.1971.tb00326.x

5/17

STRATEGIES FOR ADMINISTRATIVE R EFOR M 23sary to examine the consistency of the various goals and to establishnecessary priorities.To illustrate the parameters of this dimension, let me mention a fewpossible administrative reform goal clusters:a. Traditional administrative efficiency, in the sense of saving moneythrough form simplification, procedure change, duplication reductionand similar O.M. (organization and methods) approaches.b. Reduction of perceived weaknesses, such as corruption, favoritism,political spoils and so on. (This debugging goal is itself very hetero-geneous, leading therefore to different preferable strategy combina-tions.)c. Changing a particular main component of the administrative sys-tem so as to meet some ideal image. This again includes a large varietyof main goals, such as: introduction of a merit civil service; intro-duction of a planning-programming-budgeting system; moving to-wards automated data processing and integrated data and informationbanks; increasing the utilization of scientific knowledge; and manymore.These three goal clusters are intra-administration directed, at least intheir direct effects and their initial facade. Much more important and

often tacitly underlying such intra-administrativegoals are administrativereform goals dealing with the societal roles of the administrative systemand its interrelation with other components of the policymaking systemand even of society as a whole. Goal clusters of this type include, forinstance:d. Adjusting the administrative system to advancing overriding so-cietal objectives, such as accelerated modernization, or war.e. Changing the division of labor between the administrative system(and especially its higher levels) and the political system. For instance,either in the direction of reducing the power of the senior civil serviceand making it a more obedient servant of the political processes orincreasing the professional autonomy of the administrative system andstrengthening its influence on policy.f. Changing the relations between the administrative system and thepopulation or selected population segments. For instance, throughrelocation of decision centers (centralization vs. decentralization),democratization in the sense of participation, and democratiza-tion in the sense of making the personnel composition more reflect-ive of various population segments.Administrative reform can also serve purely political functions, such

7/28/2019 j.1467-7660.1971.tb00326.x

6/17

24 YEHEZKELDRORas making administration serve as a scapegoat, gaining support throughdistributing spoils, receiving a foreign loan, etc. While abhorred by tra-ditional administrative theory, such functions are sometimes essential,and sacrificing the administrative system may sometimes be justified toavoid an even greater evil - hough in most cases the longer range costsof such a reform are by several magnitudes larger than the benefits.The list of six goal clusters, each one of which covers a variety ofmain goals, does not exhaust the list of possible and reasonable overallgoals of adminstrative reform. But it is, I hope, sufficient to concretizethe concept of overall administrative reform goals and to demonstratethe dependence of p referable reform features on the overall goals of therespective reform, and therefore the necessity to clarify and clearly de-cide what goal-mix a particular reform should be aimed a t initially (whilerecognizing that the goals will change with time). Indeed, this is a mostimportant administrative reform strategy dimension which is seldommore than touched upon in most actual cases of administrative reformdecisionmaking.

11. REFORM BOU NDA RIESPublic administration is a complex system, which itself is a subsystemof a number of larger systems - such as the political system and thesocietal system as a whole. Simultaneously, various subsystems of theadministrative system are also components of a variety of other systems.Thus , civil service staff schools are also components of the educationsystem; professional staff members are also components of professionalassociations; administrative employees are also components of employeeorganizations and of the manpower system as a whole; and so on. Fur-ther to complicate the boundary issues, there are a multiplicity of unitswhich are in between the administrative system and o ther systems. Thus,mixed corporations are in between public administration and privateenterprise: private consultants who work for public adm inistration, in-dependent research facilities working on con tract for government, privateproducers serving as main suppliers for public administration - heseare additional illustrations of entities not easily classified as either insideor outside the administrative system. The c itizen in his role as an objectof administration, a client of administration, and (in the aggregate andin some senses) as the master of public administration, also is and is nota component of the administrative system, depending on our definitioncriteria and boundary postulates.

7/28/2019 j.1467-7660.1971.tb00326.x

7/17

STRATEGIES FOR ADMINISTRATIVE REFORM 25All these abstract observations on the undefined boundaries of theadministrative system relate to a very real and practical administrativereform strategy dimension: What are the boundaries within which anadministrative reform is to be confined? In other words, what is thedomain of institutions to be considered as appropriate objects for anadministrative reform?5A brief exploration of some concrete boundary issues often met inadministrative reforms will serve to operationalize this strategy dimen-sion and to demonstrate its applied importance:a. Politicians and political institutions. The interface between politi-cians and political institutions on the one hand and the administrativesystem on the other is so close, that many aspects of the adminis-trative system cannot be changed without also changing some politicalinstitutions. Thus, the spoils system and corruption cannot be dealtwith through changes limited to administration. Furthermore, in re-spect to many other aspects of the administrative system, interactionbetween it and political institutions is so intense that change limitedto the administrative system is useless and often even counterproduc-tive. Thus, upgrading the professionalism of senior administratorswhileleaving the capacities of politicians and political institutions low willchange the balance of power between administration and politics indirections which may often involve undesirable alienation and de-democratization. Especially when the goals of administrative reforminvolve improvement of policymaking, administration and politicsmust be regarded as intensely interacting components of the policy-making system which must be subjected to directed change as awhole.6 Indeed, one of the widespread weaknesses of administrativereforms is neglect and inability to bring about adjustments in closelytied aspects of politics, without which the administrative reform can-not achieve its goals.b. Academic institutions. A main factor shaping the quality of civilservants is the qualifiation of candidates for recruitment. The qualifi-cations of candidates for recruitment in turn are largely shaped by the

6 A different question also involving a boundary issue is: within what domainshall we look for relevant reform consequences? Related are the questions: whatshall be regarded as first order consequences? as second order consequences? asthird order con sequences? etc. These are important problem s, but they would carryus to o far into general policymaking theory to be treated in this paper.8 Compare Yehezkel Dror, Accelerated Development and Policymaking Im-provement, Civilizarions, Vol. XIX, No. 2 (1969), pp . 204-15.

7/28/2019 j.1467-7660.1971.tb00326.x

8/17

26 YEHEZKEL DRORoutpu t of academic institutions (and com peting demand). Here is amain interface between the administrative system and academic in-stitutions, which demonstrates the importance of sometimes consider-ing academic institutions as within the boundary of adm inistrativereforms. Similarly, possible functions of academic institutions as cen-ters for post-entry training, for research and for consultation, provideadditional reasons for carefully determining how far academic in-stitutions should be regarded as within the boundary of an adminis-trative reform.c. Legal institutions. Qu ite important are the roles of legal institutionsvisd-vis the administrative system. Usually recognized are the func-tions of legal institu tions in (1) controlling adm inistrative discretion,especially when the rights of individuals are involved; and (2) enforcingrules of conduct, such as criminal sanctions agains t corruption. Notless important, though less discussed, is the transferability of importan tsocial functions between courts and administration. Thus, control ofutilities, medicines, food, etc. can either be entrusted to the adminis-trative system or to the legal institutions, or to some mix between thesetwo. Therefore, clarification of the ex tent to which an administrativereform can also involve changes in the legal institutions is important.Such changes can include both new statutory laws an d/o r new courtstructures - uch as establishment of administrative tribunals andquasi-judicial units. While, therefore, the importance of legal in-stitutions as one of the boundaries of adm inistrative reform s shouldbe recognized, it is necessary at the same time to reject the tendencyin some countries to view all adm inistrative reforms m ainly from alegal perspective. A purely legal perspective is much too narrow andshortsighted to be useful for administrative reform.d. The public. More diffuse, but most important, are the problemsof including som e aspects of the public within the boundaries of anadministrative reform . Introduction of new procedures involving muchcontact with the public (such as computerization of income tax orsocial security), fighting corruption, involving the public more inparticipa tory decisionmaking, changing the recru itment base of thecivil service- hese are only some illustrations ou t of many of adminis-trative changes, the success of which depends inter uliu on changes inthe public. Ideological and political considerations usually cause us tobe suspicious of attempts to influence the public; therefo re very ca re-ful treatm ent of the boundaries of an adm inistrative reform with relatedaspects of public behav ior, opinions , and attitudes is essential.

7/28/2019 j.1467-7660.1971.tb00326.x

9/17

STRATEGIES FOR ADMINISTRATIVE REFORM 27This strategy dimension is closely related to another which I will dis-cuss later, namely, range of reform instruments. Together with overallgoals, the boundaries of an administrative reform really constitute a mostbasic decision shaping all features and elements of the reform. Thereforeit is so important not to take any a priori image of boundaries as granted ,but explicitly to consider and reconsider this strategy and decide it on thebasis of careful analysis before and while more detailed reform issuesare taken up.

111. PREFE RENC ES I N TIMETh e question of the targe t time - when does one wish main results ofadministrative reform to be produced - s another main strategy issue.There is a fundamental difference between a reform which wants toachieve savings, or more subordination to politics, or better realizationof accelerated development immediately, and a reform which wants toupgrade the capabilities of an adm inistrative system over the next ten totwenty years. This strategy dimension clearly brings out a characteristicshared by all strategies, namely, the necessity for iteration and recon-sideration of strategy decisions in the light of detailed reform plans andreform results. Thus, often the a priuri desire will be for fast results, whilemore detailed reform plans will make obvious that significant resultsneed m ore time, with consequent reformulation of the time preferences .Nevertheless, the often positive correlation between achievable resultsand required time and its sensitivity to availability of refo rm resourcesand reform instruments is quite clear, and it is wasteful to relearn throughbitter experience what is already wellknown. Also, external variables(political feasibility, critical survival needs, etc.) often establish rigidcons traints on time availability. Therefore, early establishment of cleartime preferences in respect to various reform outputs is essential as adirective for detailed reform plans. Even more important, early identifi-cation of rigid time constra ints is essential for strategy consistency, asoverall reform goals must often be reduced because of non-elastic timepreferences.

IV. RISK ACCEPTABILITYThis strategy dimension involves the degrees of risk to be accepted in theadministrative reform. Closely correlated with the next strategy dimen-sion, the issue is largely one of readiness to accept the higher risks associ-

7/28/2019 j.1467-7660.1971.tb00326.x

10/17

28 YEHEZKELDRORated usually with more innovation, or whether one prefers lower risksoften associated with more incremental changes. In theory of gamesterminology, the choices here are between maximax on the one hand andmaximin or minimax on the other hand. (But one should bear in mindthat the risks of maintaining the present situation may sometimes bevery high.) Also involved are preferences between average expectedvalue, lottery value, and similar choice principles between differentforms of risk parameters. However abstract, this is a very importantchoice, especially in view of the tendencies of risk avoidance commonlyfound in organizations on the one hand, and the tendencies towardsrecklessness often fmnd in administrative reforms motivated by sweep-ing social movements and political changes.8 Therefore , explicit judgmenton aceptable risks can significantly improve reform decisionmaking.Furthermore, explication of acceptable and expected risks carries withit important implications for detailed reform plans by encouraging suit-able hedging, contingency planning and similar risk-absorbing method-ologies.V. INCREMENTALISM VS. INNOVATIONThis dimension deals with the choice between various degrees of changein the administrative system (defined in terms of extent of change, scopeof change, and time), ranging from small incremental change of a fewadministrative details over a long period to fargoing, comprehensive andrapid administrative system redesign and even novadesign (i.e., designanew). One main variable in this choice is risk acceptability: the moreinnovation the higher often are the unpredictable c0nsequences.QWhatis often forgotten is that a higher probability of fargoing improvem ent inoutputs is also positively correlated with fargoing innovation, the latterbeing an essential (though by itself insufficient) condition for the first.Also often forgotten is the already mentioned possibility that the status7 For a stim ulating and readable treatment, see Howard Raiffa, Decisions Anal-ysis: Introductory Lectures on Cho ices under U ncertainty (Reading, Massachusetts:Addison-Wesley, 1968).6 Nevertheless, appearance may be misleading. Thus, upon closer study, it is clearthat the Nazi leadership was very careful n ot to disrupt German bureaucracy as awhole, but rather focused its resources on selected m inistries, especially police andyouth.0 This point, as well as some other advantages of incremental change (such asintellectual mana geability and political feasibility) are best discussed in David Bray-brooke and Charles E. Lindblom, A Strategy of Decision (New York: Free Press,1963).

7/28/2019 j.1467-7660.1971.tb00326.x

11/17

STRATEGIES FOR ADMINISTRATIVE REFORM 29quo and limited change may, in some dynamic conditions, be even morerisky than fargoing change. Strong organizational tendencies towards in-cremental change and scarcity of good administrative inventions makeexplicit strategy determination all the more essential for innovative ad-ministrative reforms; significant and carefully considered innovations (ascontrasted with convulsive jumps) usually are not a natural organiza-tional phenomenon and therefore need specific encouragement, whichusually does not come forth without an explicit strategy decision. Whilethus often leading towards more innovation, specific consideration of thepros and cons of innovation in administrative reforms and the re latedrisks may sometimes also serve to restrain recklessness and encouragemore careful strategies. This may involve either a more incrementalstrategy or an effort to combine innovation with risk-reducing method-ologies, such as pilot experimentation and sequential decisionmaking(i.e., decisionmaking by phases coupled to constant learning).

VI. COMPREHENSIVENESSVS. NARROWNESSThis strategy dimension deals with the degree to which the administrativereform should deal with a broad range of administrative system com-ponents or should focus on a few or even a single component. Conditionedby the reform boundaries on one side and determining the range withinwhich reform instruments may be searched for on the other side, thedegree of comprehensiveness vs. narrowness is nevertheless a distinctstrategy dimension. Thus, a reform dealing with personnel policy andorganizational structure and information processing will be comprehen-sive (even though w ithin the traditional boundaries of an administrativesystem), while a reform dealing only with recruitment for the senior staffwill be narrow er, even if it adopts broader boundaries and includes aca-demic institutions and recruitment-influencing public images within itsscope. (This illustration also shows that more com prehensive does notnecessarily mean m ore important or more significant.) This dimen-sion is conceptually important for an additional reason: together withthe degree of innovation, it categorizes an instance of directed adminis-trative change as an administrative reform, as defined by me. We cannow postula te that to qualify as an administrative reform, a case of direct-ed administrative system change must either (a) be of at least mediumcomprehensiveness plus high innovativeness; or (b) be of high compre-hensiveness plus at least medium innovativeness. (A third case to bediscussed soon is fargoing change of a single component, but one which

7/28/2019 j.1467-7660.1971.tb00326.x

12/17

30 YEHEZKELDRORhas overriding influences in its administrative system as a whole. Forinstance: new recruitment patterns into the sen ior staff.)

Innovativeness(intensity of change)

L O W

MediumHigh

Comprehensiveness(Scopeof Change)

L O W Medium HighIorderline

of Administrative ReformStill left open is the operational definition and measurement of low,medium and high, and the development of more advanced classifi-cation scales (ordinal or non-metric) than the administrative reform/adrninistrative-change-short-of-administrative-refo~ inary classifica-tion. Our strategy dimensions permit some progress in this direction.With the help of the two dimensions conditioning the scope and intensityof administrative change, we can at least order administrative change on afive-point ordinal scale: low-low; ow-medium or medium-low; medium-medium; high-medium o r medium-high; and high-high.VII. BALANCE ORIENTED VS. SHOCK ORIENTEDThe question here is, in how fa r a reform should be directed at changingadministrative system components in a mutually coordinated (bal-anced) way, o r whether a few or single components only are subjectedto fargoing change with the purpose to shock the system into a new state-of-existence - uch as increased openness to change and derou tiniation.This strategy dimension is closely related to the comprehensivenessvs. narrowness dimension,in the sense that only a comprehensive adminis-trative reform can be balance oriented and only a narrow administrativereform can be shock oriented. But it is a separate dimension, becausenot every comprehensive reform is o r should be ba lance oriented, andnot every narrow administrative reform is or should be shock oriented.

This strategy dimension is especially important because it raises forexplicit examination the widely accepted notion that every good ad-

7/28/2019 j.1467-7660.1971.tb00326.x

13/17

STRATEGIES FOR ADMINISTRATIVE REFORM 31ministrative reform must be balance oriented in the sense of striving forsynchronized change of a multiplicity of components. This a priori as-sumption is strongly reinforced by much of general systems theory,systems analysis, and planning theory (where the term comprehensiveplanning means both comprehensive, in the sense of broad scope, andbalanced).While the idea of balanced administrative reform appeals to our pref-erences for harmony and is reenforced by a number of influential tacittheories (e.g., biological analogues of the administrative system) in ad-dition to the just-mentioned decision approaches, it suffers from a num-ber of serious fallacies.10 Often very harmful is the innovation-dampingeffect of an effort to achieve balanced reform, because radical changewill often appear to be incompatible with balance. Furthermore, to pre-serve balance, we must be able to predict the consequences of directedchange - and this again pushes us to limit change to incrementalism,which is easier to predict than fargoing innovation. When what is aimedat is a radical transformation of the administrative system, shock effectswhich first disbalance the system so as to open it up for redesign andnovadesign may often be the optimal strategy, though admittedly a riskyand expensive one. Again, explicit consideration of this strategy dimen-sion should be useful in two directions: (a) by helping to overcome theu priori preference for balanced change in situations where adminis-trative reform is approached from a conservative point of view; and (b)by explicating the risks of shock change and bringing out the need forrisk-reducing measures in situations in which administrative reform isapproached from a radical and even revolutionary point of view.

VIII. RELEVANT ASSUMPTIONS OF THE FUTUREThe dependence of preferable administrative systems on a large numberof exogenous variables results in the necessity to base every reform onlo Compare Yehezkel Dror, ComprehensivePlanning: Com mon Fallacies VersusPreferred Features, in F. van Schagen (ed.), Essays in Honour o f Professor Jac.P . Thijsse (The Hague: Mouton and Co., 1967), pp. 95-99.11 Compare Albert Hirschman, The Strategy of Economic Development (NewHaven: Yale University Press, 1958), which provocatively discusses the strategy ofdeliberately unbalancing an economy in order to achieve accelerated growth.It is interesting to consider the role of compreh ensive and innovative adminis-trative reform as a shock strategy directed at the societal system as a whole. Thisis an important possibility to be considered among the overall goals for adminis-trative reform .

7/28/2019 j.1467-7660.1971.tb00326.x

14/17

32 YEHEZKELDRORassumptions concerning the expected state of those variables. To illus-trate: the state of administrative technologies (such as automated dataprocessing), availability of qualified manpow er, internal politics, externalpressures and w ars, and changes in societal goals and functions to beserved by the administrative system are som e of the variables the fu tureof which must be taken into account when a preferable reform plan isbeing developed.12This problem is especially importan t in administrative reforms with alonger-range time preference, but exists even when the main time prefer-ence is for the nearer future. Th e usual practice is not to engage in syste-matic efforts to try and identlfy possible alternative futures of relevantvariables, but ra ther either to assume that what exists now will continue inthe future, or to engage in arbitrary guesses on the future based on acci-dental information and personal bias. This is the result of the absence inmost countries (including nearly all modernizing countries) of any effortsto predict alternative futures, and of the lack of any interconnection be-tween lookout efforts and administrative eform decisionmaking in the fewcountries where some lookout mechanisms (in government or outside it)do exist. Here we run again into the already mentioned dilemma, namelythat good administrativereform requires a highly developed policymakingsystem, which in turn usually requires fargoing administrative systemredesign and novadesign. Nevertheless, much can be achieved even with-out establishmentof highly sophisticated outlook institutes.IX. THEORETIC BASESTh e importance of bringing hidden assumptions which are critical for anadministrative reform out into the open , to subject them to conscious re-examination and, if possible, to improve them with the help of systematicknowledge, structured rationality, and organized creativity - his is oneof the m ain intellectual justifications for proposing s trategy dimensionsas a framework for administrative reform decisionmaking. Nowhere isthis motive more important than in respect to the theoretic bases of anI* When we regard some of the futures of some of the variables as endogenous,that is, as themselves ob jectives for administrative system operations directed atinfluencing the futu re, then we get into one of the possible goals of administrativereform - namely, increasing the capacity of an administrative system to shape thefuture. See Y ehezkel Dror, Some Requisites of Organizations Better Taking IntoAccount the Future, in Robert Jungk and Johan Galtung (eds.), Mank i nd 2000(London: Allen &Unwin, 1969), pp. 286-90, and The Role of Futures in Govern-ment, Futures, Vol. 1, No. 1 (September 1968), pp. 4 0 4 6 .

7/28/2019 j.1467-7660.1971.tb00326.x

15/17

STRATEGIES FOR ADMINISTRATIVE REFORM 33administrative reform. Psychology brings out the fact that decisions andjudgments are strongly influenced, among other factors, by the tacittheories held by decisionmakers. Organization theory supports this find-ing in respect to organizational decisionmaking, where collectively heldtacit theories shape group and organizational decisions. Tacit theoryhere means unexplicated believed-in explanations of behavior and otherphenomena; in short, subjective images on what makes the world tick.Widespread elements of tacit theories in respect to administrative reformsinclude, for instance: belief that a merit civil service makes a lot of differ-ence; belief that civil servants are motivated by monetary incentives, byappeal to patriotism, etc.; a view of administration inwhich some com-ponents, with which one was personally involved, are regarded as im-portant reality-shaping variables; prototype images of politicians andclients; and so on. What distinguishes all tacit theories is (a) by definition,that they are not subjected to the tests of explicit consideration; (b) thatmany of them are rejected by the persons who held them, once they areexplicated; and (c) that they lag behind available scientific findings, com-parative knowledge and conclusions of distilled experience. Therefore,this strategy dimension serves first to bring out into the open at least someof the tacit theories underlying proposed reform plans, and second, totry and improve them by self-consideration and by trying to relate themto explicated theories, to available knowledge, and to processed ex-perience.

X. RESOURCE AVAILABILITYOn a different level is the strategy dimension of resource availability.Here, too, explication is needed and useful, to bring up for considerationthe often unrealistic hidden assumptions concerning resources (not onlymoney, but also qualified staff, equipment etc.) which will be availablefor the reform, or, still worse, to expose widespread lack of concern formany cost elements of a proposed reform. But in respect to this strategydimension some clear decisions are called for, concerning the resourcesto be allocated for the reform. Both indirect and direct costs must betaken into account and the time distribution of resource availability mustbe worked out, as important constraints on reform. While iteration mayresult in some revisions in available resources, more often constraints onresources are quite rigid, and clear establishment of reform budget ceil-ings will have strong influence on the reform plans. In particular, it mayavoid sudden ruptures of reform activities because the tacitly expected

7/28/2019 j.1467-7660.1971.tb00326.x

16/17

34 YEHEZKEL DRORresources do not become available. True, sometimes explicit resourceavailability strategies may prevent a reform which otherwise may havebeen executed through piece-by-piece allocations, but I think that all inall, more realistic administrative reform planning will result in morecases of successful administrative reform.

XI. RANGE OF REFORM INSTRUMENTSThis last of the strategy dimensions to be discussed in this paper sums upin a sense all other strategies, while adding to them an additional per-spective. Rational reform planning means adoption of a mix of reforminstmments which, within the given constraints, provides the highestprobability of better approximating the overall reform goals. It is, there-fore , in the selection of reform instrum ents and their setting (i.e., m anip -ulating in the objective sense of the term, without the negative overtonesand connotations) that the strategies are transformed into reform plansand reform action. An essential requisite for selecting a preferable mixof reform instruments and of their settings is to work out first a list ofavailable reform instruments with the ir main characteristics, in term s ofbenefits-costs. A distinct effort to develop a taxonomy and characteriza-tion of administrative reform instruments as one of the strategy decisionsis all the more necessary for three main reasons: (a) the tendency toignore many available reform instruments because of tacit theories, limit-ed information, and inertia; (b) the potential benefits of identificationand invention of ignored or unknown reform instrum ents, some of whichmay prove to be very effective and efficient; and (c) the im portant feed-back from reform instrument evaluation to other strategy dimensions,for instance by showing the necessity to broaden reform boundaries so asto include additional sets of instruments within the domain of reform

Th e last strategy dimension serves to reemphasize the relations betweenpolicy sciences as a whole, administrative sciences research, and reformpractice. The identification of policy instruments for directive humanaction is one of the main tasks of policy sciences. The evaluation of tha tsubset of policy instrum ents which may be available for administrativereforms in terms of relevant benefits-costs belongs to administrativesciences. Reliable mappings of possible administrative reform instru-ments with condition-sensitive evaluation of their benefits-costs (eachstrategy by itself and in various mixes) will be very helpful in concre teadministrative reform endeavors.

planning.

7/28/2019 j.1467-7660.1971.tb00326.x

17/17

STRATEG IES FOR ADM INISTRATIVE REFORM 35Even in the present underdeveloped state of policy sciences and ad-ministrative sciences, the concept of administrative reform strategiesshould be useful - as a sensitizer, as an heuristic aid, and as a guideline.This, I hope, has been demonstrated (though not proved) with the helpof the eleven administrative reform strategies discussed in this paper.What is urgently needed is intensive study and research which will permitelabora tion of various strategies in terms of applicability, effectiveness,and e fficiency under different conditions. For such prescription -orientedadministrative sciences research, the concept of reform strategy mayserve as a main foous.