The last mile for public policy

Public policy implementation hinges on thousands of decisions made by individuals at various levels, from local governments to nonprofits and service providers. Technical assistance plays a crucial role in refining these decisions. Originating from state and federal agencies, this assistance aims to guide those entrusted with executing policy.

However, current practices fall short. A federal program receiving millions in grants must first identify eligible entities, develop criteria, determine success metrics, and engage a nationwide network of stakeholders before distributing funds. A significant portion of this preparatory work involves interviews and criteria definition.

None of it matters if the funds fail to reach the organizations capable of improving the lives of those the grant intends to serve. Every major government initiative relies on hundreds or thousands of implementing partners, from local agencies to nonprofits. For instance, a state board of elections spends considerable time advising election clerks, advocating for rule changes, and educating the public on election processes.

Technical assistance is vital here too, but it's delivered in an ad hoc manner, typically at a single point in time, rather than continuously. Three key issues plague today's technical assistance delivery: bottlenecks that prevent timely expertise, scaling limitations as provider networks grow, and uneven support leading to knowledge loss.

Technical assistance is most needed when service providers face implementation challenges or edge cases, yet it often arrives too late. Without real-time guidance, providers revert to existing knowledge, undermining the value of initial knowledge sharing. Moreover, guidance often gets lost in emails and shared drives, and institutional memory fades when employees leave.

The problem worsens as federal policies evolve, budgets tighten, and turnover occurs. While much of TA funding goes to consulting engagements, consultants excel at providing accurate information but struggle with delivery and resilience. The lack of real-time connections across the TA network leads to fragmented knowledge when contracts end.

This issue is exacerbated by outdated tools and resource constraints faced by public servants. To truly transform technical assistance, we must shift from viewing it as a one-time deliverable to recognizing it as a lasting capability. This change involves treating TA as a continuously evolving infrastructure that grows stronger with use, rather than fading with turnover or contract expirations.

Modern software designed for interagency collaboration, communication, and knowledge sharing represents a pivotal shift. Such tools must capture intelligence from the entire network, ensuring that answers to questions from one county are accessible to others across the state. They must also preserve institutional memory by making knowledge available to everyone, not just the person who originally provided it.

By integrating these capabilities, technical assistance becomes a virtuous loop where experts continuously refine their answers based on emerging questions and challenges. Adding artificial intelligence to this loop can provide instant answers to routine queries while surfacing emerging implementation issues back to program staff.

This approach ensures that technical assistance is not just a fleeting service but a robust, growing capability integral to effective public policy implementation.