File for unemployment benefits in most American states and you'll encounter a website that looks like it was designed during the Bush administration, requires a browser setting your computer disabled for security reasons, and times out after 15 minutes of inactivity, losing all your data. Try to access your tax records online and you'll navigate a multi-step verification process that sends physical mail to your address and requires a return visit to the site within seven days. Apply for a small business permit in most jurisdictions and you'll be asked to print a PDF, fill it out by hand, and fax or physically mail it to an office that processes requests within 6- 8 weeks.
Meanwhile, you can order a car, file a complex insurance claim, open a brokerage account, and manage a mortgage — all from your phone, in minutes, with a user experience that has been refined by professional designers and tested with thousands of users.
This gap is not a natural phenomenon. It is the product of specific choices — choices about procurement, about prioritization, about whose time is considered valuable — that have compounded over decades into a technology environment that systematically disadvantages Americans in their relationship with their government.
And it has consequences for democracy that go well beyond inconvenience.
HOW GOVERNMENT BUYS TECHNOLOGY (AND WHY IT PRODUCES BAD RESULTS)
The federal government spends approximately $100 billion per year on information technology. State and local governments spend substantially more. This is not a small market. The problem is not that government doesn't spend on technology — it's how it buys.
Federal technology procurement is governed by the Federal Acquisition Regulation, a document of staggering complexity that runs to thousands of pages and has been built up over decades by the accumulation of requirements designed to prevent specific past failures. It requires extensive documentation before any work begins. It defines contract structures that make iterative, user-centered design — the approach that produces good commercial software — extremely difficult to execute within the procurement framework.
The vendors who thrive in this environment are not the ones who build the best software. They're the ones who best understand how to navigate the procurement process: large consulting firms and defense contractors with decades of federal contracting experience, armies of compliance specialists, and deep relationships with procurement officials. Their value proposition is not technological excellence. It is bureaucratic navigation.
The result is that the federal government consistently pays premium prices for below-market technology, built by organizations that have no competitive incentive to improve because their advantage is relational, not technical.
This is not a mystery or a scandal. It is the predictable outcome of an incentive structure that rewards the wrong things. Changing the outcomes requires changing the incentives.
WHAT GOOD GOVERNMENT TECHNOLOGY LOOKS LIKE
The United Kingdom's Government Digital Service, established in 2011, is the clearest large- scale demonstration of what good government technology looks like when the conditions for it are created.
GDS started with a simple premise: government digital services should be designed around users, not around agency organizational structures. The existing GOV.UK web presence was fragmented across hundreds of separate agency websites with divergent design, inconsistent information, and poor usability. GDS consolidated it into a single coherent platform, designed to user research standards, with clear information architecture and consistent interaction patterns.
The results were measurable. User satisfaction increased substantially. Cost per transaction fell — in some cases dramatically. Completion rates for digital services (the percentage of people who started a process online and finished it without abandoning or seeking help) improved significantly.
GDS's approach was methodological, not technological. They did user research — actually observing people trying to accomplish tasks, identifying where they got confused, and designing to remove those friction points. They iterated based on real usage data rather than requirements documents. They measured success by whether users could accomplish what they came to do, not whether the system had delivered the specified features.
This approach is standard practice in commercial software development. It is not standard practice in government technology procurement, where success is measured by contract deliverables rather than user outcomes.
In the United States, the United States Digital Service and 18F have applied similar methodologies to federal technology projects since their creation in 2014-2015. Their record includes genuine successes — the rescue of Healthcare.gov, significant improvements to VA digital services, the development of login.gov as a shared identity platform — and genuine constraints. Both organizations operate at the margins of a procurement system that systematically works against what they're trying to do.
THE PARTICIPATION STAKES
The civic technology gap is not just inconvenient. It has documented effects on democratic participation.
Voter registration is a clear case. States that have moved to automatic voter registration — where eligible citizens are registered when they interact with government services like the DMV — have seen significant increases in registration rates, particularly among young voters and voters who have historically had lower registration rates. The friction removal is the mechanism: you don't have to navigate a separate process to register; it happens as a byproduct of an interaction you were already having.
Online voter registration, where available and well-designed, increases registrations from populations with lower historical registration rates. Same-day voter registration, which requires a functional voter registration interface at polling locations, significantly increases turnout when it's implemented with adequate technology support.
Conversely, benefits access is a clear case of how bad technology suppresses participation. Research on SNAP, Medicaid, and unemployment insurance enrollment consistently finds that eligible people who don't receive benefits they qualify for cite the complexity of the application process as a primary barrier. When that process runs through a poorly designed online interface, the complexity is amplified. People who most need benefits — who are often experiencing acute stress, who may have limited technology access, who may have lower digital literacy — are the people least able to navigate a bad interface.
This is not a metaphor. It's an empirical relationship between government technology quality and the effective reach of government programs. The technology is not neutral. Its quality determines who gets access to what government offers.
THE DISTILL THE BILL PROBLEM
Civic technology is also about information, not just services.
The texts of legislation, regulatory filings, court decisions, and government reports are public records. They are, formally, the primary documents of democratic governance — the written record of decisions made in the public's name. They are also frequently inaccessible to most people without specialized training: written in legal or technical language, organized around procedural logic rather than the questions citizens actually have, and distributed through interfaces that were designed for professionals with institutional access rather than interested citizens.
The result is a systematic information asymmetry: the interests with resources to employ professional staff who can read and interpret public records have a significant advantage over citizens and advocacy organizations without those resources.
Technology can close this gap significantly. Natural language processing tools can translate regulatory filings into plain English. AI-powered search can make the content of dense legislative text findable by people who don't know the technical vocabulary. Visualization tools can make the relationships between pieces of legislation, the flow of money in government contracting, and the pattern of regulatory decisions legible to people without specialized training.
These tools exist. Some are being built by civic technology organizations, some by academic researchers, some by commercial companies. What they generally lack is the distribution infrastructure to reach the people who would benefit most from them — not the already- informed policy professionals, but the citizens trying to understand what their government is actually doing.
THE INVESTMENT CASE
The case for serious investment in civic technology is not primarily altruistic. It's economic.
Better government services cost less to deliver. A digital service that allows a citizen to complete a benefits application in 20 minutes at home costs significantly less than a process that requires multiple in-person visits to a government office, generates substantial rework from incomplete applications, and requires case workers to spend significant time on administrative rather than substantive work.
The UK government's estimates from the GDS work suggest that well-designed digital services can reduce cost per transaction by 80-90% compared to in-person or paper-based equivalents. Multiplied across the full volume of government interactions, the savings potential is enormous.
Better government technology also has economic development value. When business permitting is fast and digital, when occupational licensing can be processed quickly, when building permits don't require weeks of in-person follow-up — the economic friction cost of doing business with government falls. This matters particularly for small businesses and entrepreneurs who lack the staff resources to manage complex government processes.
THE POLITICAL WILL PROBLEM
The civic technology gap persists not because solutions are unknown or resources unavailable. It persists because the political incentives to fix it are weak compared to the institutional forces that benefit from the status quo.
The vendors who profit from the current procurement system are a concentrated interest with resources to protect their position. The citizens who suffer from bad government technology are a diffuse interest with limited ability to coordinate around this specific issue.
There is no civic technology constituency analogous to the military-industrial complex or the healthcare lobby that keeps their respective procurement systems intact. There are civic technology advocates, reform-minded officials, and organizations working on the problem. But the political economy of procurement reform works against them.
This is solvable, but it requires choosing to solve it — making government technology quality a political priority, reforming procurement to reward outcomes rather than compliance, and building the institutional capacity (along the lines of GDS or USDS) to maintain standards over time.
The technology to make American government work better is available. The question is whether we have the political will to use it.
This was the final installment of "The American Possibility," a series on the structural conditions for a renewed American dynamism.