Congress has become a less important player in the field of technology policy. Why did that happen, and what are the ramifications for technological governance efforts going forward?

I’ve spent almost 30 years covering technology policy. There was a time in my life when I spent almost all my time as a policy analyst preoccupied with developments in the federal legislative arena. I lived in the trenches of Capitol Hill and interacted with lawmakers and their staff morning, noon, and night.

In recent years, however, I have spent very little time focused on the Legislative Branch because it has effectively become a non-actor on technology policy. It is not that congressional lawmakers stopped caring about tech policy. Interest actually remains quite high—perhaps higher than ever before. Congress also continues to introduce lots of bills, host plenty of hearings, and issue mountains of press releases related to tech policy issues.

Nonetheless, all that interest and activity has not really translated into much important legislation. While it is hard to track tech-oriented legislative trends statistically because of the complication of defining “technology policy” over time, judged by substantive output, Congress has largely checked out of technological policymaking.

Think about digital privacy. How many years now have people been predicting a comprehensive “baseline” privacy bill would pass in each legislative session? It never happens. Perhaps it will this year, but if you would like to place a wager on it, I will take that bet.

Speaking of bets, for several years now, I have been wagering with friends that Congress will not pass federal legislation creating a national autonomous vehicles framework. Each session I win that bet. Keep in mind, a framework for driverless cars is far less controversial than privacy policy. Still, nothing substantive ever gets done in Congress.

Same goes for cybersecurity with lots of calls for big measures, but no final action. Folks are now also telling me to expect a big artificial intelligence bill one day soon. I sincerely doubt it. Again, I’ll bet on it if you’d like to lose some money!

Let me be clear, there may actually be some very good reasons why Congress should implement a national framework for privacy, driverless cars, and some AI policy issues. But all the wishful thinking in the world will not magically make it happen.

We need to entertain the possibility that Congress has largely checked out of the world of substantive tech policymaking and isn’t coming back. We may get a few big surprise measures here and there, as we did with clumsily-drafted FOSTA-SESTA. If anything, it is more likely that we instead see misguided legislative riders attached to non-germane measures during late night negotiations. But even haphazard efforts like those will be extremely rare. The days of Congress passing big bills like the Telecom Act of 1996 or the Cable Act of 1992 appear mostly over.

Why Congress Is No Longer the Major Player It Once Was

I think there are probably many obvious explanations for why Congress has checked out of tech policymaking, but let me try to boil it down to a couple of interrelated trends:

The “pacing problem” has intensified: The pacing problem refers to the inability of legal or regulatory regimes to keep adjust to the intensifying pace of technological change. There are just more emerging technologies than ever, and they are evolving faster than ever, too. “New technologies that used to have two-year cycle times now can become obsolete in six months, and the pace of change is not slowing,” says consulting firm Deloitte.

A growing multiplicity of technologies means more tech policy issues to cover. And those issues grow more complicated each year. As soon as lawmakers wrap their heads around one technology (if they do at all), another innovation pops up that complicates things further or crowds out their attention.

Technological convergence and blurring governance boundaries: Technology policymaking increasingly involves metaphysical questions about the underlying nature of things. For example, what is a “phone,” a “medical device,” or an “aerial vehicle”? These things used to be relatively easy to define and had well-understood meanings in federal statutes and regulations. But those concepts evolved rapidly in an age of widespread technological convergence and rapid-fire “combinatorial innovation,” with new technologies multiplying and building on top of one another in the symbiotic fashion. Basically, almost as soon as new tech laws or regulations are enacted, they are confronted with new marketplace realities and technological changes that call into question legal classifications or regulatory distinctions.

For example, today’s smartphones combine dozens of different functions that were previously quite distinct, including health tracking capabilities, mobile payment systems, and video distribution, all of which remain heavily regulated by an assortment of federal laws and agencies. But the convergence of all these capabilities in a single device that we can carry in our pockets creates massive governance challenges, not only for archaic legislative frameworks, but even for newer semantic distinctions that may seem current one moment only to be obliterated the next. These factors also make it harder to figure out who in Congress should be driving policy because technological convergence blurs previously distinct governance categories among legislative committees and the laws they have crafted.

Legislative dysfunctionalism: Policymaking processes move slowly by design. Constitutional constraints and other legal requirements demand it. But things move even slower today because of what Jonathan Rauch calls “demosclerosis,” or the “government’s progressive loss of the ability to adapt.” “[A]s layer is dropped upon layer,” he argued, “the accumulated mass becomes gradually less rational and less flexible.”

Inadequate resources are also part of the problem with Congress facing a complex, rapidly-evolving set of issues but devoting only limited resources to technical staff or studies to better understand these developments. This combined with the factors cited above has led to a never-ending “competency trap,” with lawmakers and their staffs seemingly always one step behind technological developments and societal demands or expectations.

Meanwhile, partisanship increases and the work load on many other fronts grows alongside it. There’s just a lot more on Congress’s plate than ever before. Plus, tech policy matters seemingly always take a back seat to tax, budget, entitlements, defense, and other issues.

Many people hope that boosting technology assessment efforts might help correct these problems. Perhaps better technical advice could help lawmakers ask less ignorant questions at tech-oriented congressional hearings, which have become showcases for the staggering lack of congressional understanding of modern technologies. But just adding new technology assessment capacity, such as in the form of a revived Office of Technology Assessment, won’t likely move the needle much in terms of actual legislative output. More serious structural reforms will be required.

Globalization: Many modern technologies “are truly global and call out for policy approaches that do not respect traditional national borders,” note former NITA officials Lawrence E. Strickling and Jonah Force Hill. Congress only has so much control over technologies that defy national boundaries, further complicating tech governance questions.

Yet, one would think that when America’s global competitive advantage was on the line, Congress would have greater reason to assert itself and craft frameworks to ensure US firms are not disadvantaged by a lack of policy clarity. That has not proven to be the case, however. Congressional lawmakers do plenty of huffing and puffing about the tech governance choices made by Europe, China, and other governments, but they then leave the field wide open to them (as well as lower levels of government) to craft policies that govern national markets throughout the United States.

Endless delegation: Speaking of passing the buck, Congress has been doing it for decades on tech policy by delegating massive and quite amorphous authority to technocratic administrative agencies. Over the past half century, scholars from various disciplines—economics, law, political science, history, and others—have explored the growth of what has been alternatively called the “interest group society,”  “receivership by regulation,”  “iron triangles,” and “client politics.” This literature identifies the way Congress has increasingly abdicated its constitutional role as lawmaker by shifting hard policy questions to regulatory agencies and then hoping that bureaucrats could figure out all the answers.

Delegation is even more common for the most technical policy matters, and that trend has only accelerated in recent years as the complexity increases and overwhelms lawmakers and their staff.

Ramifications for Tech Governance Going Forward

If Congress remains largely incapable of ever getting the ball over the goal line on important tech policy matters, what are some of the ramifications? There are many, but I will identify just a few of the most obvious ones:

• More tech-oriented legislative activity will shift to the states: In fact, it already has. For each of the tech policy issues I identified earlier (privacy, driverless cars, cybersecurity, and even some AI-related issues like facial recognition), states are—for better or worse—picking up the slack. We should expect that trend to accelerate. This will create an increasingly confusing patchwork of policies that will potentially raise serious barriers to entry and innovation. Nonetheless, I can’t see this trend reversing anytime soon. Perhaps Congress will finally act on privacy or driverless cars legislation if for no other reason than to preempt a crazy-quilt of contradictory policies. Of course, that’s what people have been predicting for years, and it never happens.
• “Soft law” becomes the dominate governance force for tech: Again, it already has. Soft law refers to informal, collaborative, and constantly evolving governance mechanisms that differ from hard law in that they lack the same degree of enforceability. Soft law can include things like multi-stakeholder processes, industry best practices and standards, agency workshops and guidance documents, and educational efforts. But that just scratches the surface of soft law mechanisms. For better or worse, soft law is becoming the dominant modus operandi for most modern technological governance. We can expect that trend to accelerate to fill the governance gap left by Congressional inaction. For example, we don’t have any formal “rules of the road” for driverless cars, but we do now have four iterations of Department of Transportation guidance on driverless cars. Version 4.0of the DoT guidance for automated vehicles was just released this month. Expect the “soft law-ization” of technological governance to expand considerably in coming years because it is really the only way for agencies to cope with the pacing problem and those metaphysical issues identified earlier. Because soft law is not boxed in by rigid preconceptions of what a particular technology or technological process is or entails, it is often better able to address new marketplace realities. Soft law can adapt as technologies do. With Congress out of the picture, it will have to.
• The congressional tech policy death spiral accelerates. Some may think (or at least hope) that the situation described here can’t get any worse. To the contrary, it can get radically worse. With our politics increasingly infected with bitter partisanship and rancor, what are the chances that lawmakers can work together to craft comprehensive tech policy measures? I’d say the odds are approaching zero. The Cable Act, the Telecom Act (and Sec. 230), and the Internet Tax Freedom Act all enjoyed broad, bipartisan support when they passed in the 1990s. People reached across the aisle to get things done. It didn’t always work, and sometimes it resulted in misguided policies (like the Communications Decency Act’s provisions trying to censor internet “indecency”). But bipartisan lawmaking scenarios like those seem almost unthinkable now. To the extent many lawmakers even show up at tech-oriented congressional hearings anymore, it is mostly to score points in front of the cameras for Team Red or Team Blue back home. Serious legislative oversight and policymaking is dead; it’s mostly just show-trials and media circuses at this point.

Should I Care about Congress Anymore?

If you believe this miserable thesis is correct but continue to focus on the Legislative Branch for a living, you may be asking yourself: Am I wasting all my time here? Not necessarily. Congress is still actively interested in tech policy matters. For those who hope to limit that damage Congress might do by hastily passing ham-handed, crisis-driven policy measures, your efforts in the trenches will continue to be important in curbing the worst instincts of some lawmakers. In many instances, preserving a perpetual stalemate may go down as a tremendous victory.

For example, as the debate over Section 230 intensifies—with politicians of all stripes looking to gut the most important of all Internet freedom policies—it is vital that smart people work with lawmakers and their staff to beat back misguided and destructive measures. Hopefully this becomes another instance of legislative gridlock winning out! And I think it will.

More realistically, your role will not be to stop Congress from doing insanely destructive things, it will be to just stop them from saying those things. In fact, that seems to be what a lot of people who work with Congress already do today. When I chat with various inside-the-Beltway policy advocates and industry reps today, they usually acknowledge that the prospects for actual legislation on any given issue are quite slim. They will, of course, continue to try to work with lawmakers, their committees, and their staff to either advance or stop legislative measures. Yet, they all seem to accept the utter futility of it all.

Why do they persist? Most obviously, they want to at least preserve the legislative stalemate and not cede the ground to their enemies who might succeed in getting lawmakers to do something if only one side was communicating with Congress.

But the other thing these policy advocates are hoping to achieve is better messaging. Regulatory advocates want lawmakers to use the power of the bully pulpit to put pressure on various people or groups to change behavior, even in the absence of any legislative action. By contrast, many in industry want to make sure that their technologies are understood and not endlessly demonized. Bad press isn’t good for business, even if all the congressional threats never result in final legislation. Also, those defending innovation more generally will want to make sure that even if lawmakers aren’t making any actual laws, they still better understand and appreciate the importance of new technological capabilities for improving human welfare.

Those are all good reasons not to give up your legislative advocacy. For some of us, however, the personal cost-benefit analysis just doesn’t add up. Our focus has shifted to where the real action is at: federal administrative agencies, statehouses and state administrative agencies, the courts, and the growing world of multi-stakeholder governance and other soft law efforts. Congress has checked out, but technological governance lives on in many other forms and venues.

Ken Ferree and I just filed an amicus brief with the D.C. Circuit in what could be among the most important First Amendment cases involving economic regulation in years:  Comcast’s challenge to the FCC’s cap on the maximum size of a cable operator’s nationwide subscriber-audience.  While few may feel righteous indignation at limitations targeted at large corporations such as Comcast or Time Warner, the larger principle at stake here is deeply important: Will the First Amendment provide a meaningful check on what USC law professor Chris Yoo has called “architectural censorship” (i.e., so-called “structural” regulations that “have the unintended consequence of reducing the quantity, quality, and diversity of media content”).

In a nutshell, we argue that that:

1. The provisions of the 1992 Cable Act authorizing the FCC to impose a “cable cap” are outdated in world of media abundance and vibrant platform competition.
2. Because cable is no longer the unique “bottleneck” or “gatekeeper” that it was in 1992, these statutory provisions (not just the FCC’s 30% rule) must be subject to strict scrutiny under the First Amendment as a limitation on free speech.
3. Because there are “less restrictive means” of ensuring cable operators do not impede the flow of video programming to consumers, the court should strike down these provisions.
4. Even if the court upholds the statute, it should nonetheless strike down the cap issued by the FCC in December 2007 (30% of all Multichannel Video Programming (MVPD)  subscribers as based on an outdated model of the video marketplace.

I encourage you to read our brief (below).  I’ve provided a summary below, along with some additional commentary we just couldn’t cover under our 3500 word limit.

Strict Scrutiny.  Yoo’s article Architectural Censorship and the FCC is essential reading for anyone who believes that government regulations on the size and shape of the “soapbox” can have huge effects on speech itself.   Yoo argues that the First Amendment should check this kind of regulation–however “content-neutral” it might seem–under “strict scrutiny”, which requires that the government show that a regulation is the “least restrictive means” available for advancing a “compelling government interest.”  But Yoo ultimately concludes (pp. 713-718, PDF pp. 45-50) that, under existing precedent, most “architectural censorship will be effectively insulated from meaningful judicial review.”  Yoo explains that the Supreme Court’s 1983 decision in Minneapolis Star & Tribune Co. v. Minnesota Commissioner of Revenue, “appeared to entertain the possibility of subjecting structural restrictions to strict scrutiny even in the absence of facial content discrimination or content-based motive.”  But in its 1991 Leathers v. Medlock decision, the Court “foreclose[d] any prospect that Minneapolis Star and its progeny would serve as a check on architectural censorship” by limiting the Minneapolis Star line of precedents to cases where “a statute of general application affects a small number of speakers.”  The Court reaffirmed this position in its 1994 Turner I decision, when it applied intermediate, rather than strict, scrutiny to the Cable Act’s “must-carry provisions,” which require nearly all cable operators to carry certain television broadcast signals.  Intermediate scrutiny requires only that important governmental interests that are furthered by “substantially related means.”

Unfortunate as the Leathers/Turner I line of cases is for those concerned about architectural censorship, the cable cap is exactly the sort of regulation that falls within the reduced scope of Minneapolis Star as “affect[ing] a small number of speakers” because, unlike the Cable Act’s must-carry provisions, the cap limits the speech of only the very largest cable operators.  So the question of whether the Court should default to intermediate scrutiny as it did in its 2000 Time Warner I decision (when the cap was first challenged) should turn entirely on the question of whether cable still has the “special characteristic” of “bottleneck” or “gateekeeper” power despite all the changes in the media marketplace since 1992 and even in just the last eight years.

The Modern Media Marketplace.  The subscriber limitation provisions of the Cable Act were intended to prevent cable operators from “unfairly impeding the flow of video programming.”  Yet each of the key premises behind these provisions has been disproven:

1. Increased horizontal concentration of the cable industry has, far from reducing media choices, been accompanied by an explosive growth in the amount and diversity of video content available to consumers.
2. The rate of “vertical integration” (i.e., ownership of cable programmers by cable operators), which Congress feared would cause cable operators to discriminate against unaffiliated programmers, has plummeted.
3. Cable’s share of the MVPD market has also plummeted dramatically, with the two DBS providers now sharing 1/3 of the MVPD market and representing the second and third largest MVPDs

Two charts say it all.  First, from Adam Thierer’s excellent book Media Metrics, the number of programming services (cable channels) has grown by nearly six-fold by 1992, while the rate of vertical integration has plummeted:

(That chart stops in 2006 (based on 2005 data) because the FCC still has not released the 2007 Video Competition Report, which it approved in December 2007.  Since then, Time Warner Cable has been spun off of Time Warner’s content empire, so the actual affiliation rate today is likely less than 10%.)

Second, cable’s share of the MVPD market has fallen from 95% in 1992 to ~64% today:

In 1992, when consumers had only a single MVPD option, cable might fairly have been considered a “bottleneck” or “gatekeeper.”  But today, every American has at least three MVPD choices (their local cable franchisee + two DBS operators), and can also subscribe to a Telco video service such as Verizon’s FiOS.  (“Over-building” where two cable operators serve the same area is rare.)

Internet Video.  We also describe how the availability of TV content online provides yet another distribution channel for programmers:

The last two years have seen growing numbers of Americans increasingly substituting consumption of online video for MVPD video and the Internet driving popularity of MVPD content, rather than vice versa.  But only in the last year, since the adoption of the [FCC’s December 2007 order issuing the 30% cap], has the large-scale delivery of television  content online become a reality, as large numbers of programmers have begun distributing increasing numbers of complete episodes and entire series through their own websites and/or through a new class of rapidly-growing Internet Video Programming Distributor (IVPD) websites such as Netflix, Hulu, Amazon Video on Demand, iTunes, Vuze, Sony Playstation Store, the Microsoft Xbox 360 Marketplace, Joost and Veoh.  These IVPDs already offer a staggering, and growing, library of currently-airing and archived content—as much as 90% of broadcast shows and 20% of cable shows.  These sites are supported by a growing number of set-top devices (e.g., Netflix Player by Roku, TiVo) and wildly popular game consoles (e.g., Microsoft Xbox 360, Sony PlayStation 3) that allow users to play IVPD content from broadcast and cable programmers on demand on their television, while TiVo allows users to seamlessly switch between IVPD, MVPD and OTA content.

The FCC’s decision to exclude Internet video from its analysis is hardly surprising when one considers that the economic model behind the new 30% cap comes from a 2005 study based on cable market data from 1984-2001 and that the last official data released by the agency about the video marketplace date to June 2005.  But nine months later, the agency waxed ecstatic about the promise of IVPDs when doing so supported Kevin Martin’s attempts to enforce the FCC’s non-binding 2005 “Net Neutrality” policy statement:

In August 2008, the FCC even cited [the rapid emergence of IVPDs] in support of its claim of jurisdiction over Comcast’s broadband network management practices (because of alleged harm to an IVPD that distributes content through peer-to-peer file sharing):  “consumers with [broadband] service will have available a source of video programming (much of it free) that could rapidly become an alternative to cable television.”  But the immediate competitive impact of IVPDs comes not from the fact that some IVPD users are already canceling their MVPD subscriptions, but in the ease with which IVPDs can supplement an MVPD subscription—because most IVPDs are free, while those that charge for content do so on a per-episode/show basis.  Furthermore, IVPDs have little—if any—incentive not to offer a particular program because they are not subject to the same capacity constraints as MVPDs.  Thus, even if IVPD video consumption remains relatively small in its early years, IVPDs already offer programmers a strong alternative distribution channel capable of reaching all broadband users.

Less Restrictive Means. Of course, the fact that cable no longer has a special characteristic of gateekeeper or bottleneck power does not automatically render the Cable Act’s subscriber limits provisions unconstitutional; this merely means that the government must show that no less restrictive means are available to satisfy a compelling government interest.  We suggest a variety less restrictive means that could ensure competitive video distribution and programming markets.  These include dispute resolution assisted by the FCC, enforcement of existing antitrust laws, and crafting “special obligations on cable operators with more than 30% of the MVPD market to ensure that they do not unfairly impede the flow of video programming.”

Challenging The FCC’s Rule. Besides attacking the statute, we argue that the 30% cap imposed by the FCC last year is even more obviously unconstitutional than when the D.C. Circuit struck down the same limit seven years ago in Time Warner II. To many lay observers, this argument may seem like a “no-brainer” given how much more competitive the video marketplace is than it was in 2001.  But one must understand that when the Court struck down the 30% cap the first time, it did so on the grounds that the FCC’s own rationale justified not a 30% cap but a 60% cap.  The FCC had decided that the average video programmer (network) needed an “open field” of 40% of the MVPD market to be viable.  The FCC leapt from that conclusion to a 30% cap so that even if the two largest cable companies denied carriage, the programmer would still have the required 40% “open field.”  The court found that there was no evidence that the leading two cable operators would collude to deny carriage and that the statute did not “protect programmers against the risk of completely independent rejections by two or more companies.”  In other words, the purpose of the statute was not to guarantee carriage even if, for example, a cable operator decided (exercising the same constitutionally-protected “editorial discretion” enjoyed by all media) spend part of its limited system capacity carrying a network with questionable appeal, or to raise subscription rates to cover the marginal cost of carrying the network.

But the FCC has since come up with a new “open field” model that the court must consider anew.  This time, the model more clearly supports a 30% cap–but only if one accepts the premises underlying the model and the accuracy of the data put into the model, which we do not.  We argue that their model is “based on flawed assumptions about the nature of competition for video programming” and is thus incapable of “accurately reflect[ing] cable’s present (or future) bottleneck power.”

Click the button at the top right of Scribd’s handy iPaper display to switch to full page display of the brief–or click on the top left to download the PDF itself.

PFF Amicus Brief – Cable Ownership Cap http://documents.scribd.com/ScribdViewer.swf?document_id=8630011&access_key=key-2obr4z2ohtozi1gabbay&page=1&version=1&viewMode=