Global Foundation of Public Administration

For-Profit Welfare: Contracts, Conflicts, and the Performance Paradox

Janice Johnson Dias University of Michigan

Steven Maynard-Moody University of Kansas

ABSTRACT

This article examines how financial inducements in performance contracts shape the inner

workings of a for-profit welfare-to-work training program serving long-term recipients. Our

work pays particular attention to how contract requirements shape relationships between

manager and line staff and their treatment of clients. We argue that contract design, coupled

with bottom-level management efforts to meet contractual obligations, leads to a perfor-

mance paradox—the same actions taken to achieve contractual results ironically produce

negative program practice and poor client outcomes. Thus, rigidly constructed legal agree-

ments between the government and private service providers can distort incentive struc-

tures, causing programmatic conflicts between management and staff, and do little to

reduce long-term welfare use and diminish recipients’ poverty.

Welfare for profit—the concept jars, but the 1996 Personal Responsibility and Work

Opportunity Reconciliation Act (PRWORA) invites such a depiction. PRWORA promised

to ‘‘change welfare as we know it,’’ and perhaps one of the most dramatic change was

encouraging states to contract with for-profit firms to provide welfare services. Advocates

of privatization claimed that private sector efficiency would offer poor families better

service for less money. Critics worried about firms profiting from the impoverishment

of others. Ten years after this reform, it is time to ask how for-profit services have

responded to the needs of poor families and the exigencies of street-level work. PRWORA

gave states broad latitude to reorganize welfare delivery in order to make work its focal

point (Brodkin 2000). Nonetheless, 1 decade later, we still know remarkably little about

what occurs within these private agencies.

Using a private for-profit welfare-to-work training program serving long-term welfare

recipients as our site of inquiry, this research extends previous studies by giving greater

This research is supported in part by a grant to the first author from the Ford Foundation to the Research and Training

Program on Poverty and Public Policy at the University of Michigan. For their helpful feedback, we thank Sandra K.

Danziger, Christopher Jencks, Kevin Delaney, Yeheskel Hasenfeld, David Elesh, Alford Young, Evelyn Brodkin, Mary

Corcoran, Pamela Smock, Mayer Zald, Vicki Lens, Anthony Mallon, Kristin Seedfeldt, Thomas Brock, Vincent Louis,

Mary Stricker, Robert Reilly, Shreya Janssens, and Sheldon Danziger, participants at the New Directions for Research

in Social Policy and Organizational Practices conference and the annual meeting of the Midwest Political Science

Association. Address correspondence to the author at [email protected].

doi:10.1093/jopart/mul002 Advance Access publication on July 19, 2006

ª The Author 2006. Published by Oxford University Press. All rights reserved. For permissions, please e-mail: [email protected].

JPART 17:189–211

transparency to a set of managerial and frontline practices and routines that have remained

otherwise opaque. Few studies examine for-profit agencies, and most welfare studies

choose to study managers (Lurie 2001), street-level workers (Sandfort 2000; Gais et al.

2001; Maynard-Moody and Musheno 2003; Watkins 2003), or clients (Edin and Lein

1997; Soss 2002; Hays 2003). By combining these issues, this study broadens the research

agenda and highlights the connections and implication of the ‘‘work-first’’ philosophy of

welfare reform legislation and the subcontracting of government services on the service

delivery process.

Much of the existing welfare reform literature evaluates the caseload declines and

factors that contribute to recipients’ outcomes (Council of Economic Advisers 1997;

Brauner and Loprest 1999; Schoeni and Blank 2000; Ziliak et al. 2000). Welfare organ-

izations have not been central to these discussions. Yet long-standing studies on human

service agencies have shown that what occurs during the enactment of policy is central to

understanding both policy successes and failures (Lurie 2001) and the operational proce-

dures and interpersonal relationships within these agencies are equally important to un-

derstanding such outcomes (Lipsky 1980; Handler and Hasenfeld 1997; Lurie 2001;

Brodkin 2002). Issues of managerial-staff relations, line worker discretion, and goal con-

gruence are essential parts of the policy process, and the literature on public agencies and

street-level bureaucracies maintains that client outcomes can be explained in part by

frontline discretion as exercised by street-level bureaucrats, such as caseworkers, police,

and teachers (Lipsky 1980; Goodsell 1981; Hasenfeld and Steinmetz 1981; Scott 1997;

Cohen 1998; Maynard-Moody and Musheno 2003).

The paradox of the new welfare legislation is that although it has afforded widespread

latitude and discretion to state administrators and their intermediaries, it may have simul-

taneously created greater constraint at the bottom. An underlying paradox—some would

argue ‘‘a fundamental contradiction’’—of welfare reform is that although states and local-

ities are encouraged to innovate and adapt programs to their communities, the individuals

who deliver these services to clients, the street-level workers, may be constrained in their

discretion when applying program policies to individual needs. This paradox may be most

evident in for-profit delivery of welfare as the cost-cutting incentives for management

confront the client-focused attention of frontline workers.

Contracting was promoted as the antidote for the organizational problems of public

bureaucracies. Contracting, as a form of privatization, has been used since the 1960s by the

government as a way to create greater efficiency at a fixed lower cost (Brodkin and Young

1989; Kamerman and Kahn 1989). However, prior to the 1996 welfare reform, the tradi-

tional approach to providing welfare services was in the form of noncompetitive, quasi-

grant arrangements made directly and primarily to nonprofit organizations (Nightingale

and Pindus 1997). PRWORA, by allowing states more flexibility in service provision,

made it possible for states to remove many of the existing contracting restrictions. Some

states opted to contract out services, primarily job training, and to use performance con-

tracts where payment is based solely or primarily on placement outcomes or quotas. The

institutional consequences of these changes are largely unknown.

Current research on performance contracts show that contracts typically stipulate

a focus on short-term deliverables, such as immediate job placement outcomes or quotas,

at the expense of longer term goals like employment quality and job retention (Autor and

Houseman 2005). Contract language appears to encourage some welfare programs to in-

vest their resources only in those activities that contribute directly to the financial ‘‘bottom

Journal of Public Administration Research and Theory190

line’’ (Brodkin et al. 2002; Thiel and Leeuw 2002, 267). Such efforts are likely to clash

with the case management perspective.

Drawn primarily from social work, the case management approach is based on in-

dividual assessment; it demands detailed, emotional interaction with clients, taking time to

discover their areas of interest and identifying the obstacles preventing them from accom-

plishing their individual goals (Anderson 2001; Segal, Gerdes, and Steiner 2004). In

theory, the work-first and case management philosophies complement welfare goals

(Anderson 2001, 167). But in program practice, reconciling the process-oriented philoso-

phy of case management with the work-first demands of ‘‘take a job, any job, now’’ may be

difficult. The work-first approach assumes that clients have few obstacles to employment

beyond personal motivation, and any work—even low-wage work with no benefits, little

security, and limited chances for advancement—is viewed as the best antidote to wel-

fare dependence. Case management, on the other hand, recognizes that individuals face

diverse barriers to employment; it sees good jobs, not just job placement, as the outcome,

not the treatment.

The difference in philosophy may be exaggerated when coupled with contracting out

to for-profit agencies where management is called upon to maximize output while also

minimizing costs. 1 These goals are likely to conflict with the casework goals that many

service staff bring to the job. Under the new accountability model, frontline staff, 2 who are

trained as case managers and deliver services in these new, for-profit welfare-to-work

agencies, may find themselves trapped between their understanding of clients’ needs

and the demands of their supervisors to place clients quickly and cheaply in jobs.

At the organizational level, the existence of these different perspectives is expressed

in normative conflicts over how best to move recipients from welfare dependency to self-

sufficiency. Using the narratives of managers, line workers, and clients, we examine the

ways these managers, street-level workers, and clients interpret the welfare directives to

better understand how the new policy imperative of work first merges with case manage-

ment approaches to help recipients. We present new findings on the competing demands

between policy demands and practice.

Using data from an intensive study of one for-profit job-training program, a local

subsidiary of a national program, we provide a detailed account of how the focused

attention to work-first service goals and contract compliance produced a performance

paradox. We examine how the performance contract contributes to tensions within this

program and produces unintended consequences for program practices and its long-term

1 The same may be true for nonprofit agencies that enter contractual relationship with the state. In the age of

performance contracting where public administrators of nonprofit agencies are being held to the same performance

outcome standards as their for-profit counterparts, nonprofit agencies may not be so inclined to take contract-centered

approach in their delivery of services (for fuller discussion of how contracting shape nonprofit service delivery, see

Sommerfeld and Reisch 2003).

2 No longer functioning solely as educators, staff in job-training agencies—those agencies responsible for

addressing recipients’ educational and employment barriers and assisting them in finding employment—are now

required to function as caseworkers. New responsibilities include, but are not limited to, goal setting, advocating for

clients, providing supportive counseling, brokering relationships, monitoring employment-related services, and

providing job links (Anderson 2001). In addition, today’s frontline workers in welfare-to-work programs, now

called ‘‘case managers,’’ are required to have more ongoing, direct, and ‘‘helping relationships’’ with clients. The

importance of this change in role from bureaucrat to caseworker cannot be overemphasized as it shifts power and

discretion to staff. As one client put it, ‘‘Generally the caseworker who has your case has your life in their hands . . .’’

(Anderson 2001, 167).

Dias and Maynard-Moody Contracts, Conflicts, and the Performance Paradox 191

welfare participants. We investigate how this job-training program responds to the com-

peting demands of state regulators, executive management, frontline staff, and poor

women recipients. We ask, how do conflicting interests shape the ways frontline work is

conducted and the way frontline discretion operates?

Though focused on a single site, we maintain that the findings of this study may be

generalizable to other private job-training welfare programs similarly structured. Likewise,

our results have bearing on the nature of the gulf between policy ideals and policy imple-

mentation. After a discussion of the methodology, we present the background of the job-

training program and discuss the implementation of the performance contract mandates.

Having laid out the key areas of programmatic conflict, we then offer three sets of narra-

tives from the managers, line staff, and clients. We conclude by offering some explanations

for why these programmatic conflicts occurred and some policy recommendations.

METHODS

The data for this article come from an 18-month qualitative study conducted between

October 2002 and April 2004 3 of four welfare-to-work training programs (Johnson 2004).

This article focuses on one site, WorkOpts, a for-profit welfare-to-work training program

for long-term welfare recipients. 4 Given the focus on issues of process and the relations

between people and organizational practice, we employ a grounded theory approach.

The grounded theory approach demands a dialectic relationship between data gath-

ering and theoretical analysis. Developed by Glaser and Strauss (1999), grounded theory

relies on the interactive process wherein the researcher is constantly comparing theory to

what he/she has learned from the data-gathering process. Simultaneously, the information

she gathers transforms the kinds of questions she asks.

Data Collection Strategies

The research involved multimethods: organizational, interview, and observational. Orga-

nizational data include contract stipulations. Interview data come from 22 clients, 17

street-level workers, and 5 program supervisors and managers. Data were also gathered

from over 1,200 h of participant observations of orientation sessions, classrooms, lunch-

rooms, and lobbies. The three different levels of information allow for triangulation and

depth of information on the relationship between the actors and their environment.

The 60 semistructured individual and group interviews provide information on the

individual perspectives of these actors. Some participants were interviewed twice. Gener-

ally, interviews lasted between 1 and 2 h. The majority of interviews were tape-recorded

and followed a formal protocol, but some discussions were impromptu. Most interviews

took place in a private room at WorkOpts, although several occurred at coffee shops, and

one interview was conducted in the respondent’s home. Group talks were equally impor-

tant to the study because these conversations allowed respondents to share individual

experiences in a large setting while allowing for explication of ‘‘the group’s point of view’’

and the contrasting perspectives of the interviewees (Agar and MacDonald 1995, 81).

3 The first author returned during the summer of 2004 to gather additional data.

4 The names of the individual, program, and the city are pseudonyms. Individual and agencies were promised

anonymity.

Journal of Public Administration Research and Theory192

Beyond interviews, the study also included observations that allowed us to make

sense of the interview data. Lincoln and Guba (1985) argue that observations permit

researchers to acquire the fullest understanding of the phenomenon they are studying

because the researcher must interact with participants in their natural context. To guard

against having a biased view of the program, observation times varied about every 6 weeks.

Typically, observations occurred on Mondays, Wednesdays, and Fridays, but over the

course of the study, observations also took place on Tuesdays and Thursdays. Mondays

and Fridays were the busiest days at WorkOpts; however, the program appeared to work in

similar ways throughout the week.

We also collected contract information as a way to understand how the larger orga-

nizational structure shapes office relations and influences service delivery. Denzin (1989)

posits that this kind of data gathering can be helpful in elaborating the observational

framework, and Drasgow and Schmitt (2002) suggest that this process of exploring agency

documents, particularly the formalized rules, can be useful in developing a theory about

how process mechanisms impact outcomes.

Data Analysis

Each transcribed interview was analyzed using ATLAS.ti, which offers the ability to

manage, extract, compare, and explore meaningful excerpts from extensive amounts of

data. Data collection and analysis were iterative. Interviews were transcribed and analyzed

as soon as they were completed, and these findings would, to a degree, shape subsequent

interviews. In this way, new information was collected as the research proceeded and

questions arose. We began with a search for broad themes and patterns in the conduct

of frontline work, specifically the different set of incentives for managers, line workers,

and clients.

FINDINGS

The Political Economy of the Job-Training Agency

WorkOpts, a for-profit agency, is a subsidiary of Headly Inc., a national marketing research

and consulting firm with more than 4 decades of experience in serving many of the largest

corporations in the United States. They had no previous experience in providing welfare

services and only recently, in 2003, began to provide job training to welfare recipients in

Porter City, a large northeastern city. Its first job-training program, JobOpts, collapsed after

1 year in operation for failing to meet contractual job placement quotas. Despite JobOpts’

failure, Headly Inc. successfully negotiated with the city for additional programs. The city

agreed to fund two new Headly programs, WorkOpts and Exodus. 5

At the start of the research study, WorkOpts had been in operation 6 months. The

program began as a demonstration project. During its initial phase, WorkOpts had 30

employees, including six case managers and two job developers, who provided service

to about 60 clients, many of whom had been on welfare for less than 5 years. At that time,

clients came from six of the 14 welfare districts in Porter City. Each of the four case-

workers on staff worked with 10–12 clients—a decidedly low caseload. According to

5 Exodus collapsed after 5 months of operation.

Dias and Maynard-Moody Contracts, Conflicts, and the Performance Paradox 193

caseworker testimony during the initial phase of the program, ‘‘We had lots of time with

clients to figure out what they needed. We were able to work with them.’’

Contract

In the second phase of the program, Headly executives negotiated a new contract that

required the WorkOpts program to accept all ‘‘extended Temporary Assistance for Needy

Families (TANF)’’ clients from all 14 Porter City welfare districts. ‘‘Extended TANF,’’ as

they were called, are recipients who have been on welfare for 5 or more years and are

among the most difficult to place in employment.

A summary of program features is provided in table 1. For $3 million, WorkOpts

contracted with the city to serve at least 1,200 extended TANF clients per year, approx-

imately 100 clients monthly. 6 With this decidedly hard-to-employ population of recipients,

WorkOpts’ contract required them to place only 10 percent of its population in work. No

minimum wage specifications were set for the WorkOpts’ clients. These minimal contract

expectations are, on one hand, realistic, given the nature of the clients, but, on the other, set

Table 1 Contract Terms

Program features WorkOpts

Type of contract Performance based; agency is required to

place 10 percent of its clients into jobs

Employment timeline Employ at least 90 days

Contractually defined eligible

welfare clients

On welfare 5 years or more considered

extended TANF

Actual years of clients’

welfare experience

11 years

Origin of clients Entire city

Mandated maximum case

management caseload 35:1

Length of time in program 6 months but cyclical

Total required hours of

work activities

30 (includes paid or unpaid work activity at

worksites and classroom instruction, commonly

referred to as job readiness or wraparound classes)

Work training 20 h of paid or unpaid work

Payment schedule $1,053.34 for enrollment in program: $1,316.67

for program participation for 6 months, $2,633.34

for initial job placement a

Minimum average wage required

for placement

None

Number of clients required to

serve annually 1,200

Contract length 1 year renewable

Contract cost $3,000,000

Cost per client $2,500

Cost per job placement $25,000 a At both programs, a client is considered enrolled if they have attended 5 consecutive days of programming without any absences.

6 Porter City has 5,804 extended TANF recipients (Source: Porter City, Division of Income Maintenance).

Journal of Public Administration Research and Theory194

such low expectations that the meaning of ‘‘success’’ and ‘‘failure’’ becomes blurred. The

contract defines success as 90 percent failure.

In exchange for service, WorkOpts was paid according to its ability to enroll, retain,

and place clients in employment: $1,053.34 per client for enrollment in program,

$1,316.67 per client for program participation for 6 months, and $2,633.34 for initial

job placement (see table 1). According to the program manager, in addition to the con-

tractual obligations to the city, Headly Inc. demanded that WorkOpts show at least an

8 percent profit.

Services

WorkOpts’ contract with the city also stipulated that all enrolled clients receive case

management services defined as individual assessment, job coaching, educational services,

and assistance with employment obstacles. According to state guidelines, caseload size

was not to exceed 13 clients per caseworker. Despite the emphasis on case management,

the contract did not attach a fee to delivery of service as they did to job placement out-

comes, a disjuncture that proved telling.

In addition to case management services, WorkOpts was required to provide recip-

ients with 30 h of work activities. These hours were divided into 20 h of on-the-job training

(OJT), paid or unpaid, and 10 h of ‘‘job readiness’’ activities intended to inculcate recip-

ients with a strong work ethic and proper workplace behaviors. 7 WorkOpts subcontracted

OJT services to ‘‘The Mission,’’ a nonprofit agency, whose main responsibility was to

identify and place welfare clients at various not-for-profit human service agencies across

the city, referred to as ‘‘worksites.’’ In exchange for working at these worksites, recipients

received their regular benefits and an additional check for their work hours. 8

Clients

Over the course of the research study, WorkOpts enrolled 1,454 extended TANF clients.

According to their records, 152 or 10 percent of WorkOpts’ clients gained employment.

Headly Inc., WorkOpts’ parent company, hired 20 of these individuals. Without these in-

house hires by Headly Inc., WorkOpts would not have met the minimal 10 percent place-

ment rate stipulated by the contract. Full program data were not available for all Work-

Opts’ clients because none were collected. The program did, however, collect detailed

demographic information, literacy, and mathematical test scores. What information was

available showed that clients had significant barriers to employment. WorkOpts’ clients

read at or below the eighth grade level. The majority of the clients were African American

women and single heads of household.

For many of these long-term clients, welfare had become a way of life. Clients made

typical statements such as: ‘‘I feel like I have been on welfare all my life.’’ ‘‘My mother was

on welfare.’’ Recipients had come to rely on welfare to support themselves and their

families, and though many had worked in the past, they had limited work skills and few

7 In Porter City, job readiness activities had a variety of meanings. At WorkOpts, these services intended to prepare

clients for work were called wraparound activities. This included classroom lectures about the job search process;

however, with the permission of their case managers, some clients were allowed this time to go on independent job

searches where they filled out job applications.

8 Working clients were paid the minimum wage. Some clients who worked at worksites did not earn any wages

because they used up their 6-month federal allotment. Unpaid clients worked side by side in the same agencies as paid

clients.

Dias and Maynard-Moody Contracts, Conflicts, and the Performance Paradox 195

stable employment spells. To the critics of the old Aid to Families with Dependent Chil-

dren policy, WorkOpts’ long-term recipients symbolize what is wrong with welfare—

intergenerational recipients with low levels of education and many children.

To understand the way that these barriers influenced recipients’ employment, the first

author followed 13 clients for 1 year. 9 These clients’ characteristics are shown in table 2

and are compared with national data on TANF recipients. The ethnographic sample

employment outcomes are shown in table 3. The average person in this group had been

on welfare for over a decade and had five children and more than two-thirds lacked a high

school diploma. The majority of these clients reported having childcare problems. More

than half had physical health problems, such as high blood pressure, that required medi-

cation. Many also reported having back problems, spinal injuries, ‘‘nervous breakdowns,’’

and feeling depressed more than twice a week. Three clients revealed that they had

considered suicide. On average, these WorkOpts clients had at least three significant

employment barriers. 10

At the end of their 6 months training at WorkOpts, none of the 13 clients in the

ethnographic sample found employment. After two additional months of observations,

two of the 13 clients found employment; 11 remained unemployed. Of the two employed

clients, one client reported that she found her own part-time employment as a housekeeper

on weekends for $10 an hour. The other recipient found full-time employment through the

program as a security guard but quickly lost her job for lying about her educational

credentials. 11

Although only two of the 13 respondents in the ethnographic sample found

employment, this employment rate is higher than the overall employment rate for Work-

Opts’ clients.

Staffing

To manage the new client population, WorkOpts employed additional staff, totaling 45. 12

Staff included one program director, five department managers, and one data specialist.

WorkOpts hired 38 frontline staff: case managers, classroom instructors, receptionists, and

job developers. The increase in the size of the client population increased caseload size.

Although the contract required the caseload size not to exceed 35, with more clients,

caseload size varied from 70 to 100 per case manager because many clients did not find

employment and were therefore kept on the books. Case managers maintained ongoing

involvement with clients whether they were ‘‘actively’’ participating in the program or not.

Workers would intermittently call clients just to check in, or in some cases, clients would

call caseworkers in cases of life ‘‘emergencies’’ (e.g., childcare fell through, paycheck did

not arrive, pending sanctions). 13

9 This small, nonrandom sample of participants is not statistically representative of the client population. Program

data of educational tests show that the women interviewed are similar to the client population with regard to educational

attainment and marital diversity. These women, like the larger WorkOpts sample, read at low levels, and many were

unmarried mothers.

10 See Danziger, Kalil, and Anderson (2000) for a fuller discussion of recipients’ employment barriers.

11 The client falsely told her employers she had a high school diploma. When she was unable to produce a copy of

the diploma, she was fired.

12 The number of staff changed several times over the course of the research project.

13 Clients also cycled in and out of the program. As the only program of its kind, all extended TANF clients were

required to attend WorkOpts. Unsuccessful clients returned to the program after they completed their 6 months or if

they lost their employment.

Journal of Public Administration Research and Theory196

Table 2 Ethnographic Sample

Barriers to employment National percentage WorkOpts

Race a

Black 39 77% (10)

White 34 8% (1)

Hispanic/Latina 22 15% (2)

Marital status b

Single/never married 66.0 100% (13)

Ever married 22.4 0

Prior work experience (1 5 yes) 92% (12)

Average number of years on welfare 2.0 c

11 years

Age d

25 years or younger 77% (10)

26 years or older 23% (3)

Average Age e

31 years 33 years

Average number of children f

2 children 5 children

Average number of children , 13 3 children

Education g

Less than high school 46.3 69% (9)

High school graduate/GED 47.5 31% (4)

Post high school — 0 (0)

Transportation (no car) h

30.0 i

77% (10)

Health problems

Mother has work disability j

46% (6)

Child with health problem 15.7 k

15% (2)

Mental health problems l

4.3 38% (5)

Criminal background (1 5 yes) 23% (3)

Domestic violence (1 5 yes) 3.2–3.4 m

15% (2)

Childcare problems n

46% (6)

Employed o (1 5 yes) 25.3

p 15% (2)

Number of employment barriers

Average number of employment barriers 3 barriers

0 0% (0)

1 15% (2)

2 23% (3)

3 15% (2)

4 or more 23% (3) a U.S. Department of Health and Human Services. 2004. Fiscal Report on the characteristics of TANF adults. b Ibid. c Harris (1996, 407–26). d Department of Health and Human Services, ‘‘Fiscal report.’’ e Ibid. f Ibid. g Ibid. h If a respondent reporting not having a license or access to a car, they were coded as having a transportation problem. i U.S. Census Data. 2000. Public Use Micro Data Sample. j If respondent reported having a physical health problem such as back and spinal injuries that limited the type of work they can do, they

were coded as having a work disability. k The 1994 National Longitudinal Survey of Youth: percentage of all mothers aged 29–37 with children who have one of six limitations. l If respondents reported being depressed or having considered suicide, they were coded as having a mental health problem. m The 1993 Commonwealth Fund Survey and 1985 National Family Violence Survey: percentage of all women aged 18 and over who

report current severe physical abuse. n If a respondent reported not having reliable childcare or not having someone to help them take care of their children when they were

looking for work or at work, they were coded as having a childcare problem. o Employed after 6 months in program. p Department of Health and Human Services, ‘‘Fiscal report.’’

Dias and Maynard-Moody Contracts, Conflicts, and the Performance Paradox 197

Many line staff were familiar with high caseloads from their prior work experience in

human service jobs. With two exceptions, all other staff had previously worked in Porter

City for a state agency, the majority in mental health. Eight of the 10 case managers had

been case managers before, and the majority had prior experience working with welfare

recipients. Among the staff, only two held Masters degrees, one in Social Work and the

other in Sociology; several others held Bachelors degree in Social Work. The program

director completed only a high school diploma. Thus, the management and staff had

considerable experience working in human services. Many were conversant with the

welfare rules, but few had experience working for and understanding the demands of

a for-profit corporation. Many line staff thought working for a for-profit company would

allow them to help clients—most believed in the message of ‘‘ending welfare as we know

it’’—and increase their own pay. For them, at least initially, working for WorkOpts was an

opportunity. However, many found a stark difference in the work cultures of public human

services and those of the for-profit company. Their optimism quickly soured.

Frontline staff at WorkOpts resembled the racial distribution of the client population:

32 of the 38 staff were black, 5 were white, and 1 was Asian. Three of the five managers

were white. Men occupied the majority of the management positions. None of the man-

agers and job developers were former welfare recipients. Four members of the frontline

staff were former welfare recipients, and eight other line staff members had immediate

family members who were currently receiving cash assistance from the state. The eco-

nomic and social border separating the worker from the client was thin.

Implementing Competing Philosophies

Meeting the contract’s job placement goals and profit quotas created considerable tensions

within program management. The evidence we have marshaled suggests that the easiest

way for WorkOpts’ management to meet the city’s contract goals and make a profit was to

minimize the time and effort line staff devoted to each client. Individualized attention

distracted from the goal of immediate job placement, which compromised the parent

company’s economic gains. Yet many of their hard-to-place clients came to WorkOpts

with multiple employment and life problems. They lacked the work habits and skills to

easily succeed in the job. Many had several young children or a history of continuing health

and substance-abuse problems. Nonetheless, management expected the line staff to treat

this diverse group with a standardized, rapid-paced program. The contract obligated short-

term service and job placement for only 10 percent of their clients, and this quota drove

management’s expectations and demands. Under these conditions, WorkOpts management

had little incentive to go beyond minimal services.

Table 3 Ethnographic Sample Employment Outcomes

WorkOpts

Employed 15% (2)

Average hourly wage $8.50

Job-offered benefits (1 5 yes) 50.0%

Unemployed 85% (11)

N 5 13 13

Journal of Public Administration Research and Theory198

From the onset of the program, the contract inducements competed with the frontline

staff and client motivations and desires. This conflict led to enduring and debilitating

organizational problems. Prior to operating under the new contract, frontline staff had

smaller caseloads and thus more individual time with clients. According to caseworkers’

testimonies, under the case management model, WorkOpts staff placed roughly 40–50

percent of their previous clients into unsubsidized employment. 14

They were then able to

concentrate their energies toward life coaching, educational development, and job fit rather

than quick job placements. Under the new contract, caseworkers had considerably higher

caseloads and less individual time with clients. Where previously line staff were able to

spend 30 min to an hour with individual clients, with higher caseloads and more emphasis

on quick job placement, they now had less than 10 min with each client. Whereas staff

were able to place almost half the clients into employment, under the new contract, staff

struggled to place even 10 percent. In total, the change in contractual stipulations from case

management to a work-focused program produced considerable organizational conflicts

that affected program delivery and client outcomes. From these conflicts emerged three

distinct narratives of what the agency should be doing.

Management, street-level workers, and clients held strikingly different views of what

was needed to move recipients from welfare to work. The primary clash centered on

management’s ‘‘work-first’’ approach and staff’s orientation toward ‘‘social work’’ or

casework practices. 15

These sets of beliefs proved incommensurable. In total, the differ-

ences of views ignited an ideological war between frontline workers and managers in

which the clients were treated more or less as pawns in a struggle over which approach

was best. In sociological terms, it was a conflict between immediate market attachment and

human capital development. In organizational terms, it became a conflict between man-

agement and staff. In policy terms, it was a failure of both sides to achieve the goals of

welfare reform.

We elected to use personal narratives from managers, line staff, and clients as a way

of inviting readers to see these individuals not just as mere subjects but rather as people

whose stories can serve as powerful research tools. Generally, these viewpoints are

flattened or silenced by more traditional research methodologies (Maynard-Moody and

Musheno 2003; Ospina and Dodge 2005).

Management’s Narrative

For the majority of WorkOpts mangers, implementing welfare is about getting line workers

and welfare recipients to comply with the ‘‘rules.’’ Their narratives reveal a keen focus on

compliance and efficiency. Emotionally detached and financially motivated, management

focused their energies on achieving program results as stipulated by the contracts. For four

of the five managers, WorkOpts’ success was tied to their own financial achievement. They

simply could not afford to have the program fail. When asked why they worked at Work-

Opts, the executive manager replied, ‘‘It pays the mortgage.’’ For others, ‘‘I am trying the

for-profit area to pay my bills.’’ Still others responded with ‘‘because of the money and to

compensate me.’’

14 According to worker testimonies, the program served between 70 and 100 clients in total during the demonstration

phase of the program.

15 The terms social work and casework are used interchangeably throughout the paper; however, within the disciple

of social work some distinction exists (see Segal, Gerdes, and Steiner 2004).

Dias and Maynard-Moody Contracts, Conflicts, and the Performance Paradox 199

Managers’ personal success was tied to the program’s meeting its contractual require-

ments. In that vein, they pushed line staff to dedicate their time to the financially rewarded

contract goals of immediate job placements. They also wanted clients to abide by the rules

of the program and take a job, any job. To achieve these goals, managers worked tirelessly.

Besides the personal motivation, the program supervisor explained the impetus behind his

push toward the work-first approach:

Well [the contract] gives you directions. The contract does have direct guidelines. I mean it’s

clearly defining; [It says] this is how you are going to submit the paperwork to the fund

source. This is how blah, blah, blah, blah, so it clearly delineates a process. . . . We have

guidelines that we have to follow, we have those rules that we got to follow like they [staff

and clients] have to follow. . . .

The rules laid out by the contract governed program service options. The contract

required two simultaneous program activities, classroom and work activities, and dictated

what activities clients must receive and specified areas of flexibility and discretion. Despite

being granted some flexibility to be innovative, management chose a narrow reading of the

contract. In practical terms, this meant fully adopting the work-first model of practice at the

expense of the social work or case management model that line staff supported. Describing

the contrast between these perspectives, one manger explained

This type of [welfare-to-work] industry requires fast-paced thinking, quick thinking,

effective thinking, resourcefulness . . . [It’s not] traditional social work, [where] there is

a process in working with a client. There are expectations and rules and regulations and that is

also in the welfare industry, but there isn’t that immediacy as much as . . . within the welfare

industry. In the traditional social work . . . something that needs to get done [can get done]

within a few weeks or a month or something . . . there isn’t as much immediacy.

The manager’s use of the word ‘‘industry’’ accurately reflects the business mode that

frames supervisor understanding and decision making.

He later elaborated

This is not an industry for someone looking to break into social work, etc. I look at it as above

a social work level because in the social work level, the traditional social work, there is

a process in working with a client. But there isn’t that much immediacy. Th[is] industry is

unique in terms of the social work approach. You really can’t have that social work model so

to speak in working with clientele.

Displeased with line staff’s social work or case management approach, managers

described frontline workers as ‘‘taking too much time’’ with clients. According to manag-

ers, staff needed to abandon their slow social work approach in favor of the ‘‘rush-rush’’

style. 16

Such actions would help case management meet the immediate and short-term

results of the city government and the profit expectations of Headly Inc.

Though financial gains and the contract guidelines heavily influenced the manage-

ment’s dedicated focus to quick labor force attachment, the management’s actions were

also motivated by their personal beliefs about the recipients’ job prospects. Managers were

generally pessimistic about clients’ chances of employment and upward mobility. The

16 Client comment.

Journal of Public Administration Research and Theory200

‘‘take a job, any job’’ approach seemed the best course of action; it meant that they and

their organization succeeded even if the clients did not.

Sure [clients] will be able to hook up a job but they’re not going to be able to hook up that $20

an hour job. And so it becomes a catch 22 [for them]. I want the job that is $20 an hour but on

the other hand I don’t have the skills that are going to get me there. So it’s a difficult situation.

When probed on what kind of services WorkOpts offers that may afford recipients

some upward mobility, this manager ignored the issue of skill deficits and personal cir-

cumstance and put forth an alternate explanation for the recipients’ unemployment, mo-

tivation. Retreating to a common stereotype, he argued that

It’s the motivation. Those individuals that have secured appointments they’re not necessarily

college graduates. They do have again eighth, ninth grade reading level and have three to four

children so they do face the same barriers that those that are not. I think it’s the mindset of do

I want to go now or is this the summertime and I just want to chill in the summer and wait

until September comes around. I think that’s part of it.

All managers agreed that one of the biggest obstacles to clients’ employment was

their motivation. When asked what these clients needed to help them, management shared

the standard response, ‘‘I mean these folks [clients] need some, I think, intense clinical

work, quite frankly, just years of depression.’’ However, when queried on what resources

the program provides for these clients, the program manager responded

I think that we have bounced around [getting a psychologist] a little bit, but never given it

much thought. I haven’t seen in the contract or guidelines any indication of needing

a psychologist. I think that’s the reason why there’s no funds allocated for it. . ..

Given the clients’ educational and skill deficits and their difficult personal situations,

motivation is a relatively simple and superficial problem: one that can improve particularly

through therapy. Although managers argued that a psychologist might have been helpful in

motivating recipients, they failed to hire a therapist or subcontract with any agency that

provides this type of service. Cutting costs took precedence over providing a needed client

service.

Street-level Workers’ Narrative

In contrast, frontline workers expressed a typical ‘‘social work narrative,’’ which could be

described as ‘‘people first.’’ Their narratives reveal a focus on helping clients achieve long-

term success.As previousworkhasshown,frontlineworkin social service istied toworkers’

personal and occupational identities (Maynard-Moody and Musheno 2003). Line staff at

WorkOpts were no different. In their words, frontline staff wanted to work in human service

to ‘‘help people.’’ They wanted to use their professional power to make a difference in their

clients’ lives. Line staff tried to tailor their services to what they perceived as clients’ needs

but were limited by management actions, program resources, and contract requirements.

The contract required case management but did not reward its implementation. The

provision of case management services was not tied to any financial inducements or

penalties. With little program support from management and no legal and financial support

for their people-first philosophy, frontline staff were constrained in their ability to help

clients enhance their job skills and prospects. In the end, line staff at WorkOpts were

reduced to client processors, not social workers.

Dias and Maynard-Moody Contracts, Conflicts, and the Performance Paradox 201

Many staff understood that their social work philosophy was not compatible with the

managers’ work-first approach. Summarizing the differences in viewpoints, one line work-

er’s words reflected the perspectives of her peers:

We are in the field of social work [but] a social worker is someone who gets a little

more into the person’s personal life. They go to the house maybe more. They’re more

interested in maybe some of the quality aspects of the life, ‘‘Do you have leisure time

with your children?’’ You know nice things like that. . .. We have a limited amount of

time with clients.

Echoing a similar sentiment, one caseworker said,

We have a limited amount of time, you know, with the person . . . it’s gotta be fast. To me,

they’re [management] looking for results here. ‘‘How many placements did you get? What

were the results?’’ It’s like, almost like a statistical thing. They’re worried about statistics,

where from our end we see it as, ‘‘Okay, it’s not just statistics.’’

Without enough time to work with clients and with management’s emphasis on

placement quotas, frontline staff were not able to explore ‘‘the quality aspects’’ of the

client’s life or implement their social work skills. Workers complained that

You’ve got like maybe ten minutes for each participant, sometimes less than that; so basically

the relationship is really bad. I mean you don’t have time. I wouldn’t use the word

‘‘relationship’’ because you don’t have any relationship; you just fill out the papers [and] ask

basic questions. . . .

Though some staff were willing to abide by these time constraints, others were not.

Some would extend their time with clients ‘‘to work through the barriers that [clients] have

and [help them] realize that those barriers are having a hindrance on them getting a job.’’

One of the major obstacles facing WorkOpts’ clients was their limited human capital;

many lacked the education and skills to move beyond poverty-level employment. Street-

level workers believed that the case management approach could help clients increase their

human capital to make employment gains. Describing the dismal situation for clients, one

worker gave an account of the clients’ employment barriers:

I would say at least eighty percent of the clients that come through the door haven’t

made it through high school—don’t have their diploma. So realistically, we’re

looking at clients who are only gonna get minimum wage. $5.15 an hour ain’t

gonna cut it. Not when you have three kids, or six kids, or however many, it’s not gonna

cut it.

Frontline staff strongly believed that clients needed remedial educational classes such

as Adult Basic Education and courses for General Equivalence Degree (GED). Many line

staff complained about the lack of resources at WorkOpts to help clients. Echoing the

viewpoint of their peers, one staffer summarized the perspective. ‘‘You cannot get some-

body ready for a job if they have no skills, if they have no education, if they have no

training. What job are they gonna have?’’ When asked what the program needed, the

overwhelming frontline response was:

My main issue [with the program] is the job thing. They emphasize the job more than the

education, and to me, it should be education, then job. Because the more education you got,

then the more you can make on your jobs. That would be my main issue.

Journal of Public Administration Research and Theory202

For line staff, helping clients remove such employment barriers was a personal im-

perative. Staff defined ‘‘barriers’’ broadly. According to their definition, obstacles to em-

ployment included educational deficiencies, depression, and social isolation. For example,

line staff wanted to expand the clients’ horizons, introduce them to new places, and help

clients become less socially isolated. In discussion, workers would comment:

Some [clients] have never been downtown except to go to the mall. Some people have

never seen [The Monument] and lived here all their lives. Some people have heard of Willow

Creek [a near by suburb], they heard of Salem [another suburb] but have never been there.

I think that’s really sad because their world is so small. And trying to expand their world

is a difficult task.

Another worker echoed the desire to address clients’ social isolation:

Poor people, it’s easy to oppress them because you keep them in their neighborhoods

and you surround them with liquor ads, drug ads and cigarette ads . . . you don’t educate

them. It’s sad. It’s sad. I wish I could take them on field trips . . . I’m like let me show

you something.

These strong feelings about clients’ social isolation and lack of opportunities fueled

street-level workers’ desire to help. But despite their personal impetus, line staff were

constrained by the management’s guidelines. When queried on how much flexibility they

had to design individualized programs for their clients, caseworkers responded, ‘‘Not

much. I mean the guidelines are very specific. [The management] give us, you know, what

THEY want to do’’ [emphasis original].

Regardless of the pressure from management to reduce their individual time with

clients, some frontline staff refused, openly objecting to management demands. For line

staff, helping clients remove their employment barriers was a personal imperative. Despite

the consequences of the managements’ objection, line staff banded together and organized

client initiatives.

I did the sexual workshops, sexual health. I’ve got [Black Alliance], ‘‘Blacks Educating

Blacks’’ about sexual health to come out and do a two-day workshop and, boy, you never

heard so much howling and screaming and fussing. Management said, well, we don’t want to

do that again.

With limited success, line staff continuously tried to push their agenda forward. As

one frontline staff put it:

One way we wanted to [motivate clients] was not only though our own role models and

behaviors and the way we treated them, but was by bringing in other professionals to talk to

them and let them know that, ‘‘Yes, maybe you made mistakes, maybe your past has been you

know this, that and the other, but that is not the way your future has to be.’’ But that was

circumvented by management because what we came to find out was it wasn’t about

motivating the women to make a transition in their own life, a long-term change, it was about

meeting an immediate number. Just get them jobs. Bottom line—get it done.

Management did not reward such initiatives. Instead, they became angered by what

they perceived to be the staff’s insistence on ignoring the contract’s demands. Seeking to

curtail line staff’s action, management constrained frontline workers’ initiatives by can-

celing all motivational speakers because it distracted from the job placement focus.

Dias and Maynard-Moody Contracts, Conflicts, and the Performance Paradox 203

Unable to give up their fight for broader service options, line staff remained vigilant in

their efforts. When group actions failed, individuals took up the struggle alone. Demon-

strating her protest, one staff member described her actions with clients:

I chose to speak my mind. I choose to let paperwork go. I chose to listen to the client. . ..

When I go into the classroom I close the door and I tell them what I want. How [is

management] going to know? Who is going to tell them? [I tell clients], ‘‘If you’re not going

to keep the job, what’s the point in accepting it? Even better yet what is the point in going on

the job interview for a job you don’t really want?’’

This type of advice, although undermining management directives, allowed street-

level workers to achieve dignity at work. In their eyes, they were doing ‘‘good’’ by helping

clients navigate the welfare system to achieve their individual goals. As other scholars

(Maynard-Moody and Musheno 2003) have found, when staff perceived the rules as unjust,

they often follow their own ethics.

Management responded to this rebellion by confronting staff. Suspected of willfully

disobeying directives, a staff member reported her exchange with management:

I was approached by the [Program Officer, highest ranking administrator] who mentioned

that she thought I was not a good job developer and when I asked her to clarify that she said

the [job placement] numbers were really low and that proved that I was bad job developer . . .

[She said] I’m too much of a purist; that job development is a dog-eat-dog profession and that

I just wasn’t cut out for it. . .. She literally yelled at me when I tried to justify some of the

reasons for the low number [and said] I care too much about the clients and I’m thinking,

‘‘Well isn’t that why you hired me in the first place?’’

The incongruence in viewpoints often led to open and hostile exchanges between

outcome-focused program supervisors and client-centered line staff. On a regular basis,

managers and line staff could be found in the hallways hashing out their complaints with

each other. Quite often, each would storm off to their respective corners without any

resolve.

The inability to reach some form of reconciliation between these two perspectives

damaged the program. Disagreement with management’s work-first philosophy led to the

resignation of two key employees, the most highly educated and experienced, which left

the division understaffed. The result was a higher caseload, exceeding the state require-

ments. The disagreement had a ripple effect: 1 month following the departure of these two

employees, a high-ranking manager, who strongly empathized with clients, also resigned.

She too cited management’s ‘‘tunnel’’ vision and disregard for clients’ long-term interest as

reasons for her resignation. Before she left, she gave a teary-eyed testimonial about her

displeasure with the company.

We are just another welfare-to-work, WorkOpts is a for profit. . .. The company gets

$12,000 for every participant we can get to stay in a job for 90 days . . . and guess what the

jobs are, wiping the shit off of old people’s butts in nursing homes, they are cleaning toilets at

the vet. What am I doing . . . like why am I working here? We are not helping anybody. It’s

horrible I hate it . . . I think [case managers] try really hard to make the program mean

something. We had motivational speaker; we were doing something that was meaningful . . .

and the clients told us when I did my little evaluation every Friday that this was different, that

they didn’t just feel like they were another hot dog being processed in the chute. Now I am . . .

almost embarrassed to work here.

Journal of Public Administration Research and Theory204

These departures left the program with fewer employees who had only worked in

public social service agencies and were themselves former welfare recipients. These

remaining employees had less experience and limited employment networks that they

could leverage on the clients’ behalf, thus further compromising the program’s ability to

connect recipients to work.

The unreconciled differences in the views of managers and workers reflected a clash

embedded in the policy and contract that resulted in debilitating effects for program

operations. By demanding that clients receive case management services but compensating

the program only for job placement, the contract created the basis for the hostile interaction

between management and staff. With no incentives to put the clients’ needs first, the

contract set up a dichotomy between the desires of the staff and those of the management.

Clients’ Narrative

Different from line staff and managers’ viewpoints, clients’ perspectives emphasized the

need for educational credentialing, skill development, and job tailoring in their struggle for

upward mobility. Neither the management’s quick labor force attachment model nor the

case managers’ life skills program addressed clients’ perception of their own needs. In

general, clients believed in the underlying philosophy of work but wanted training, edu-

cation, or the right type of job to make work worth it. Common in most client narratives

was a high level of optimism, but many felt constrained by their limited education and

few employment skills. Believing they were still going to make it, many clients were

hopeful; but for some, welfare generally, and WorkOpts specifically, was crushing, or

had crushed, their dreams. Rather than fostering hope, welfare and WorkOpts had pro-

moted a system of spirit-breaking work that they considered menial. For some clients,

particularly African Americans, this kind of work made them think of the sharecrop

economy their grandparents had fled during the northern migration. Drawing on the slavery

analogy, some clients would comment, ‘‘Welfare is modified slavery. They put you in

jobs that they wouldn’t even want to be in. Where they wouldn’t want to take their kids,

their daughters, their cousins, their aunts, none of their family members just so they can

keep their jobs.’’

The contract required that clients work in these ‘‘slavery-like’’ conditions. The work-

first contract demanded that recipients like ‘‘Macy,’’ a 10-year veteran of welfare who

dropped out of school with only an eighth grade education, find a job within 6 months.

Tearfully defending her employment status, she explained, ‘‘I’ve been looking for a long

time. Being so that I don’t have no high school diploma or no GED is like they don’t want

to hire you because you don’t have none of that.’’ Macy’s plaint is familiar to many

WorkOpts clients; one-third of the ethnographic sample lacked a high school diploma,

and a significant number of clients read at or below the eighth grade level.

Clients experienced the one-size-fits-all approach at WorkOpts as ‘‘You want the

help, but they are not giving you the help that you want. It is like a rush thing with them—in

and out. They are just giving you little jobs. It don’t work like that.’’ Unaware or unfazed

by clients protest to the rushed job placement, the program did not deviate from a narrow

approach to its mandate.

With respect to the 10 h of classroom instruction called ‘‘wraparound’’ classes, the job

readiness activities also did not focus on what the clients desired: educational credentialing

for better job attainment. These classes were built on the basic assumption that long-term

recipients lacked ‘‘life skills,’’ such as stress reduction, time management, and budgeting,

Dias and Maynard-Moody Contracts, Conflicts, and the Performance Paradox 205

needed for employment. The classes were geared toward either educating clients to a proper

work ethic and etiquette or providing them with life skills.

Though some clients may have agreed that they lacked work ethic, more pressing for

these clients was the need for educational development, skill training, and career advance-

ment. They were frustrated by managers and case managers’ attention to their life matters

at the expense of attention to their educational needs, skill development, and need to get

good jobs. Expressing the shared feelings of her peers, one client commented

I don’t just want a job, I want basically a career, that’s why I want to do something in data

entry where I know I have room for advancement and move up in the company. I don’t just

want to work at a store oh, no, I can’t put up with that. [If it’s a choice between just taking

a job then] I’m going to stay on welfare and get all this money that they’re giving me.

Linda had been on and off welfare for 15 years because, in her view, her lack of high

school diploma, not her work ethic, kept her from securing stable employment. She hoped

that by attending WorkOpts she would finally earn her diploma, even a nursing degree. In

March 2003, she reported that

They [WorkOpts] suppose to help me get the GED. I been here since November [2002],

I have not yet started one class for my GED. I only need one more test. And I’ve been to

different places, and it’s hard to come here and to meet my [work] hours, and to go to another

place to do my GED. I can’t do that. I can’t travel two different places at one time. Now they

said they were gonna help you get your GED in here . . . they said that in the beginning of the

month. This is the end of the month, and I have not yet heard nothing. I put my name on the

list . . . ‘‘Oh, we’re gonna talk to you later.’’ And it’s been months, days been running around,

and still nothing yet. I’m like, ‘‘Well when am I gonna start?’’

Linda’s experience was common; she represented the norm at WorkOpts. During

orientation sessions, clients were promised that they would get a chance to earn their

GED, get their driver’s license, and learn computer skills, but none of these services were

provided. Indeed, management reassigned the GED instructor to teach job readiness clas-

ses. No replacement was hired for the GED classes.

In general, WorkOpts’ clients were discouraged and aggravated with the disconnect

between job readiness classroom instruction and their employment needs. Expressing

a sentiment shared by other clients, one client reported that

They [the classes] full of crap. Crap, crap, crap that’s the only thing. The wrap around class

that I’ve been to, the last one I been to was about stress. Relieving stress. Yes, I had some

stress to relieve too and I released it. I released my stress and I let them know that I was

stressed about coming in here. I got tired of coming in here; I got tired of them telling me oh,

we’re going to find you a job today. Oh, we’re going to do this for you today. I’m tired of

hearing it. Don’t tell me unless you going to actually help me.

These classes, created to inculcate recipients with job readiness, had little to no utility

in the clients’ eyes. Job readiness classes did not bring them any closer to getting the nec-

essary educational credentials, skill development, or jobs that they sought. In fact, any client

responded to the lack of GED classes with open hostility. Displays of their anger were

present in the classrooms. In a room filled with 40 or more of their peers, recipients loudly

voiced their complaints about curriculum, but often they transferred this dissatisfaction to

their instructor and fellow classmates. On any given day, clients could be found in verbal and

Journal of Public Administration Research and Theory206

physical altercations in the classroom. Describing the situation, one client says, ‘‘Classes are

out of control . . . people screaming, arguing with each other. . .. Girl[s] . . . yelling at the teacher, who’s trying to instruct the class. It [is] really hectic.’’ Many clients would disrupt

the class saying, ‘‘I’ve been through this. I don’t feel like doing this shit no more.’’ Physical

and verbal altercations ensued between instructors and students, and on more than one

occasion, the police had to be called to calm the crowd. The classroom situation became

so unmanageable that many clients stopped attending classes, choosing instead to hang out

in the lunchrooms or smoke outside. Some even dropped out of the program entirely, thus

lowering the program’s enrollment and jeopardizing their own welfare benefits.

The client narrative demonstrates dissatisfaction with both the work-first and case

management perspectives. Recipients were reluctant to leave welfare for jobs that would

not advance their position. Classes that focused on life skills instead of education and job

attainment irritated them. In short, the client narrative highlights the flaws of the mana-

gerial and caseworker perspectives. Both points of view were equally unsatisfactory in

getting recipients from welfare to work.

DISCUSSION

WorkOpts, although a single organization, is not an idiosyncratic program but rather an

example of what happens during the implementation process of welfare services when

legislators attempt to bring the market model of contracting to social service reform. The

programmatic conflicts at WorkOpts are due in part to the contract incentives and the

clashing views among managers, line staff, and clients of what is needed to move recipients

from welfare to work.

The policy and contract mandated implementing job placement and case manage-

ment, but without any incentives to make the two elements work together, the contract only

served to aggravate traditional organizational dilemmas. Although in theory the program

could feasibly implement both mandates, management had little incentive to do both. They

were financially rewarded for pushing forth the message of work first in the short term at

the expense of providing individually tailored services for clients.

However, the decision to ignore case management for job placement exacerbated

problems within the program. Organizations place demands and constraints on their lead-

ers. What is most important is the leaders’ construction of, and reaction to, the available

and constrained choices; these responses shape organizations’ failures and successes (Lynn

1987; Denhardt 1993). Managers at WorkOpts were unable to balance the priorities of the

contract with the demands of line staff and clients. Opting to ignore line staff requests,

management’s actions fueled conflict between themselves and their line staff and between

line staff and clients. The result was a philosophical war with practical consequences over

whether to prioritize the recipients’ needs or their contract first. In practical terms, this

difference of positions led to a battle over whether to enhance clients’ human capital or

push for quick labor market attachment. Driven by the contract, management succeeded in

designing a program that emphasized rapid and often-temporary labor force attachment

over skill or educational development but at a great expense.

To achieve low-cost rapid job placement, management sought to curtail staff atten-

tion to clients’ needs, but those efforts were never fully successful. Staff continually

revolted, and this led to a chaotic and hostile environment, not efficiency. This clash

between cost-cutting managers and service-oriented street-level worker was predictable,

Dias and Maynard-Moody Contracts, Conflicts, and the Performance Paradox 207

almost inevitable. Frontline staff were hired for their devotion to client-centered service—

they were social workers by either training and/or orientation—yet at WorkOpts they were

being asked to forgo their training and ethics for an economic model that they maintain did

not benefit them or their clients. Previous scholarship shows that conflict is likely where

line staff and management have divergent points of view and differ about which perspec-

tive is more effective.

Frontline workers’ responses to managements’ actions underscore findings from pre-

vious research. Worker discretion can be limited and constrained but not eliminated despite

concerted management effort (Brown 1981; Brehm and Gates 1997). The WorkOpts case

illustrates that even in strict regulatory environments frontline workers will exercise in-

dividual discretion or agency. The WorkOpts case also shows, paradoxically, that attempts

to eliminate frontline worker flexibility and to focus exclusively on contractually defined

objectives can contribute to program failure.

Whereas the conflict between management and frontline workers is telling of what

occurs during implementation, clients’ narrative provides an insightful and often over-

looked part of the welfare story. Their narratives suggest that the debate between work-first

and case management philosophies is largely irrelevant because neither perspective

addresses clients’ views of their needs. From the clients’ perspectives, the work-first

approach overlooks their realities. The majority of clients had worked in the low-wage

market in dead-end jobs; they did not want to return to them. In practice, the work-first

philosophy assumes that they have very few obstacles to employment beyond personal

motivation issues and lack of exposure to the world of work. The client’s narratives un-

dermine these theories. The case management/social work approach, on the other hand,

assumes things about clients that are equally untrue. More than stress reduction, clients

wanted quality jobs. The social work approach ignores their need for skill development.

Like the work-first approach, case management assumes that problems lies within the

client and the solution rests with the social worker. The inadequacy of both perspectives

to be useful to clients suggests that an alternative approach is needed.

Policy Implications

It is also no surprise that these organizational clashes emerged within welfare agencies that

were rushing to implement welfare reform and undergoing major organizational and in-

stitutional changes. This case example highlights the emerging conflicts between manag-

ers, line workers, and clients in one setting. It also demonstrates the poor work outcomes

that can result in these programs that are directed and funded by government contracts.

The evidence from this case shows for-profit agencies, such as WorkOpts, caught in

a difficult position. Beholden not just to clients and staff but also to their stockholders,

these programs are driven by financial objectives that encourage them to maximize output

while minimizing expenditures. Therefore, even as they provide services to recipients, they

may have a disincentive to attend to clients’ entrenched employment deficits, such as low

literacy levels and mental health problems. Indeed, they may be motivated to ignore them

completely, particularly in cases where the cost of attending to their clients’ needs is high.

The evidence also shows that state contracts can contribute to the problem or defuse it.

In this case, the performance-based contract exacerbated the problem. These findings show

the folly on the part of the government of rewarding A while hoping for B (Kerr 1975).

Although government administrators profess to value the importance of case management

Journal of Public Administration Research and Theory208

services, as defined by the one-to-one interaction between worker and clients and the

development of client-centered action plans, financial rewards were linked solely to job

placement outcomes. The contract focused little to no attention on the quality or length of

clients’ employment or on the reduction of their employment barriers.

If not properly stipulated, the external pressures of performance-based contracts

can set up distorted incentive structures for agencies that view welfare contracts merely

as a financial, not service, opportunity—as it did in this case. Facing fiscal and legal

obligations and perpetual resource shortages, program managers operating under this

framework are encouraged to ignore issues of organizational process such as institutional

cohesion, workload distribution, and clients’ needs. Responding to Porter City’s work-first

demands and Headly’s performance pressures, WorkOpts management focused simply on

immediate results. Paradoxically, this myopic attention on job placement produces both

organizational and client failures.

External demands from government contractors can invite job-training programs to

ignore clients’ needs. By tying payment solely to short-term employment outcomes,

government administrators encourage private job-training agencies to pay more attention

to the monetary bottom line and less attention to clients’ needs. Unfortunately, when man-

agers follow this dictum, their actions contribute to internal organizational strife and may

have little effect on clients’ poverty and welfare dependency.

Our results call into question the way privatization of welfare services is currently

executed. National lawmakers contend that the broad ideals of welfare reform policy are to

reduce welfare dependency and move poor families toward self-sufficiency. State and city

administrators and job-training managers are left to translate these lofty goals into practi-

cal reality. However, in the process of implementation, the state and city governments

design welfare contracts that reward welfare-to-work agencies for accomplishing one-

dimensional goals, such as immediate and short-term job placement. Further, these con-

tracts set benchmarks for success so low that job-training agencies are not encouraged to

help clients achieve long-term economic success. Indeed, the actions of the government

encourage job-training agencies to focus their attention to the minute details of satisfying

their contracts. Such actions may actually frustrate recipients already mired in poverty.

Further research can help us disentangle whether these contradictions are inherent and

impossible to overcome in these systems or workable under different types of contractual

arrangements and policy design. Additional investigations may also provide clues about

service approaches that might successfully realize greater work attachment and long-term

client economic stability.

POSTSCRIPT

Headly Inc., WorkOpts’ parent company, hired enough of its own clients to achieve the

contract-stipulated 10 percent success rate and was rewarded by an additional contract to

provide services in a wider area of the city. In the new world of welfare reform, pro-

grammatic failure can mean organizational success.

REFERENCES

Agar, Michael, and James MacDonald. 1995. Focus groups and ethnography. Human Organization

54:78–86.

Dias and Maynard-Moody Contracts, Conflicts, and the Performance Paradox 209

Anderson, Steven. 2001. Welfare recipients view about caseworker performance: Lessons for developing

TANF case management practices. Families in Society 82 (March–April): 165–74.

Autor, David H., and Susan N. Houseman. 2005. Do temporary help jobs improve labor market outcomes

for low-skilled workers? Evidence from random assignments. Working Paper Series 05–20,

National Poverty Center.

Brauner, Sarah, and Pamela Loprest. 1999. Where are they now? What states’ studies of people who left

welfare tell us. Washington, DC: Urban Institute.

Brehm, John, and Scott Gates. 1997. Working, shirking, and sabotage: Bureaucratic response to

a democratic republic. Ann Arbor: Univ. Michigan Press.

Brodkin, Evelyn. 2000. Investigating policy’s ‘practical meaning’ street-level research on welfare policy.

Working Paper 162, Joint Center for Policy Research.

Brodkin, Evelyn, and David Young. 1989. Making sense of privatization: What can we learn from

economic and political analysis? In Privatization and the welfare state, ed. S. Kamerman and

A. Kahn, 121–56. Princeton, NJ: Princeton Univ. Press.

Brodkin, Evelyn, Carolyn Fuqua, and Katarina H. Thoren. 2002. Contracting welfare reform:

Uncertainties of capacity-building within disjointed federalism. Working Paper 284, Joint Center

for Policy Research.

Brown, Michael K. 1981. Working the street: Police discretion and the dilemmas of reform. New York:

Russell Sage Foundation.

Cohen, Marcia B. 1998. Perceptions of power in client/worker relationship. Families in Society, July,

433–42.

Council of Economic Advisers. 1997. Technical report: Explaining the decline in welfare receipt,

1993–1996. A Report by the Council of Economic Advisers, Washington, DC.

Danziger, Sandra K., Areil Kalil, and Nathaniel J. Anderson. 2000. Human capital, physical health,

and mental health of welfare recipients: Co-occurrence and correlates. Journal of Social Issues

56:635–54.

Denhardt, Robert B. 1993. Theories of public organization. Belmont, CA: Wadsworth.

Denzin, Norman K. 1989. The research act: A theoretical introduction to sociological methods. 3rd ed.

Englewood Cliffs, NJ: Prentice Hall.

Drasgow, Fritz, and Neil Schmitt. 2002. Measuring and analyzing behavior in organizations.

San Francisco, CA: Jossey-Bass Publishers.

Edin, Kathryn, and Laura Lein.1997. Making ends meet: How single mothers survive welfare and

low-wage work. New York: Russell Sage Foundation.

Gais, Thomas L., Richard P. Nathan, Irene Lurie, and Thomas Kaplan. 2001. Implementation of the

personal responsibility act of 1996. In The new world of welfare, ed. Rebecca Blank and

Ron Haskins. Washington, DC: Brookings Institution.

Glaser, Barney G., and Anselm Strauss. 1999. The discovery of grounded theory: Strategies for

qualitative research. Chicago, IL: Aldine Publishing.

Goodsell, Charles. 1981. Looking once again at human service bureaucracy. Journal of Politics

43:763–78.

Handler, Joel F., and Yeheskel Hasenfeld. 1997. We the poor people: Work, poverty and welfare.

New Haven, CT: Yale Univ. Press.

Harris, K. M. 1996. Life after welfare: Women, work and repeat dependency. American Sociological

Review 61:407–26.

Hasenfeld, Yeheskel, and Daniel Steinmetz. 1981. Client-official encounters in social service agencies.

In The public encounter: Where states and citizen meet, ed. Charles T. Goodsell, 83–101.

Bloomington: Univ. Indiana Press.

Hays, Sharon. 2003. Flat broke with kids: Women in the age of welfare reform. Oxford, NY: Oxford

Univ. Press.

Johnson, Janice. 2004. Separating policy hopes from policy realities. An examination of the inner-

workings of welfare-to-work training programs. PhD diss., Temple Univ.

Kamerman, Sheila B., and Alfred J. Kahn. 1989. Privatization and the welfare state. Princeton, NJ:

Princeton Univ. Press.

Journal of Public Administration Research and Theory210

Kerr, Steven. 1975. On the folly of rewarding A while hoping for B. Academy of Management Journal

18:769–83.

Lincoln, Yvonna S., and Egon G. Guba. 1985. Naturalistic inquiry. Beverly Hills, CA: Sage

Publications.

Lipsky, Michael. 1980. Street-level bureaucracy: The dilemmas of the individual in public services.

New York: Russell Sage Foundation.

Lurie, Irene. 2001. Changing welfare offices. Welfare Reform and Beyond Policy Brief. Number 9.

Washington, DC: The Brookings Institute.

Lynn, Laurence E. 1987. Managing public policy. Boston, MA: Little, Brown and Company.

Maynard-Moody, Steven, and Michael Musheno. 2003. Cops, teachers, counselors: Stories from the

front lines of public service. Ann Arbor: Univ. Michigan Press.

Nightingale, Demetra Smith, and Nancy Pindus. 1997. Privatization of public social services: A

background paper. Prepared for the U.S. Department of Labor, Office of the Assistant Secretary for

Policy. Washington, DC: The Urban Institute.

Ospina, Sonia, and Jennifer Dodge. 2005. Narrative inquiry and the search for connectedness:

Practitioners and academics developing public administration scholarship. Public Administration

Review 65 (4): 409–23.

Sandfort, Jodi. 2000. Moving beyond discretion and outcomes: Examining public management from

the front lines of the welfare system. Journal of Public Administration Research and Theory

10:729–56.

Schoeni, Robert, and Rebecca Blank. 2000. What has welfare reform accomplished? Impacts on welfare

participation, employment, income, poverty, and family structure. Working Paper 7627, National

Bureau of Economic Research, Cambridge, MA, February.

Scott, Patrick, G. 1997. Assessing determinants of bureaucratic discretion: An experiment in street-level

decision making. Journal of Public Administration Research and Theory 7 (1): 35–57.

Segal, Elizabeth A., Karen E. Gerdes, and Sue Steiner. 2004. Social work: An introduction to the

profession. Belmont, CA: Wadsworth/Brooks Cole.

Sommerfeld, David, and Michael Reisch. 2003. Unintended consequences: The impact of welfare reform

in the United States on NGOs. Voluntas: International Journal of Voluntary and Nonprofit

Organizations 14:299–320.

Soss, Joe. 2000. Unwanted claims: The politics of participation in the US welfare system. Ann Arbor,

MI: University of Michigan Press.

Thiel, Sandra V., and Frans L. Leeuw. 2002. The performance paradox in the public sector. Public

Performance and Management Review 25:267–81.

Watkins, Celeste. 2003. The incomplete revolution: Constraints on reform in welfare bureaucracies.

PhD diss., Harvard Univ.

Ziliak, James P., David N. Figlio, Elizabeth E. Davis, and Laura S. Connolly. 2000. Accounting for the

decline in AFDC caseloads: Welfare reform or the economy? Journal of Human Resources

35:570–86.

Dias and Maynard-Moody Contracts, Conflicts, and the Performance Paradox 211