The federal government can and should require information reporting from states to assess the welfare reform effectiveness and to assure that state programs perform according to federal mandates. However, as proposed, reporting requirements impose a significant personnel and financial burden upon the states. To minimize this burden, the federal government must be more scientific, more collaborative and pay greater attention to the reasonableness and feasibility of mandated state reporting requirements.
The federal government must recognize that existing state information systems were primarily designed to provide automated support for transaction processing and case management -- not data collection for analysis and reporting -- and that a variety of data elements might satisfy federal information needs equally well. Therefore, data elements should be selected from those already used by states to fulfill their management needs.
The process for determining what data elements are available should include consultation with states and other potential providers of data. Otherwise, data for federal reporting will be collected as an afterthought and likely be inaccurate, and state resources will divert from using information systems to promote self-sufficiency toward meeting federal reporting requirements. The current and apparently arbitrary and closed process of selecting the proposed reporting requirements has created further distrust between the states and federal government, and will undermine the national effort to reform welfare.
PROBLEMS WITH REPORTING REQUIREMENTS
Failure to Employ a Systematic and Methodical Approach to Identify Reporting Requirements -- While state and local governments agree that outcome measures are necessary to understand the impact of programs on families, and performance measures may be necessary to assure state compliance with legislated mandates, many reporting requirements do not seem linked to outcome measures or performance indicators.
The NPRM (notice of proposed rulemaking) imposes 178 reporting requirements -- 110 more than the emergency data requirements, which were more than previously required under Aid to Families with Dependent Children. The state and local governments responsible for all this data collection wonder how all of this data will be used. Failure to employ a collaborative, methodical approach to determine outcome measures, performance indicators and data requirements has created the perception of a scattershot attempt to collect potentially relevant data without assessing each reporting requirement's degree of relevance, feasibility of collection and cost.
The Requirements Are Costly -- The extent of these reporting requirements requires significant personnel and fiscal resources. Although the Office of Management and Budget (OMB) and state estimates of the personnel time needed to fulfill the reporting requirements differ significantly, even the OMB estimate of 33 minutes per case requires significant personnel resources. Unfortunately, there is no comparable cost estimate for updating and creating the information systems and business processes necessary to collect this data. States must build new systems, rebuild old systems, and redefine data fields and business processes.
In the majority of states, county governments administer human services systems, and each of these counties will encounter the same technology problems. Much of the data requested either does not yet exist in state human service databases or, if available from other programs, cannot be linked to TANF recipients, nor can that information be accessed in a way useful for reporting. The American Public Welfare Association (APWA) estimates that states will spend $1.1 billion to update current systems and add new functionality to meet TANF program reporting and eligibility requirements alone. When the costs to meet the new set of reporting requirements are taken into account, the total expense is likely to be significantly greater.
The Requirements Divert Time and Money -- If the creation and upgrade of these information systems helped administer state welfare reform programs, then the creation of such systems would be an investment in the effectiveness of state programs. However, preliminary research conducted by the Rockefeller Institute of Government, APWA and the National Association of State Information Resource Executives indicates that the development of systems to meet federal reporting requirements competes with the development of systems to manage welfare reform. In essence, states and counties have one set of system requirements to administer their welfare programs and another to fulfill federal reporting requirements. If reporting requirements were more closely aligned to the tasks necessary to deliver human services, significant efficiencies could be obtained.
Redundant Process of Outcome Determination -- Under the authority granted in PRWORA, states and counties define their welfare reform outcome measures. For instance, New York state includes a significant number of additional reporting requirements in its authorizing legislation.
While most states will be interested in defining outcomes in a way that satisfies local requirements, facilitation by the federal government might promote some commonality between the data elements collected to measure achievement according to local outcome measures and federal reporting requirements. Most states are finding the process of defining their data requirements is difficult and would likely welcome coordination between their efforts, other states and the federal government.
Requirements Antagonize States -- The spirit of devolution contained within PRWORA created a unique opportunity to change the relationship between HHS and states from enforcement and punishment to collaboration. However, by insisting upon reporting that only seems to regulate states, and by utilizing a process that appears to be closed and capricious, HHS undermines its authorizing role and continues its antagonistic relationship with states.
Collected Data Will Be Inaccurate -- Despite the inadequacy of information collection systems and processes, states will provide the required data rather than have their block grants cut or face other sanctions.
Also, since data collection for the reporting requirements is not part of the service delivery process, data will be collected and verified haphazardly by local and county officials, third-party service providers and others with insufficient resources for accurate data collection and verification. This inaccurate data will be consistently questioned by researchers and not prove useful for future public policy and program decisions.
COLLABORATIVE IDENTIFICATION OF REQUIREMENTS
While state and local governments would benefit from additional time and money to build appropriate systems that meet federal reporting requirements, the development of an alternative set of reporting requirements promises to be the most beneficial.
Outcome measures to determine policy success, program performance indicators to measure program efficiency and the information systems used to collect and report on these indicators and measures are inextricably linked. For instance, data that is a byproduct of the service delivery process -- such as workforce skills, educational attainment or job training program enrollment possessed by the client -- is data likely collected as a case management function. Data not integral to the service delivery process is likely to be old, false or both. Unfortunately, PRWORA simply indicates that all data might be useful, rather than trying to determine what is needed to achieve goals and what is available from existing service delivery systems.
Today, government at every level is turning to outcome measures to assess their policies and programs. However, these outcome measures must be developed in concert with the data requirements and business processes necessary to measure them. In the NPRM, failure to link reporting requirements to outcome measures, information system development or business process reform foretells of arbitrary outcome measurement, inaccurate data and tax dollars wasted on redundant information systems. Fortunately, it's not too late to change.
Larry Singer -- an expert on strategic computing with 12 years experience in the information technology industry serving all levels of government -- is president of Public Interest Breakthroughs Inc., of Vienna, Va.
April Table of Contents
