What Government Can Learn From Moore's Law

To make Washington more like Silicon Valley, we need expiration dates on legislation.


In a fit of nostalgia, I recently watched You've Got Mail with my 11-year-old daughter (don't judge me!). For those who haven't seen the 1998 chick flick, it tracks the first antagonistic, later tender, interactions between an independent bookseller (Meg Ryan) and the owner of a corporate bookstore chain (Tom Hanks) that just moved in across the street. As movie fate would have it, Ryan and Hanks develop an anonymous Internet romance even as his megastore is putting her little shop out of business.

To the modern eye, the technology in the film appears almost comically outdated: dial-up Internet connections, phones attached to answering machines, people getting their news from actual newspapers. In fact, the movie's plot wouldn't make sense today, given the plethora of online dating services and applications like FaceTime and Snapchat that allow for deeper, more intimate online communication than was available in the '90s.

The movie is also a reminder that exciting new technologies and the business models that sustain them are unlikely to dominate forever. However great an innovation, eventually, it may be replaced by something better, faster, or cheaper. Viewers in 1998 couldn't be blamed for thinking big chains like Hanks' Fox Books would permanently dominate the bookselling industry. In reality, the upstart online book retailer Amazon was already on its way to the top of the food chain. Remember Borders, Waldenbooks, or B. Dalton? Today in the world of brick-and-mortar bookstore giants, only Barnes & Noble remains.

In an open marketplace, a business that doesn't evolve to offer better goods and services at ever-more-affordable prices simply won't survive. That reality is particularly well-understood in places like Silicon Valley, which has been shaped by a folk understanding of Moore's Law, named for Intel co-founder Gordon Moore, who first observed in the 1970s that the number of transistors that fit on a computer chip doubles every two years, yielding cheaper and more powerful computers at a rapid rate. The result is a world in constant motion where risktaking is rewarded almost above all else.

"In today's fast-paced information age, everyone is expected to be constantly innovating and reinventing their business at the speed of 'Moore's Law,'" my Mercatus colleague Adam Thierer says. "Firms have to tear up their business plans every couple of years."

But while virtually all industries are engaged in a constant race to meet consumer needs, there's one sector where no such impetus need be present. "Governments never tear up any old law; they stay on the books seemingly forever," Thierer says. As a result, taxpayers at the federal, state, and local levels end up getting the same or worse services at higher and higher prices—exactly the opposite of what happens in the private sector.

Take K-12 education. During the last 40 years, the federal government has spent $2.1 trillion to school our kids, with spending per pupil tripling in real terms over that time. Neal McCluskey at the Cato Institute calculated that government at all levels invested an average of $166,773 on the 13-year education of a high school senior who graduated in 2011, compared to $57,602 (in real dollars) for a 1970 graduate.

Despite the dramatic increase, students' National Assessment of Educational Progress scores in reading, math, and science have shown no notable change since the 1970s. This means that each dollar we spend today buys us less education per capita than it used to.

The same is true of the war on drugs. In the last 30 years, the government has spent more than $1 trillion and put millions of people behind bars trying to reduce the supply and use of substances like marijuana and cocaine. That policy has been nothing short of a tragic failure.

A 2005 working paper by Carnegie Mellon's Jonathan Caulkins and Sarah Chandler uncovered an inverse relationship between incarceration rates and drug prices: Contrary to what you might expect, the more people we locked up for drug offenses, the cheaper dealers' black-market products became. A 2013 study in the journal BMJ Open likewise found that "despite increasing investments in enforcement-based supply reduction efforts aimed at disrupting global drug supply…illegal drug prices have generally decreased while drug purity has generally increased since 1990."

On the demand side, the National Survey on Drug Use and Health shows that past-month use of illicit drugs rose from 8.3 percent in 2002 to 10.2 percent in 2014, while past-year use rose from 14.9 percent to 16.7 percent. Taxpayers are funding a program that doesn't work.

Every year billions of taxpayer dollars are wasted in improper payments made by the lumbering federal bureaucracy. The Food and Drug Administration's horrifyingly slow process for testing and approving new pharmaceuticals means people die waiting for access to treatments that could have saved them. The U.S. Postal Service loses money just by existing, as does the government's national passenger rail service, Amtrak.

Total federal spending in 2000 was about $1.8 trillion. For 2016, that number is almost $4 trillion. We've added $13 trillion to the national debt over that same period, and what did we buy with all that extra spending? Not much. In fact, the bloat may actually be hurting the American people: Research has shown that when government grows, the economy shrinks. In 2011, for instance, economists at the European Central Bank produced a paper arguing that growth in government adversely affects economic performance.

For consumers, the contrast between government and markets is striking but not surprising. A business that regularly raised its prices while the quality of service went down wouldn't last long since customers would quickly turn to someone else. We have no such exit option when government fails us. Making the problem worse, the incentives facing elected officials and bureaucrats to prudently manage taxpayers' money are very weak. Interest groups are able to exploit this anodyne environment to obtain their own goals, often at the expense of the public welfare.

We deserve better. As Thierer puts it, we should be "applying a variant of Moore's Law" to government by forcing it to scrap its business model every so often the same way private sector institutions already do.

How do we do it? Where government and technology intersect, Thierer suggests that new laws and regulations should include a provision automatically sunsetting them 18 months to two years after enactment. "Policy makers can always re-enact the rule if they believe it is still sensible," he says. All the better if the change were retroactive, forcing legislators to re-evaluate rules and restrictions already on the books.

The short lifespan is necessary for a fast-moving industry like tech. But something similar should be done for all government programs so that they too are re-examined periodically. Would the war on drugs exist in its current devastating form if Congress had to pass and the president had to sign each piece every few years? I doubt it.

My years working in public policy have taught me that in practice, the only way to reduce spending is for Congress not to take action at the moment a program is set to expire. For instance, in March 2012 lawmakers rather accidentally imposed on the federal government a set of across-the-board spending cuts called "sequestration." The cuts went into effect despite strong congressional opposition, but only because the two parties could not find a mutually acceptable way to stop them.

The way to force policy makers to change their behavior and become more fiscally responsible is to put an expiration date on every program and regulation Congress passes. Under such a regime, the inertia and inefficiency that plague Washington would play to the favor of limited government advocates. This would finally allow us to start getting rid of some bad programs and replacing them with better ones, just as Amazon shook up the bookstore world.