Active Contract Management: How Governments Can Collaborate

Active Contract Management: How Governments Can

Collaborate More Effectively with Social Service

Providers to Achieve Better Results

I. Introduction

Many critical functions of government social service

agencies involve contracting with private service

providers. Increasing the effectiveness of

procurements is therefore essential to improving

governments’ ability to deliver social services. Often

government agencies assume that their role is

complete once a contract is signed and shift to a

narrow focus on processing invoices and enforcing

compliance. However, some of the most important

work for government comes during the course of the

contract, when real-time improvements to service

delivery can drive better outcomes for the people

being served. Agencies should use procurement and

contracting to establish the foundation for an

ongoing collaboration with contracted service

providers to strategically improve performance.

Active contract management (ACM) is a set of

strategies developed by the Harvard Kennedy School

Government Performance Lab (GPL) in partnership

with government clients that apply high-frequency

use of data and purposeful management of agency-

service provider interactions to improve outcomes

from contracted services. We have seen these

strategies adopted by agencies responsible for child

welfare, workforce development, prisoner re-entry,

developmental disabilities, and addiction treatment.

Like PerformanceStat, ACM consists of high-

frequency, data-informed meetings designed to

produce action that improves performance. Where

PerformanceStat is often implemented in order to

improve core government functions, ACM focuses on

collaborations between government agencies and

social service providers to improve contracted

services. ACM empowers leaders to detect and

rapidly respond to problems, make consistent

improvements to performance, and identify

opportunities for reengineering service delivery

systems.

This policy brief describes the problems that ACM

aims to solve, discusses the benefits of using these

strategies, and outlines elements of effective ACM

systems. Examples in this brief draw primarily on

the GPL’s engagements piloting ACM strategies with

state and local governments across the country.

While this brief is focused on how ACM can drive

social outcomes, we have seen these same strategies

improve results of other government contracts,

including for waste collection and road construction.

II. How typical government contract

management practices fall short

Typical contract management practices fall short in

three ways: 1) government agencies do not

purposefully attempt to improve service provision,

2) agencies fail to collect and use data to improve the

delivery of contracted services, and 3) agencies fail

to collaborate effectively with providers to improve

outcomes.

Piloting Active Contract Management

The GPL has helped more than a dozen

government agencies implement ACM strategies

as part of its technical assistance, including:

 In New York, the GPL assisted the state in a

Pay for Success (PFS) project that provided

job training and reentry support services for

ex-offenders leaving state prison with the

goal of reducing recidivism.

 In Seattle, the GPL collaborated with the

city’s Human Services Department (HSD) to

reorient homelessness service contracts to

focus on helping clients achieve stable

housing.

 In Rhode Island, the GPL helped the

Department of Children, Youth, and Families

(DCYF) reprocure and better manage

services offered to children and families as

part of an agency-wide turnaround effort.

 In Illinois, the GPL worked with the state’s

Department of Children and Family Services

(DCFS), the Department of Juvenile Justice

(IDJJ), and local probation departments to

expand wraparound services for justice-

involved foster youth.

1. Lack of purposeful attempts by

government agencies to improve

outcomes of contracted services

Governments often lack strategies for systematically

improving the results that contracted services

achieve for clients. In our work, we have observed a

homelessness services agency that monitors the

number of showers provided, without tracking the

percentage of homeless families housed; a

behavioral health agency that lacks information on

how service providers delivering the same outpatient

therapies perform relative to one another; and a

child welfare agency that provides annual funding to

community providers without checking to see how

client needs may have shifted.

In these and other governments, contracting is

treated as a back-office function, handed off to

administrative units focused on compliance and

invoicing, which are disconnected from the

programmatic objectives of the agency. Staff track

billing accuracy and volume metrics in isolation

from any performance management activities.

This compliance focus is particularly costly in the

context of overly prescriptive contracts, in which

time-consuming change orders are needed in order

to make even minor operational adjustments. What

quality assurance activities exist occur at the case

rather than systems level, and often focus on

determining whether providers are meeting

licensing requirements and basic standards of care.

Incidents of poor performance are addressed in

isolation, and necessary improvements are specified

only in one-off provider-specific corrective action

plans. For example, in one agency responsible for

providing vocational training to TANF recipients,

providers are regularly relicensed without any

systematic efforts to address the state’s near-bottom

of the nation ranking on employment outcomes for

this population.

Rarely do agencies look across programs, providers,

and internal operations to measure the actual results

of service delivery or identify opportunities for the

entire system to work better.

2. Failure to collect and use meaningful

data to improve service delivery

For many government programs, data is not used

effectively to monitor, inform, or improve service

delivery—the right data is not collected or shared

between governments and providers, and data that is

collected is not used to drive action.

Lack of data usage often begins with the limited

availability of actionable operational data. Often

governments aren’t collecting the right data from

social service providers. Providers delivering similar

programs often report on different data measures.

Incomplete, inconsistent, or late data reporting is

common when data collection is not prioritized.

Frequently, government administrative data is not

shared with providers in a timely manner, so that

even the most data-driven providers have trouble

accessing information needed to adjust operations.

Similarly, data on provider operations is often

opaque to government, making it difficult for

government to track key operational indicators. This

fragmentation hinders attempts to calculate

important metrics that require both government and

service providers to share data with each other.

In one child welfare agency, providers of

maltreatment prevention services were never given

access to information about the abuse and neglect

outcomes of their clients after their care had ended,

substantially impairing their ability to assess

program success.

Even when the right data is collected, government

officials often fail to use it in a way that can drive

management decisions. Contract management

offices generally have limited data and analysis

capacity. Without available staff that can review data

and ask questions to drive operational or practice

reforms, provider data is used only retrospectively to

validate contract compliance, rather than to make

real-time improvements. The data is not used to

identify high performing providers or best practices

or to inform future policy decisions. When data is

published, it is often in aggregate format and offered

without interpretation, making it difficult to discern

the connection between specific practices and

results.

For example, we observed one human services

agency that annually published a report showing

how client outcomes varied for different types of

services received by agency clients. However, raw

data about provider-specific results was not shared

with the agency’s frontline staff so that referrals

could be adjusted, nor used by the agency’s contracts

units to inform award decisions in upcoming

procurements.

The failure to collect the right data and use it is

incredibly frustrating to both providers and agency

staff – who are spending so much of their time

reporting and tabulating data that is not being used

effectively.

3. Lack of productive collaboration with

service providers

Often, government does not regularly engage with

service providers to assess how service delivery is

going or what actions could improve it. Once a

contract is negotiated and signed, contact between

the agency and providers is focused on compliance

with contract terms, invoicing, or responding to

emergencies. Contract monitors may conduct annual

site visits to ensure adherence to contractual

guidelines or licensing requirements, but these visits

rarely involve conversations about how performance

can be improved.

Performance problems are often addressed

punitively resulting in penalties without shared

learning. With relationships between agencies and

providers often adversarial in nature, few

opportunities exist to build the trust necessary for

jointly driving reforms.

Additionally, providers rarely receive information

comparing their performance to that of their peers,

which limits opportunities to identify and replicate

best practices.

III. Active contract management

Active contract management consists of high-

frequency data-informed collaboration between

government agencies and service providers that is

focused on outcomes. The GPL has been developing

and piloting ACM in our work with state and local

governments across the U.S. In each of these

jurisdictions, the government found that

implementing this new approach to managing

contracts allowed both the government and

providers to solve problems and make sustained

improvements in service delivery.

We observe three primary benefits when agencies

actively manage the performance of their contracted

providers:

 Reactive troubleshooting: Real-time, rapid

identification of performance problems allows

governments and providers to make immediate

course corrections. In the New York State

criminal justice project, the state and service

provider identified an issue with low enrollment

numbers at the project’s outset, and immediately

adjusted referral processes to address this.

 Incremental improvements: Persistent

attention on critical performance issues can

facilitate ongoing improvements in results. This

work often involves agency leaders reviewing

data on whether programs are doing

better this month than last month (and this year

compared to last), and holding agency staff

accountable for continually refining their

processes and those of their contractors to

produce rising performance trends over time. In

Rhode Island, persistent attention to conducting

clinical child assessments resulted in steady

improvements in completion rates, which

enabled services to be better tailored to client

needs.

 Systems reengineering: Often, dramatic

performance improvements can be achieved by

revising protocols for matching clients to

services, partnering with other government

agencies to overcome care coordination siloes, or

shifting resources to the most effective programs

and providers. In Illinois, DCFS recognized a

multi-week lag between reporting of

circumstances that present risks to youth in

state care and notification to the agency.

Handwritten reports were faxed to a central

office, requiring manual transcription into a

state computer database. As of February 2017,

the agency was in the process of automating this

system, which should decrease the time between

the occurrence of incidents and potential

referrals to providers for additional services

from several weeks to less than a day.

These benefits are made possible by three common

components of ACM strategies:

1. High-frequency reviews of real-time

performance data

Frequently reviewing real-time, operationally

meaningful performance data enables agencies and

providers to rapidly identify and address service

delivery problems before they become ingrained or

unfixable. It also creates feedback loops for

providers and program staff to assess and refine

changes to service delivery.

In the New York State criminal justice project,

project partners monitor data on a biweekly basis to

determine the percentage of individuals being

released from prison attending the first day of job

training services. When this percentage falls below

targets, immediate course corrections are designed

and implemented.

Seattle’s HSD is also using performance data to

collaborate with providers to monitor progress,

detect problems, and resolve issues in real time.

Data are reviewed monthly across six key

homelessness metrics (including permanent

housing, housing stability, and returns to

homelessness)

that are used to drive towards more

effective service delivery.

2. Regular, collaborative meetings between

service providers and agencies

Regular, collaborative meetings enable service

providers and agencies to swiftly troubleshoot

problems, solve specific operational challenges, and

identify opportunities for broader systems

transformation. These meetings often feature

substantial data reporting and analysis, case reviews,

and the sharing of best practices.

A typical active contract management meeting has

three agenda items: 1) follow up on action items

previously identified and a summary of lessons

learned to-date, 2) discussion of progress against

high-priority performance measures and joint

troubleshooting of any problems that may have

emerged, and 3) in-depth examination of a topic

critical to success but not regularly reported or

reviewed.

The Seattle HSD is establishing three types of

meetings as part of its ACM strategy:

 Monthly data sharing and contract check-ins

between HSD and service providers, where

progress reports that measure outcomes on six

key homeless metrics for each provider are

reviewed in order to troubleshoot program-

specific problems.

 Quarterly internal HSD meetings, in which HSD

staff assess system-wide performance on key

outcomes and indicator metrics to obtain

internal consensus on strategic programmatic,

funding, and policy decisions to improve results.

 Quarterly executive meetings between HSD and

service providers, where homeless population

outcomes are reviewed, general trends are

For more information, see pages 5-6 of the GPL brief on

the Seattle homelessness services project

identified, and the relative performance of

service providers is used as a learning

opportunity.

In collaborative meetings such as those being held in

Seattle, reviewing data is often only the start of the

conversation. It is important that agencies and

providers use lessons from the data to develop and

implement strategies to improve outcomes.

It is also often useful to convene groups of providers

targeting similar populations to facilitate peer

learning, share effective practices, and resolve

common barriers. Improvements in service delivery

frequently require adjustments by both the referring

agency and the service provider, and can be

replicated across multiple providers offering similar

services.

In Rhode Island, DCYF facilitates monthly ACM

meetings with leadership teams from each of its four

front-end family preservation service providers. Ad

hoc working groups consisting of agency and

provider staff meet more frequently as needed to

address specific performance issues raised in

executive meetings. “This active contract

management experience has provided us with new

clarity on the goals we are trying at achieve,”

observed one provider of child welfare services in

Rhode Island. “This clarity makes it easier for my

staff, and informs how we need to adjust our

program model moving forward.”

3. Forward-looking performance

management roadmaps

Performance management roadmaps help agencies

proactively direct in-depth analytical attention

toward topics and practices that are critical to

success but may not be regularly reported or

reviewed. Roadmaps can be organized as running

calendars of priorities for deeper investigation. Often

additions will be made to follow-up on specific

operational challenges uncovered in progress report

reviews. Without such roadmaps, it is easy for

agencies and providers to inadvertently focus solely

on reactive adjustments or incremental

improvements, missing opportunities for systems re-

engineering.

Some roadmap topics benefit from advanced

quantitative analysis, such as regression-

discontinuity techniques to evaluate referral

decisions. In other circumstances, it may be helpful

to pull and review individual case files to inform

interpretations of aggregate time series data.

Rhode Island’s DCYF has organized ACM according

to a forward-looking 12-month performance

improvement roadmap, which is regularly updated.

This roadmap contains the plan of which in-depth

analyses will occur each month at the ACM

meetings. Throughout the course of the first year of

ACM, the agency and providers examined family risk

factors associated with higher service needs,

assessed providers’ strategies for identifying and

matching families to programming, and determined

which service components are most critical for safely

keeping families together, among other topics.

When beginning ACM with a new group of

providers, it is useful for early in-depth topics to

focus on identifying key performance metrics for

regular review and generating benchmarks and

targets.

IV. How can governments maximize the

effectiveness of ACM practices?

Implementing ACM nearly always requires agencies

to restructure internal operations in several ways:

picking the right measures to track, using data to

drive action, creating a culture of partnership to

improve performance, and elevating the status of

contract activities.

 Select the right measures to track:

Choosing appropriate measures for regular attention

is a critical first step in transitioning provider

reporting requirements from compliance to

performance. When agencies track activity

measures, they should chose indicators that are

linked to successful service delivery, such as time

from referral to service enrollment, program take-up

and completion rates, and spending per client. For

example, for services designed to stabilize

individuals in crisis, reaching clients within hours of

referral can be the difference between

hospitalization and successful diversion to

outpatient care.

Medium and longer-term outcome metrics, such as

recidivism rates, are critical to track as well. Input

metrics, such as meetings conducted or staff-hours

billed are rarely helpful for assessing provider

performance.

Once measures have been selected for tracking,

agencies and providers should jointly identify

benchmarks against which performance trends can

be contextualized. A combination of historical, peer,

and best-practice targets are helpful in determining

performance goals.

 Use data to drive action:

Data should be used to identify the right problems,

creatively craft solutions, and subsequently drive

sustained action. Dashboards, which enable regular

attention to a small set of metrics, are often helpful

but rarely sufficient without active interpretation

and pre-meeting analysis of operational

implications. Generating real understanding

requires a capable analyst to look at data,

contextualize it in relation to program operations,

and ask questions about it to drive a reform.

However, it can also often be as simple as examining

why a performance trend has (or hasn’t) changed

from the previous period, or why one vendor has a

different client mix than another.

Rhode Island DCYF discovered that families who

regularly participated in family preservation services

for at least 3 months were half as likely to experience

subsequent child welfare involvement compared to

those who completed less than 3 months of service.

Based upon this insight, providers created a handout

to help convince newly referred families to engage in

services.

Supplementing analysis of aggregate performance

data with in-depth review and discussion of

individual case files is often critical to correctly

interpreting the aggregate trends. Reviewing a small

number of cases from a mix of clients with good and

poor outcomes can often distinguish among

alternate explanations for performance trends and

point the way to the right implementation changes.

In some cases, it can be helpful to have multiple

providers learn from each other based on relative

performance. For example, six providers partnering

with Illinois to expand wraparound services for

justice-involved foster youth review weekly data on

success engaging families in team meetings.

Providers who ranked lower actively seek out those

who ranked best to learn from them on how to

improve their performance. Frontline staff from all

six providers also regularly meet to share best-

practices and brainstorm solutions to barriers to

care.

 Create a collaborative performance

management culture:

It is important to develop an active contract

management culture that encourages a collaborative

Further analysis suggested that this difference was not

simply the result of better functioning families being more

likely to agree to services.

partnership with providers and shared ownership by

government for improving client outcomes.

Thoughtfully considering the frequency, timing,

structure, and attendees of regular meetings to

maximize productivity is critical. The tone set by

agency leaders should be constructive rather than

punitive, as the focus should be on generating

solutions toward the common goal of improving

service delivery. In one jurisdiction, the program

director of a provider participating in ACM reflected

on this, saying, “Our program supervisors appreciate

the chance to do deep dives in collaboration with

agency staff – it’s no longer us versus them.”

Commitment by agency leadership is also critical.

Regular executive attention to provider results

throughout the life of the contract – and not just

when problems arise – demonstrates to providers

that performance improvement is an agency priority.

If major barriers arise, it is helpful to have robust

existing relationships between executives in order to

resolve problems as quickly as possible.

Similarly, agencies should create mechanisms for

frontline staff to immediately elevate provider

performance problems to ACM staff and agency

leadership. For example, one agency is considering

an online tool through which frontline caseworkers

can report to the contracts unit any problems

accessing transportation, clothing, and other

contractually required resources. Data generated

from these reports will help the agency detect

common issues with specific providers, develop

strategies to address these problems, and assess the

success of remedial actions. Agency data and

evaluation units should regularly engage with front

line staff to get their suggestions for which processes

could most benefit from deep dives during active

contact management meetings.

 Regularly share government

administrative data with providers:

It is as important for agencies to share government

administrative data back with providers as it is for

providers to consistently report high-quality data to

agencies. Regular sharing of government

information about what happens to clients after they

complete provider services, for example, enables

providers to learn and adjust service delivery

operations.

Some agencies establish joint management

information systems in which the same database is

used by the government for maintaining

administrative records and by providers for tracking

day-to-day client management and service delivery.

In Seattle, HSD and homelessness providers have

access to the same database, enabling each to

generate the same performance reports.

Other agencies chose to share administrative data

with providers by regularly generating reports from

agency datasets; however, if not automated,

sustaining this practice requires ongoing staff

attention. Alternatively, linking a small portion of

payment to one or more outcome measures for

which providers don’t otherwise have access

institutionalizes these data feedback loops for

providers without requiring additional government

capacity.

 Elevate the status of procurement and

contract management activities:

Adequately resourcing agencies’ contract

management functions is critical for improving

provider performance. Contract managers must be

perceived both internally and externally as senior

leaders responsible for driving a key part of the

agency’s mission. For example, to oversee its active

contract management practice, Rhode Island’s DCYF

established a Contracts and Compliance unit that

reports to the director’s office.

To fully implement ACM practices, staff will also

need to have time and training to effectively review

performance data, flag problems, and work with

providers to implement any necessary changes. Staff

must be empowered to work across agency units to

access information, breakdown barriers, and

influence change.

Since active contract management is time intensive,

agencies may need to streamline other contract

office work to free up time for performance work.

Seattle’s adoption of active contract management for

homelessness services was paired with a

consolidation of contracts that significantly reduced

invoicing and change orders.

Additionally, training current contract managers or

hiring a new type of employee to actively manage

contracts may be necessary for building

organizational capacity. For example, Rhode Island

conducts workshops to teach agency employees how

to use provider data to detect, diagnose, and respond

to unexpected performance trends.

Seattle HSD considered whether it should hire

dedicated staff responsible for the fiscal monitoring

of contracts to allow contract managers to solely

focus on performance improvements or add contract

For more information, see page 5 of the GPL brief on the

Seattle homelessness services project

managers so that each could handle both fiscal

monitoring and performance for a smaller case load.

Other jurisdictions separate contract compliance and

performance management duties to allow

specialization and to more fully leverage the

technical capabilities of data-fluent staff.

V. Questions on Spreading and Sustaining

Active Contract Management

As we continue to help governments implement

ACM, we are testing solutions to questions about

spreading and sustaining this practice.

How can active contract management be

spread across an entire agency when

resources are unavailable to sustain this

high-intensity approach for every contract?

Agencies will need to prioritize which contracts are

important enough to employ the full set of ACM

techniques. Vendors who are delivering high impact

services or new services, vendors who are at a high-

risk of running into obstacles or have had

historically poor results, and vendors whose

contracts are large in dollar value or in complexity

should be prioritized to receive the most attention.

One way to reach a broader audience with scarce

resources is by using performance improvement

sprints, which focus brief but intensive attention on

a rotating set of contracted providers. Sprints aim to

solve a more limited set of performance issues than

what would otherwise be included in an agency’s

performance improvement roadmap.

Alternatively, a select set of universal performance

metrics could be required across all vendors. One

way to do this is to link a portion of payment for

services to one or more common indicators of

results, such as sustained employment for a certain

period following referral to services. Linking even a

very small amount of payment to results ensures that

data collection will be sustained throughout the

period of the contract.

How can agencies provide necessary

technical assistance to support providers

when performance gaps emerge?

Many contracted service providers, especially those

providing health and human services, lack

sophisticated data, finance, or strategy capabilities of

their own. While ACM can often help these providers

identify opportunities to improve service delivery, it

does not provide the in-depth support necessary for

providers to wholly revamp internal operations.

There are a few options for agencies to support

providers when performance gaps emerge. Some

agencies connect contracted providers with third-

party capacity building resources such as the U.S.

Small Business Administration or local business

associations. Others are exploring how to give

providers access to resources available to state

agencies, such as technical assistance organizations

or Lean process improvement consultants.

While we are unaware of any examples, agencies

might also experiment with joint data and analysis

training for government contract managers and

provider program leaders.

What is the potential role of funders in

requiring reporting around coordinated

outcomes metrics?

Many service providers receive philanthropic

funding in addition to government contracts.

Philanthropic grants typically include reporting

requirements on metrics that are disconnected from

those that government agencies prioritize through

ACM.

Philanthropies have an opportunity to coordinate

with government to require consistent reporting

around coordinated outcomes metrics in a way that

allows government, providers, and philanthropy to

identify successes, pinpoint challenges, and work

together to improve services.

VI. Conclusion

Government too often assumes that responsibility

for achieving outcomes is handed off to providers

once a service is contracted out. To achieve effective

delivery of social services, government must

maintain its share of responsibility for results during

the period of service delivery and be an active

participant in refining systems to improve client

outcomes. ACM offers government a set of strategies

for undertaking these partnerships with providers.

These strategies can be an important part of a

broader results-driven contracting effort to improve

the results achieved with contracted dollars.

The Government Performance Lab is grateful for support

from Bloomberg Philanthropies, the Corporation for

National and Community Service Social Innovation

Fund, the Dunham Fund, the Laura and John Arnold

Foundation, the Pritzker Children’s Initiative, and the

Harvard Kennedy School Government Performance Lab.

For more on reforming procurement practices, see the

GPL’s policy brief on Results Driven Contracting.