But there will probably be impacts. Federal agencies send about $1.1 trillion a year to state and local governments. That money accounts for 36 percent of state revenues on average, and states move a big chunk of those funds to local governments to sustain services from transportation and health care to food and housing assistance. But even in potentially leaner budgetary environments, state and local governments can do much to sustain and improve the effectiveness of vital public services.
How might they do that? They may have to make budget cuts of their own, but government services and infrastructure are real necessities. So, state and local governments must do less with more. That can only happen with big assists from process modeling and redesign, automation, and sharper decision-making. All involve technology.
For example, consider that the proposed cutbacks on federal Medicaid funding could shift a burden of more than $44 billion to states. That would increase average state Medicaid expansion spending by 26 percent. What roles can technology play in easing that pain?
On the decision-making front, generative AI can put the power of advanced analytics into the hands of planners and decision-makers without the need for specialized IT skills. Staff could use natural language queries to run scenarios based on varying levels of federal and state Medicaid support; the added costs of emergency room visits if different numbers of people end up uninsured; the likely impact of medical providers bailing on Medicaid patients because of lower reimbursements; and even the potential cost advantages of mandating AI screening and diagnosis to improve care and reduce costs.
Process modeling and redesign, aided by increasingly sophisticated tools to speed up historically painstaking efforts, might involve a thorough reassessment of how a state’s Medicaid program operates and how patients interact with it.
Automation then builds on that process redesign. On the front end, that means incorporating AI chatbots and copilots to answer questions, help patients navigate the health-care bureaucracy, get more timely care and identify coverage gaps. On the back end, automation means using AI and machine learning to verify documents, automate payroll and payments, optimize procurement, and spot fraud, waste and abuse.
AI — relevant, responsible, reliable AI — is a common theme here, as it must be in government applications if effective efficiency is the goal. For example, if federal transportation dollars shrink, what benefits might state and local governments get from revenue-boosting congestion pricing or public-private partnerships paid for through future toll revenues? How will those options affect how and with whom a department of transportation does business? Agentic AI, which combines machine learning, large language models and other AI technologies, is poised to propose solutions to these sorts of complex problems with minimal human interaction.
There are prerequisites to incorporating agentic or other AI effectively, and they’re familiar by now. AI needs data — ideally, diverse, fresh data — to draw its connections. A platform with a data management structure combining core system data with that of integrated third-party software drives both IT efficiency and more potent AI capabilities.
While agentic AI may be a new concept, states in particular are no strangers to harnessing AI and machine learning to advance the public good. Intelligent spend management across sourcing and contracts, procurement, payments, and supplier management has been a focus. Pennsylvania is using advanced data analytics to gather financial transactions, grants, contracts and other siloed information and distill it all into holistic views of spending by type of spend, agency, vendor and so on. That same system is serving up candidates for internal audits based on hard data, rather than using older manual, subjective processes.
The Indiana Management Performance Hub’s Transparency Portal leverages data analytics and data visualization tools to deliver detailed data on vendors, expenditures, revenues and assets to state leaders, personnel, government watchdogs, and others to help improve how the state spends its taxpayer dollars. California aims to use AI to analyze proposed legislation to make sure it's not duplicative and to check for revenue or expense impacts on laws already on the books.
Federal cutbacks may be coming for state and local governments. But even if they don’t, government efficiency efforts that incorporate process modeling and redesign, automation, and AI support for those making decisions in the complex, dynamic environments can help governments everywhere be more effective — and, yes, efficient.
Don Ingle is industry executive advisor for state and local government and education at SAP.
