Policy Analysis - Thomas Dye

9
POLICY ANALYSIS what governments do, why they do it, and what difference it makes 1 .!? 1 ti d 13IMQ-? ' 0 (vickyl---j(-04...,),-K-Q. Fts^.60k- rbc-a-s- George Bush takes the oath of office as president of the United States. (AP/Wide World Photos) 1

Transcript of Policy Analysis - Thomas Dye

Page 1: Policy Analysis - Thomas Dye

POLICY ANALYSISwhat governments do,

why they do it,and what difference it makes

1

.!?

1 tid13IMQ-?' 0 (vickyl---j(-04...,),-K-Q.

Fts^.60k- rbc-a-s-

George Bush takes the oath of office as president of the United States. (AP/Wide World Photos)

1

Page 2: Policy Analysis - Thomas Dye

I

2 Policy AnalysisPolicy Analysis

POLICY ANALYSIS IN POLITICAL SCIENCE

This book is about public policy. It is concerned with w:Lacgovernments do,why they do it, and what difference it makes. It is also about political scienceand the ability of this academic discipline to describe, analyze, andexplainpublic policy.

Definition. J Public policy is whatever governments choose to do or not to do.' Fovernments do many things. They regulate conflict within society;they organize society to carr on conflict with other societies; they distributea great variety of symbolic rtfry AVria. (Material services to members of thesociety; and they extract money from society, most often in the form of taxes. Thu/public pofirrer–r y regtor, organize bureaucracies,distribute benefits, or extract taxes—or all these things at once.

Scope. Governments in the United States directly allocate about 35percent of the Gross National Product, the sum of all of the goods and ser-vices produced in the nation each year. About two-thirds of the governmentalsector of the GNP is accounted for by the federal government itself; the re-maining one-third is attributable to eighty thousand state, city, county, town-ship, school district, and special district governments combined. Overallgovernment employment in the United States comprises 16 percent of thenation's workforce.

Public olicies ma y deal with a wide varlet of substantive areas—p et_er y, environment, foreign affairs, education, welfare, police,

hi hwa s taxation, housin g; social security health economic opportunity,urban develo ment, inflation and recession, and so on. They may range fromthevital to the trivial—from the al ocation of of billions of dollars fora mobile missile system to the designation of an official national bird.

Political Science. Public olic is lot a new concern of • lithe earli st writings of political philosophers reveal an interest in the policies

uis e y governments, the forces shaping these policies, and the impactof these policies on society. Yet ►e major focus of attention of politicalscience has never really been on policies themselves, but rather on the in-stitutions and st,-uctures of government and on the political behaviors andprocesses associateil with policy making,.

"Traditional" political science focused its attention primarily on theinstitutional structure and philosophical justification of government. Thisinvolved the study of constitutional arrangements, such as federalism, sw-aratioriof_power, and judicial review; powers and duties of official bodiessuch as Congress, president, and courts; intergovernmental relations; and

X-E-oryN.

the or 1 o eration e, executive, and judisialagen-cies. Traditional studies described the institutions in which public policy wasformulated. But unfortunately the linkages between important institutionalarrangements and the content of public policy were largely unexplored.

Modern "behavioral" political science focused its attention primarily.on the processes and behaviors associated with government. This involved the study of the sociological and psychologiCal bases of individual and group (C.behavior; the determinants of voting and other political activities; the func-tioning of interest groups and political parties; and the description of variousprocesses and behaviors in the legislative, executive, and judicial arenas.Although this approach described the processes by which public policy wasdetermined, it did not deal directly with the linkages between various pro-cesses and behaviors and the content of public policy.

5 1.6-R La mow, cse...ct rt.fpc. _

Policy- tuc Today ►y_ri,olitical scientists have shifted the focusCSoct. et..vrer•km.... palvlar.)(-0 • ,..

/ tp_public o th–e– scrit:decau.ses and consequencesotgotyLvaent activity. This focus involves a description of the content o publicpolicy; an analysis of the impact of social, economic, and political forces onthe content of public policy; an inquiry into the _effect stitu-tional arrangements and political processes on public policy; and arrevalua-

jiln olicies on society, in terms of both, expected and unexpected consequences.

DEFINING PUBLIC POLICY: PLAYING WORD AMES

FrrAAADavy, CX-D /-0

'Sec insert, "Defining Public Policy: Playing Word Games."

a P P . fr)

mental actions and an overall program of action toward a given goal. But the prob-lem raised in insisting that government actions must have goals in order to belabeitsd "policy" is that we can never be sure whether or not a particular actionhas a goal, or if it does, what that goal is. Some people may assume that if a govern-ment chooses to do something there must be a goal, objective, or purpOse, butall we can really observe is what governments choose to do or not to do. Realistical-

CV) i vin li .41.( CAA-0 OC( r"'") C/6/A.l) LC1-9 =A-

otA p. pc) •This book discourages el, orate academicdiscussions of the definition of public

policy—we say simply that public policy is whatever governments choose to door not to klo.Books, essays, and discussions of a "proper" definition of publicpact' have proven futile, even exasperating, and they often divert attention frontthe study of public policy itself. Moreover, even the most elaborate definitionsof public policy, upon close exa • • it to boil down to the same thing.For example, political scient st David Easton defintspitheauthoritative allocation of values c society"—but it turns out that on-lx, tlu%government can "authoritativelLat on the "whole" society, and everything.

'the ro y nt .houses t or not to f values"Political scion st)Tlarold L.asswel nd philosoph define

policy as "a projected program of goals, values, and practices, an political scientistsays, "ILts_en there be a goal, ob.

, or purpose." These definitions imply a difference between specific govern-

Page 3: Policy Analysis - Thomas Dye

Policy Analysis

ly, otriiiiitai7157r6rtc policy must include all actions of government and notwhat governments or officials say they are going to o. e may wis tat govern-ments act in a "purposeful goal-oriented" fashion, but we know that all toofrequently they do not.'

Still another approach to defining public policy is to break down this generalnotion into various component parts. Political scientist Charles O. Jones asks thatwe consider the distinction among various proposals (specified means for achievinggoals); programs (authorized means for achieving goals); decisions (specific ac-tions taken to implement programs); and effects (the measurable impacts of pro-grams). But again we have the problem of assuming that decisions, programs, goals,and effects are linked-. Certainli in many policy areas we will see that the decisionsof government have little to do with announced "programs," and neither are con-,nected with national "goals." It . may be unfortunate that our government does,not function neatly to link goals, programs, decisions, and effects, but as a matter.of fact, it does not.

Political scientis

behaviors consistent and repetitiveness on the •art ofIt and those who abide by it." Now certainly it would be a wonderful thing if govern-ment activities were characterized by "consistency and repetitiveness"; but it isdoubtful that we would ever find "public policy" in government if we insist onthese criteria. Much of what government does inconsistent and nonrepetitive.

So we shall stick with our simple definitioni public policy is whatever governmentsnot to do>1ote that we are focusing not only pn government action,

but also on government inaction, that is, what government chooses clo. Wecontentl ent inaction can havej,t2st as great an iLact on societ • asgovernment action.See David Easton, 'the Political System (New York: Knopf, 1953), p. 129; Harold U. 1.asswelland Abraham Kaplan, Power and Society (New Haven: Yale University Press, 1970), p. 71; CarlJ. Friedrich, Man and His Government (New York: McGraw-Hill, 1963), p. 70; Charles ()Jones,An Introduction to the Study of Public Policy_(11oston: Duxbury, 1977), p. 4; I kin/. Etilau'andKenneth Prewitt, Labyrinths of Democracy (Indianapolis: Bobbs-Merrill, 1973), p. 465; and Fishileclo, "Policy Analysis," British Journal of Political Science, 2 • .y 1972), p. 85.

WHY STUDY PUBLIC POLICY?

Policy Analysis 5

can ask ‘,11,11 impact ■ ittlilie policy has on societ y and its political system.By asking such questions we can improve our understanding of the linkagesbetween socineconotnic forces, political processes, and public policy. Anunderstanding of these linkages contributes to the breadth, significance,reliabilit y . and theoretical development of social science.

12.e.mYyt,eft-NatoProblem Soteing. Public policy can also be studied foirprofessiatual reasons:1 9

understanding the causes and consequences of public policy permits ys toapply social science knowledge to the solution of practical problems. Fac-6.71. Tnowledge is a prerequisite to prescribing for the ills of society. If certainends are desired, then the t uestion of what oli i would best in-7177.7

re airing scientific studthese ends is factual uestion, in ten

c(;co i-v)titi^° C<,(

Policy Recommendations. Finally, public policy can be studied for(political inirpo.Q1 to ensure that the nation adopts the "right" policies to

achieve the "right." goals. It is frequently argued that policial science shouldnot be silent tit- impotent in the face of great social and political crises, andthat polii ical scientists have a moral obligation to advance specific publicpolicies. An exclusive torus on institutions, processes, or behaviors is fre-quently looked upin I ;I s "dry," -irrelevant," and "amoral," because it doesnot direct hill to t he really important policy questions facing Americansociety. Policy :;:.lidi•s can he undertaken not only for scientific and profes-sionalp■i_Eposes lint also to inform >olitical discussion, advance the level ofpolitical ait\\ c •H■liev. Of course, these ,are very su vet l ye purposes—Americans do not always agree on what con-stitutes the "right" policies or the "right" goals—but we will assume thatknowledge is preteralde to ignorance, even in politics.

QUESTIONS IN POLICY ANALYSISar.vvvot/A,What can 1\ C I C;I 1 I about public policy?

o(A

tion of public polic : o is is define

choose to do or

t supply still another defini-anding decision' characterized by

ke0 • •

olicy sttue".* an produce rofession statements, about (limy to achieve desire goals)

In other words,of "if ... then ..."

Description. First of all, we can descrthei public policy we can learn what government is doing (arid not doing) in tts;jfa defnn e, education, civil

leak h. the environment, taxatio d so on, ann. A factual basts of intoraTinon narT7rTicy is realty- an indispensable part of everyone'seducation. What does the Civil Rights Act of 1964 actually say aboutdiscrimination in employment? What did the Supreme Court rule in the Bakkecase about affirmative action programs? What is the condition of,the na-tion's social security program? What do the Medicaid and Medicare programspromise for the poor and the aged? What is the purpose of the MX missile,

Why should political scientists devote greater attention to the study of public,_policy?

Selena ic Understanding. First of all, public policy can be studied forpurelytpentificraiTajuiderstand ing the causes and consequences of policydecisions im roves our knowledge of societ . Public can beviewedas a. and we can ask what socioeconomic forces an.pplitirn1 sysw characteristics operate to shape the onto t of policy. Alter-native , public policy can be viewed as an inde endent variable, and we

i9OCZA1 Tx-ZIA' LaA., zawo ao,A, oci/i-At.3.,a,cA , d.c, — SA- )-i---A- 4 °'4 //743°1CLA7)VOWS - C•Wiel.)— -RX.A.A.6 en-,

OR P eoreto acAik-i rldP • - A-54IfYNa414)4. "'QC- '24)1'4A) 1(evx:A e...r.4.0 va/1/s. ;--1...,0(4.7,) .ji-4141.A

GA/) . .A-1'‘CA-kri) •

(I)

Page 4: Policy Analysis - Thomas Dye

policy Analysis 7

to attack critical issues with the tools of systematic inquiry. There isan implied assumption n policy anal sis that develo in se • is knowled eabout the forces shaping public r ictan the consequences of ublic cp21isyis itself a sociali=-elevant activity.. and that such analysis is a prerequisiteto rescri bon v c and activism.

0 j.4.- Specifically, public analysis involves:

1. A primary concern with explanation rather than prescription. tong, recommenda-tions—if they are made at all—are subordinate to description and explana-tion. There is an. implicit judgment that understanding is a prerequisite to pre-,scription, and that understanding is best achieved through careful analysisrather than rhetoric or polemics.A L-ori.s search for the causes and conse uences o u • This search in-volves the use of scientific scan ar s o in erence. Sophisticated quantitativetec e p u in establishing valid inferences about causes andconsequences, but they are not really essential.

3. An effort to develop and test general propositions about the causes and consequences ofpublic-policy and to accumulate reliable research :ndin s o ener e. The ob-ject is to develop general theories about pu is policy that are reliable and thatapply to different governmental agencies and different policy areas. Policyanalysts clearly prefer to develop explanations that fit more than one policydecision or case study—explanations that stand up over time in a variety ofsettings.

6 Policy Analysis

en)the B-2 bomber, or the Trident submarine? flow much arc we really spend-7 ing on national defense and on social welfare? How much money are we

do paying in taxes? How much money does the federal government spend eachyear? These are ( xamples of descriptive questions.

Viipublic olicy. Wiyispal gover o whatCauses. Second, we can inquire about (he cause 6r determinant of

they do? We might inquire about the effects of political institutions, processes,and behaviors on public policies. For examp le: nake any difference)

r7 iiiiix and spending levels whether Democrats or Republicans control the

Ore (I presidency and Congress? What is the impact of interest group conflict onfederal aid to education? What is the impact of lobbying by the special in-terests on efforts to reform the federal tax system? We can also in uire aboutthe effects of social, economic, and cultural forces in s aping public policy.For example: What are the effects of changing public attitudes about raceon civil rights policy? What are the effects of wars and recessions on govern-ment spending? What is the effect of an increasingly older population onthe social security and Medicare programs? In scientific terms. when} the causes of public .olic , policies become the de iendent variables, and their

olitical, social, economic, and cultural determinants become theriu ependent varia es.various

Consequences. Third, we can inquire about t e consequences,of public policy. What difference, if any, does pub lcple's lives? For example: Does capital punishment help to deter crime? Arewelfare programs a disincentive to work? Is busing an effective means ofending racial inequalities in education? Do liberal welfare benefits resultin larger numbers of poor people? Does increased educational spending pro-duce higher student achievement scores? In scientific terms, when we studythe consequences of public policy, ` policies become the inde endent variables,and thew political, social,econotnic, zuld cultura impacts on society become •

e -dependent verrirble.t.

POLICY ANALYSIS AND POLICY ADVOCACY

It is important to distinguislitratcy ana/ysii from Explainingthe causes and consequences of various policies is not equivalent to prescrib-ing what policies governments ought to pursue. Learning why governmentsdo what they do and what the con er uen s of th%La,..cs ons are is riot thesame as saying what governments g t o or brThw iti about changes in:what they do. Polic advocac re uires the sorganization, and ac olicy analysis encourages scholars and students

"012 wrna(-0POLICY ANALYSIS IN ACTION: ACHIEVINGEDUCATIONAL OPPORTUNITY

One of the more interestin exam les of olicy analysis over the ears hasbeen the social science research one ua educational opportunit and howto achieve it. "Educational opportunity" has been one of the most controver-sial topics in American politics, and social science has played an importantrole in policy making in this area. However, as we shall see, the more contro-versial the policy area, the more difficult it is to conduct policy research.

1214A`Cl°Early Research—the Coleman Report

The early landmark research on educational opportunity in Americawas sociologist James S. Coleman's Equality of Educational Opportunity, fre-quently referred to as the "Coleman Report."' The Coleman Report dealtprimarily with the consequences of educational policy—specifically, the im-pact of schools on the aspiration and achievement levels of pupils. Although

'James S. Coleman, Equality of Educational Opportunity (Washington, D.C.: GovernMent Print-ing Office, 1966).

s • r tonic, persuasion,

CA-CA-1 AAg.A Wm 4-)1,v)7{A eYx14^ /". 62/4 o(,,, . Y o x0A-Ict/vo

Page 5: Policy Analysis - Thomas Dye

Policy Analysis

Coleman's study was not without its critics,' it was nonetheless the first com-prehensive analysis of the American public school system and included dataon 600,000 children, 60,000 teachers, and 4,000 schools.

The results of Coleman's study undermined much of the conventionalwisdom about the impact of public educational policies on student learningand achievement. Prior to the study, legislators, teachers, school adminis-- - -trators, school board members, and the general public assumed that factorssuch as the number of pupils in the classroom, the amount of money spenton each pupil, libraitand labOritory facilities; teachers' salaries, the qualityof the curricultn,- and Other •tharicteristics of the school affected the quail-,ty of education and educational opportunity. But systematic analysis revealedthat these factors had no significant effect on student learning or achieve-ment. Even the size of the class was found to be unrelated to learning, al-though educatOri had asserted the importance of this factor for decades. Inshort, the things that "everybody knew" about education turned out not tobe soh

The only factors that were found to affect a student's learning to anysignificant degree were (1) family background and (2) the family backgroundof classmates. Family background affected the child's verbal abilities and at-titudes toward education, and these factors correlated very closely withscholastic achievement. Of secondary but considerable significance were theverbal abilities and attitudes toward education of the child's classmates. Peer-group influence had its greatest impact on children from lower-class families.Teaching excellence mattered very little to children from upper- and middle-class backgrounds; they learned well despite mediocre or poor teaching.Children from lower-class families were slightly more affected by teacherquality.

Reanalyzing Coleman's data for the U.S. Commission on Civil Rights,Thomas F. Pettigrew and others found that black students attending pre-dominantly black schools had lower achievement scores and lower levels ofaspiration than black students with comparable family backgrounds who attendedpredominantly white schools.* When black students attending predominantlywhite schools were compared with black students attending predominantlyblack schools, the average difference in levels of -achievement amounted to

'For reviews of the Coleman Report, see Robert A. Dentler, "Equality of EducationalOpportunity: A Special Review," The Urban Review (December 1966); Christopher Jenks, "Educa-tion:The Racial Gap," The New Republic (October 1,1966); James K. Kent, "The Coleman Report:Opening Pandora's Box," Phi Delta Kappan ( January 1968); James S. Coleman, "EducationalDilemmas: Equal Schools or Equal Students," The Public Interest (Summer I966); James S. Coleman,"Toward Open Schools," The Public Interest (Fall 1967); and a special issue devoted to educa-tional opportunity of Harvard Educational Review, vol. 38 (Winter 1968).

'U.S. Commission on Civil Rights, Racial Isolation in the Public Schools, 2 vols. (Washington,D.C.: Government Printing Office, 1967).

Policy Analysis 9

more than Iwo grade levels. On the other hand, achievement levels of whitestudents in classes nearly half-black in composition were not any lower thanthose of white students in all-white schools. Finally, special programs to raiseachievement levels in predominantly black schools were found to have nolasting effect.

Policy Implications

The Coleman Report made no policy recommendations. But, like agreat deal of policy research, policy recommendations were inferred fromits conclusions. First of all, if the Coleman Report was correct, it seemedpointless to simply pour more money into the existing system of publiceducation—raising per pupil expenditures, increasing teachers' salaries,lowering the number of pupils per classroom, providing better libraries andlaboratories, adding educational frills, or adopting any specific curricularinnovations. These policies were found to have no significant impact onlearning.

The findings of the Coleman Report undermined the logic of Title Iof the Elementary and Secondary Education Act of 1965 (see Chapter 7).This piece of congressional legislation authorized large amounts of federalassistance each year for "poverty impacted" schools. The purpose of this pro-gram was to remedy learning problems of disadvantaged children by increas-ing spending for special remedial programs. But the Coleman Report. im-plied that compensatory programs have little educational value. They mayhave symbolic value for ghetto residents, or political value for officeholderswho seek to establish an image of concern for the unde-privileged, but theyare of little educational value for children.

The reaction of professional educators was largely one of silence.Perhaps they hoped the Coleman Report would disappear into historywithout significantly affecting the longstanding assumptions about the importance of money, facilities, classroom size, teacher training, and curricula.Perhaps they hoped that subsequent research would refute Coleman's find-ings. Daniel Moynihan writes:

The whole rationale of American public education came very near to crashingdown, and would have done so had there not been a seemingly general agree-ment to act as if the report had not occurred. But it had, and public educationwill not now be the same. The relations between resource input and educa-tional output, which all school systems, all legislatures, all executives have ac-cepted as given, appear not to be given at all. At very least what has heretoforebeen taken for granted must henceforth be proved.'

'Daniel P. Moynihan, Alaximunt Feasible ilisunderslmuling (New York: Free Press, 1969),p. 195.

Page 6: Policy Analysis - Thomas Dye

Policy Analysis 1 1

10 Policy Analysis

The U.S. Commission on Civil Rights used the Coleman Report to but-tress its policy proposals to end racial imbalance in public schools. Inasmuchas money, facilities, and compensatory programs have little effect on stu-dent learning, and inasmuch as the socioeconomic background of the stu-dent's classmates does affect his or her learning, it seemed reasonable to arguethat the assignment of lower-class black students to predominantly middle-class white schools would be the only way to improve educational oppor-tunities for ghetto children. Moreover, because the findings indicated that 'the achievement levels of middle-class white students were unaffected byblacks in the classroom (as long as blacks were less than a majority), the com-mission concluded that assigning ghetto blacks to predominantly whiteschools would not adversely affect the learning of white pupils. Hence, theCommission called for an end to neighborhood schools and for the busing

of black and white children to racially balanced schools.

Research on Busing

However, in 1972, Harvard sociologist David Armor reviewed theavailable evidence of the effect of busing on achievement levels of blackstudents.' His conclusions: busing black students out of their neighborhoodsto predominantly white schools did not improve their performance relativeto that of white students, even after three or four years of integrated educa-tion. His interpretation of the impact of busing on the achievement levelsof black students indicated that black students were not being helped "inany significant way" by busing, and he urged consideration of the questionof whether psychological harm was being done to black students by placingthem in a situation where the achievement gap was so great. Note that Ar-mor was not contradicting the Coleman Report. Coleman was observing blackchildren who were attending predominantly white schools not as a resultof deliberate government action, but rather within the previously existingpattern of "neighborhood schools." In contrast, Armor was observing blackchildren who had been deliberately reassigned to integrated schools by

government action.The policy implications of Armor's work appeared to support op-

ponents , of governmeia-mandated racial balancing. Other social scientistsdisputed Armor's review of the relevant research findings, including ThomasF. Pettigrew, who originally used the Coleman data in support of busing.'They argued that Armor's work undermined progress toward an integratedsociety and reinforced racism. But Armor replied that social science findings

'David J. Armor, "The Evidence on Busing," The Public Interest, nu. 2$ (Summer 1972),

90-126.'Thomas F. Pettigrew et al., "Busing: A Review of 'The Evidence,'" The Public Interest,

no. 31 (Spring 1973). 88-113.

cannot be used only when they fit the political beliefs of social scientists andignored when their policy implications are painful.8

Busing and "White Flight"

Coleman himself reentered the fray in 1975 with the publication of

a new report, Trends in School Desegregation. , This "Second Coleman Report" •

appeared to counter earlier implications about busing as a means to achieve

equality of educational opportunity. In examining changes in segregationover time in twenty-two large cities and forty-six medium-siz ed cities, Cole,man found that an increase in desegregation was associated with a loss ofwhite pupils—"white flight." This white response to desegregation wasgreatest in large cities with large proportio ns of black school pupils, whichwere surrounded by predominantly white, independent, suburban schooldistricts. Coleman predicted that the long-run effect of white-pupil loss inthese cities would offset government efforts to desegregate public schoolsand contribute to greater rather than less racial isolation. As Coleman

explained:

There are numerous examples of governme nt policy in which the result of theinteraction between policy and response is precisely the opposite of the resultintended by those who initiated the policy. It is especially important in the

case of school desegregatio n to examine this interaction, because many of the

actions taken by individual s , and some of those taken by their local govern-ment bodies, have precisely the opposite effect to that intended by the federalgovernment. The most obvious such individual action, of course, is a move ofresidences to flee school integration."

Coleman had not lost his earlier belief in the achievemen t benefits of

school integration. But he believed that large-scale busing had so many

'David J. A I nn it, "The Double Double Standard," The Public Interest, no. 31 (Spring 1973),

119-31. Still another reaction to the Coleman Report is found in the work of Harvard educator

Christopher Jenks, Inequality: A Reassessment of the Effect of Family and Schooling in America (New

York: Basic Books, 1972). Jenks reanalyzed Coleman's data and conducted additional research

on the impact of schooling on economic success. He found that school quality has little effecton an individual's subsequent success in earning income. He concluded, therefore, that noamours of educational reform would ever bring about economic equality. Jenks assumed thatabsolute equality of income is the goal of society, not merely equality of opportunity to achieveeconomic success. Because the schools cannot ensure that everyone ends up with the same in-come, Jenks concludes that nothing short of a radical redistribution of income (steeply pro-gressive taxes and laws preventing individuals from earning more than others) will bring abouttrue eqalit in Aerica. Attempts to improve die educational system, therefore, are a waste

of timeu

andy effort.

m Thus, the Coleman findings have been used to buttress radical arguments

about the ineffectiveness of liberal reforms.'James S. Coleman et al., 7'rends in School Desegregation 1968-1973 (Washington, D.C.: Ur-

ban Institute, 1975)."Ibid., p. 2.

Page 7: Policy Analysis - Thomas Dye

12 Policy Analysis Policy Analysis 13

_

negative consequences, including white flight from the cities, that busingwas self-defeating as a means of achieving equal educational opportunity.

Critics of the notion of "white Bight" argue that there are many reasonsbesides desegregation which encourage white migration out of central cities."Many before and after studies of black students bused to majority-whitemiddle-class schools report significant gains by the black students in verbaland mathematical skills over several years. And almost all researchers reportthat desegregation has little, if any, effect on the achievement levels of whitestudents.' 2 There is very little racial interaction in desegregated schools ex-cept in supervised activities such as sports, band, and cheerleading.

Educational Reform: What Works?•

Early educational policy analysis, especially Coleman's research in the1960s and 1970s, documented the ineffectiveness of many policy alternatives.Spending more money on the existing educational system was unproduc-tive: as federal aid to education increased during the 1970s, student achieve-ment test scores declined. Racial imbalances in many large city school systemsincreased despite busing, and the achievement scores of inner city blackpupils failed to improve. But social science can take little comfort frompredicting what does not work. Social science has a societal responsibilityto search for policies that do work.

In recent years, spurred on by evidence that American students per-form poorly in comparison with students of other advanced industrial na-tions," social scientists as well as policy makers throughout the United Statessought to identify policy alternatives that could improve the educationalachievement levels of pupils (see Chapter 7). Again, sociologist James S. Cole-man and his associates produced the most influential analysis of school fac-tors that affected achievement levels.

While Coleman's early research indicated that nearly all differencesin student achievement levels are attributable to home and peer group in-fluences and very little to differences in public schools, Coleman remainedpuzzled by the consistently higher achievement scores of pupils in Catholicschools. Indeed, Coleman documented the fact that overall differences inverbal and mathematical achievement levels between tenth-grade studentsin Catholic schools and public schools was over two full grade levels." But

"See Edward J. Hayes, Busing and Desegregation: The Real Truth (Springfield, Mass.: CharlesC. Thomas, 1981).

"See Nicolaus Mills, ed., The Great School Bus Controversy (New York: Teachers CollegePress, 1973); and Gary 0i :field, Must We Bus? (Washington, D.C.: Brookings Instilution, 1978).

"National Commission on Excellence in Education, A Nation at Risk (Washington. D.C.:Government Printing Office, 1983).

"James S. Coleman, Thomas Hoffer, and Sally Kilgore, High School Achievement (New York:Basic Books, 1982).

he quest ion remained whether this difference was a product of differenteducational policies and practices in Catholic and public schools, or dif-ferences in the types of pupils and families who chose to send their childrento Catholic schools. Coleman was able to control statistically for differencesin race, religion, socioeconomic status, family type, and other studentbackground characteristics, by comparing achievement levels of Catholic andpublic school students with the same backgrounds. It turned out that sttidentswho attended Catholic schools scored over one full grade level higher thanpublic school students from comparable fancily backgrounds. In short, while familybackground is mbst important, schools can and do make a difference ineducational achievement. Yet the "Catholic school effect" was not a resultof smaller classrooms, or higher teachers' salaries, or better facilities; on thecontrary, Catholic schools scored poorly on these traditional educationalpolicy measures.

Many public school educators were angered by the implication thatCatholic schools produced better results than public schools. Some fell backon the old argument of "progressive" education that verbal and mathematicachievement scores are not really good measures of educational outcomes.Others argued that Catholic schools were "creaming" off the better students

' and leaving poorer students in the public schools. Still others repeated themyth that Catholic schools expelled problem children that public schoolshad to accept. (Actually Catholic schools receive many problem and delin-quent pupils from families who send them there because "they needdiscipline.")

Catholic school students not only performed better on achievementtests than students from comparable backgrounds in public schools, Catholicschool students also learned more over time. Coleman and his associatesfollowed up on their earlier study of tenth graders by testing these samepupils in the twelfth grade. Catholic school students raised their scoressignificantly more than public school students over these two years. Thegreatest gains in achievement in Catholic schools occurred among studentsfrom disadvantaged backgrounds: blacks, Hispanics, and children of parentswith low education levels. Moreover, the dropout rate for public schoolstudents in Coleman's study for this two-year period was 14.3 percent; inCatholic schools it was only 3.4 percent. And the lower dropout rate forCatholic schools occurred among students of all family backgrounds.

Why do Catholic schools work better than public schools? Colemanhimself attributed the different results of Catholic and public schools to"social capital"—strong relationships within the family and between the fam-ily and the school. The families of Catholic school pupils deliberately choseto send their children to Catholic schools. When the school is perceived bythe student as an extension of the family, or in cases where no strong familyties exist and the school is a substitute for the family, students perform well.

Page 8: Policy Analysis - Thomas Dye

Agreement among family, pupil. iassitiates.,ind teacher on values and normsproduces a iigh expectation for at-I lie ,. et nem and an orderly and disciplinedlearning environment. These factors help explain why students from com-parable family backgrounds pet 1! win ter in Catholic schools than in publicschools.

Reforming Public Schools

Thus, social science research indicates that children's educational per-formance improves when the schtiiil is perceived as an extension or substitutefor the family. The family act of rhoocing a school for the child, rather thanhaving the child assigned to a school by government authorities, helps toestablish the linkage between famil y and .school.

What are the implications of this research for public policy? Theresearch provides a rationale for reforming education to encourage paren-tal choice of schools. Some reformers use the research to support the useof educational vouchers given to parents to spend at any school they choose,public or private. The state would redeem the vouchers submitted by eachschool in amounts equivalent to the costs of educating each child. Schoolswould be encouraged to compete for students since educational funds wouldflow to schools on the basis of enrollment. Competition would encouragemagnet schools to allow freedom for principals and teachers to determinegoals, curriculum, discipline, and structure within their schools, and todevelop programs that appeal to parents, their "customers." The "best"schools would have excess demand and might have to turn pupils away; otherschools would have to improve themselves or close.

Alternatively, a parental choice system might be limited to publicschools. This would be a less radical reform, but it would still encourageparental involvement in education and competition among schools. Severalstates and school districts have already experimented with choice plans:parents choose a public school for their children and state educational fundsflow to the schools on the basis of enrollment. (See Chapter 7.)

However, not many in the public educational establishment—schooldistrict administrators, state education officials, colleges of education,teachers' unions—are prepared to support these refOrms. Parental choicethreatens the traditional power of educators to assign pupils, determine cur-ricula, establish goals, and monitor student progress, with only minimalparental involvement.

Policy Analysis and Political Conflict

The point of this brief discussion is that policy analysis sometimes pro-duces unexpected and even embarrassing findings, that public policies donot always work as intended, and that different political interests will inter-

pret their findings of policy research differently—accepting, rejecting, orusing these findings as they fit their own purposes.

POLICY ANALYSIS AND THE QUESTFOR "SOLUTIONS" TO AMERICA'S PROBLEMS

It is questionable that policy analysis can ever provide "solutions" toAmerica's problems. War, ignorance, crime, poor health, poverty, racialcleavage, inequality, poor housing, pollution, congestion, and unhappy liveshave afflicted people and societies for a long time. Of course, this is no ex-cuse for failing to work toward a society free of these maladies. But our striv-ing for a better society should be tempered with the realization that "solu-tions" to these problems may be very difficult to find. There are many reasonsfor tempering our enthusiasm for policy analysis, some of which are il-lustrated in the battle over educational policy.

Limits on Government Power. First of all, it is easy to exaggerate theimportance, both for good and for ill, of the policies of governments. It isnot clear that government policies, however ingenious, could cure all or evenmost of society's ills. Governments are constrained by many powerful en-vironmental forces—wealth, technology, patterns of family life, class struc-ture, child-rearing practices, religious beliefs, and so on. These forces arenot easily managed by governments, nor could they be controlled even ifit seemed desirable to do so. In the final chapter of this volume we will ex-amine policy impacts, but it is safe to say here that some of society's prob-lems are very intractable. For example, it may be that the only way to ensureequality of opportunity is to remove children from disadvantaged familybackgrounds at a very early age, perhaps before they are six months old.The weight of social science evidence suggests that the potential for achieve-merit may be determined at a very young age. However, a policy of remov-ing children from their family environment at such an early age runscontrary to our deepest feelings about family attachments. The forcibleremoval of children from their mothers is "unthinkable" as a governmentalpolicy. So it may turn out that we never really provide equality of oppor-tunity because cultural forces prevent us from pursuing an effective policy.

s.

Disagreement Over the "Problem." Second, policy analysis cannot of-fer "solutions" to problems when there is no general agreement on whatthe problems arc. Coleman's research assumed that raising achievement levels(measures of verbal and quantitative abilities) and raising aspiration levels(the desire to achieve by society's standards) were the "problems" to whichour efforts should be directed. But others argue that racial segregation in

i

Policy Analysis 1514 Policy Analysis

Page 9: Policy Analysis - Thomas Dye

16 Policy Analysis

the schools is constitutionally impermissible, whether or not integration im-proves the achievement levels of students. In other words, there is no realagreement on what societal values should be implemented in educationalpolicy. Policy analysis is not capable of resolving value conflicts. At best itcan advise on how to achieve certain outcomes, after these outcomes havebeen agreed upon as valuable; it cannot determine what is truly valuablefor society.

Subjectivity in Interpretation. Third, policy analysis deals with very sub-jective topics and must rely upon interpretation of results. Professionalresearchers frequently interpret the results of their analyses differently. Socialscience research cannot be "value-free." Even the selection of the topic forresearch is affected by one's values about what is important in society andworthy of attention. Years ago, sociologist Louis Wirth observed:

Since every assertion of a "fact" and the social world touches the interests ofsome individual or group, one cannot even call attention to the existence ofcertain "facts" without courting the objections of those whose very raison diirein society rests upon a divergent interpretation of the "factual" situation."

Limitations on Design of Human Research. Another set of problems insystematic policy analysis centers around inherent limitations in the designof social science research. It is not really possible to conduct some formsof controlled experiments on human beings. For example, researchers can-not order middle-class white children to go to ghetto schools for several yearsjust to see if it has an adverse impact on their achievement levels. Instead,social researchers must find situations in which educational deprivation hasbeen produced "naturally" in order to make the necessary observations aboutthe causes of such deprivation. Because we cannot control all the factors thatgo into a real-world situation, it is difficult to pinpoint precisely what it isthat causes educational achievement or nonachievement. Moreover, evenwhere some experimentation is permitted, human beings frequently modifytheir behavior simply because they know they are being observed in an ex-perimental situation. For example, in educational research it frequently turnsout that children perform well under any new teaching method or curricularinnovation. It is difficult to know whether the improvements observed area product of the new teaching method or curricular improvement or mere-ly a product of the experimental situation. Finally, it should be noted thatthe people doing policy research are frequently program administrators whoare interested in proving the positive results of their programs. It is impor-tant to separate research from policy implementation, but this is a difficultthing to do.

"Louis Wirth, Preface to Karl Mannheim, Ideology and Utopia: An Introduction to the Sociologyof Knowledge (New York: Harcourt Brace Jovanovich, 1936).

Policy Analysis 17

Complexity of Human Behavior. Perhaps the most serious reservationabout policy analysis is the fact that soci21liarolAmsaresosoupU com plex socialscientists are unable to make accurate predictions about the impact of pro-Posed policies. Social scientists simply do not know enough about individual andgroup behavior to be able to give reliable advice to policy makers. Occasionally policymakers turn to social scientists for "solutions," but social scientists do nothave any "solutions." Most of society's problems are shaped by so manyvariables that a simple explanation of them, or remedy for them, is rarelypossible. A detailed understanding of such a complex system as human societyis beyond our present capabilities. The fact that social scientists give so manycontradictory recommendations is an indication of the absence of reliablescientific knowledge about social problems. Although some scholars arguethat no advice is better than contradictory or inaccurate advice, policy makersstill must make decisions, and it is probably better that they act in the lightof whatever little knowledge social science can provide than that they actin the absence of any knowledge at all. Even if social scientists cannot predictthe impact of future policies, they can at least attempt to measure the im-,p_a_ct of current and past public policies and make this knowledge availableto decision makers.

P01!1C`,Q11AINLJI"kaALYSISAS AR6tAT Ne'')D Ct9r).t4A)AFT 'CLP‘ i°(CtctAvv-k412ijw°L-CWv-1):1"Pottya. r4t-ecaAA ^-%-t"-eko

Understanding public policy is both an art am! a craft. It is anbrjecauseit requires insiativity, and imagination in identifying societal prob-lems and describing them. in devising public policies that might alleviatethem• and then, in finding out whether these policies end up making thingsbetter or worse. It is a era Because t ese as s usua y require someknowledge of economics, political science, public administration, sociology, law, and statistics. Polhesetraditional academic discs phi

We doubt that there is any "model of choice" in policy analysis—thatis, a single model or method that is preferable to all others and that con-sistently renders i he best solutions to public problems.

Instead we are in agreement with political scientist Aaron Wildaysky,who wrote:

Policy analysis is onc activity for which there can be no fixed program, forpolicy analysis is synonymous with creativity, which may be stimulated by theoryand sharpened by practice, which can be learned but not taught."

Wilda ysky goes on to warn students that "solutions" to great public ques-tions are not to be expected.

"Aaron Wilda ysky, Speaking Truth to Power (New York: John Wiley, 1979), p. 3.