Virulent Word of Mouse

October 11, 2009

Can a government encourage economic experiments?

Can the US government adopt a policy that encourages innovation in the Internet? I think so. Recently, I got to say so in a public hearing at the Federal Communications Commission. As part of its Congressional assignment to write a broadband plan the FCC has held these hearings. The first half of this post summarizes my views about encouraging economic experiments in the broadband ecosystem, and the second half summarized my response to an interesting email from a listener about whether the government should fund experiments with municipal wifi.

A bit of background: I have been writing about innovation for a long time, but in the last few years I have become more interested in policy for innovative industries, such as the Internet.  Why write on this topic? Because these settings give rise to some challenging issues and policy questions for many well-meaning policy makers. I am hoping to be useful.


My comments focused on my recent paper, Glimmers and Signs of Innovative Health in the Commercial Internet. The paper focuses on the question, “What are the signs of healthy behavior in an innovative industry, such as the Internet?” This is the type of question that arises in almost every policy proposal for the Internet. It brings together many different strands of thinking into one essay.

More to the point, this essay provided the fodder for my two cents of policy prescription. I participated in the workshop on economic issues in broadband competitionfcc-logoSo far there have been 29 workshops between August 6th and October 9th held at the FCC. If you are curious, check out the group’s URL, (and it even includes blogs). The other participants on October 9th were economists too: Judith A. Chevalier from Yale; Joseph Farrell from UC Berkeley, presently as the FTC;  Marius Schwartz from Georgetown; and Carl Shapiro, from UC Berkeley, presently at the DOJ. If you are really curious, see this link to our session.

My testimony stressed the same issues found in my paper. We live in a world of vastly decentralized innovation. Innovation comes from many sources — established firms, users, university students, volunteers, entrepreneurs, local governments, and plenty of other unsavory actors. Government plays a role in this setting because regulated firms play a role, and government policy for regulated firms can help/hurt the entire ecosystem.

More to the point, it is possible for society to get the most benefit from so much innovative energy if policy goes beyond a simplistic approach. The policy must be tailored for its goals. That is not straightforward to do.

So how do policy makers know a market innovates well or poorly? What benchmarks should an observer use to judge market conduct? Here is the Cliff-notes version of my view. I suggest focusing on four types of behavior: economics experiments; vigorous standards competition; entrepreneurial invention; and the absence of unilateral bargaining.

A economic experiment describes market wide behavior oriented towards learning about economic value. Vigorous standards competition describes behavior oriented towards the rapid introduction of new standards to support new products and services. Entrepreneurial invention describes the range of participants (both large and small) involved in a market who undertake risky investments. The absence of unilateral bargaining refers to the lack of prevalence of powerful firms proffering take-it-or-leave-it deals to other firms.

These four criteria are arranged so that more is better, and less is worse (e.g., more experimentation beats less). If you want to know why I think more of these more of these four behaviors is better, then read the paper or listen to my presentation (which starts about minute 20 in the workshop and ends about minute 38).

In any event, I said my two cents and that was that.  It is a drop in the bucket of the vast record compiled at all these hearings. I hope the somebody finds it useful.

ShowCover.aspxI got home and found a very interesting email in my inbox from Stan Stringfellow. We have never met, but Stringfellow has a very memorable last name. I looked him up online. I found plenty of evidence on the Internet that Stan is a sophisticated thinker (he wrote a book about Sun Servers, for example). He seemed worthy of a serious conversation, so I looked at this proposal.

Stringfellow’s proposal concerns the next generation of experiments in municipal wifi, which he wrote about in his blog.  He intends to forward the proposal to the FCC in response to their solicitation for ideas, known as a Notice-of-Inquiry in lawyer-speak. I will give only a summary here. You can read the full version on his blog.

Stringfellow proposes to have the government use some of its stimulus money to foster further experiments in municipal wifi. He offers many reasons. Three struck a chord with me as potentially interesting.

First, while a quite a few small scale muni wifi networks have been built, Earthlink and others never fully finished building the large scale examples of municipal wifi networks. Society could learn from a large scale implementation and daily operation; Second, there have been recent upgrade to the design of wifi, enabling it to support higher bandwidths (in engineer-speak from 802.11b/g to 802.11n). This gives these networks enough functionality to be more successful where they had not in the past; Third, the timing of prior buildouts was too eager and too soon. Now the timing is right for a big-scale implementation from which others could learn.Muni Wifi Router

Stringfellow asks me: is this the sort of economic experiment I am hoping the government will try to encourage? Sometimes specific proposals can illustrates new questions and highlight interesting angles. This one does, so let me comment on it here.

First, a bit of truth in advertising. I am not making an endorsement. I will not endorse any proposal, Stringfellow’s or anyone else’s. It is not about the merits; it is just not the sort of thing I should do. It is not my place to assess every proposal, nor would I want to. It is, however, my place to try to observe and comment, and this proposal fosters that goal. Got it?

Second, my paper articulates a set of norms for understanding how innovative a market is or is not. Stringfellow proposes a specific experiment. Those are different. I want to consider the proposal because it illustrates something about economic experiments. Ok?

Third, I want to do this even though I am somewhat skeptical the proposal will work or will be viable. I am intrigued by it for a simple reason: because because nobody in Washington has thought of it. More to the point, the last decade’s experience with Wifi illustrates a key lesson. There is value of letting users, suppliers, and just about anyone with a idea experiment with new applications in wireless.

Fourth, this proposal is consistent with one of the lessons of wifi. That is, if market participants have the ability to decide how to employ the spectrum, they will come to a different conclusion about how to deploy it than some bureaucrat in DC who negotiated the precise application with other politicians and other agencies in the federal government. In wifi’s case, the spectrum it uses was originally designated for short range uses, such as garage door openers and wireless handset inside houses, but some hard work by some foresighted thinkers in the FCC made sure the spectrum did not have tight restrictions on its use. The allocation was regarded by traditionalists as “garbage spectrum”, a throw-away. To be sure, Wifi was not the anticipate use.

Fifth, there are many ways to interpret the success of Wifi, but I prefer to interpret in light of the US spectrum system, which allocates spectrum for very circumscribed purposes.  Circumscribed allocations truncate economic experimentation, limiting application development before it ever gets started.  Stringfellow’s proposal is a good example of something society would never get without letting others conduct economic experiments with spectrum.

Sixth, and perhaps more concretely, I think wifi technology and its descendants still have a long way to go. Yes, I think wifi will continue to get better. It is a very cost effective way to accomplish many things, and, with the recent increases in bandwidth, it is way more cost effective than lots of other proposals for doing all sorts of activity.

Seventh, I am not sure which implementation will work. Like Stringfellow, I am pretty sure something will work eventually, and at some point we will find out. Indeed, like Stringfellow, I am also convinced that timing will play a big role in finding out. However, I am more agnostic than he is about whether the timing is right just now. Stringfellow might be right or he might not be. Beats me.

Finally, Stringfellow would like the government to fund experiments, but I am not so sure about that. There are lots of levers government can use to increase experiments (such as allocating more spectrum to public use). More to the point, I am not convinced it should play any role as a provider of funds for an experiment of this magnitude, as if the FCC takes on a VC role. Partly this is because VCs do a lot more than fund start-ups. VCs monitor their investments, give managerial advise, and lubricate new solutions to many of the hurdles that young firms face. The best VCs are really good at this because it is their secret sauce. I am quite skeptical that anyone in a government job could match the best VCs in such activity.

In addition, there is something ironic about government bureaucrats choosing to fund a project in a setting where the applications’ principal strength is the discretion it gives users of spectrum other than the government. Maybe I am old fashioned, but it makes me uncomfortable.

So there you have it. More economic experiments in the broadband ecosystem would be a good thing for society. That broad goal implies a set of principles for understanding whether a market’s participants have behaved in ways fostering innovation or not. It would be useful to use such guidelines in policy instead of relying on ad hoc assessments about whether a market is innovative or not.


  1. I am thrilled to have someone of Shane Greenstein’s stature critique my ideas, and I want to thank him for doing it.

    I guess I’m not surprised that no one in Washington has thought of the approach I am advocating. It seems quite unusual, even to me. But, I was really struggling to conceive of a solution to a problem that I see in the wireless broadband ecosystem in the USA – a problem that no one appears to have an adequate answer for. Let me put it in a very specific and concrete way:

    What if Earthlink tried a major experiment with municipal Wi-Fi, and the experiment failed. What if there were reasonable explanations for that failure – such as issues with market timing, business/technology ecosystem immaturity, lack of funds to continue the experiment, and more – but that none of this mattered, because no private sector company was left that was willing or able to continue the experiment on a large enough scale to develop the necessary wide-spread momentum to make the industry profitable. Does all this imply that municipal Wi-Fi is a bad idea?

    This is all the more vexing if one is convinced for a variety of reasons that increasing the competitive stance of large-scale Wi-Fi hotzones and municipal Wi-Fi within the overall “coopetive” framework it shares with other approaches such as 3G and 4G/WiMAX can go a long way toward addressing the national broadband goals set forth by Congress.

    I’d like to respond to a few of Shane Greenstein’s comments:

    “Stringfellow would like the government to fund experiments, but I am not so sure about that. There are lots of levers government can use to increase experiments (such as allocating more spectrum to public use). More to the point, I am not convinced it should play any role as a provider of funds for an experiment of this magnitude, as if the FCC takes on a VC role. Partly this is because VCs do a lot more than fund start-ups. VCs monitor their investments, give managerial advise, and lubricate new solutions to many of the hurdles that young firms face. The best VCs are really good at this because it is their secret sauce. I am quite skeptical that anyone in a government job could match the best VCs in such activity.”

    I very much like the idea of allocating more spectrum for public use, such as “White-Fi”, which would use the “whitespace” spectral bands between the new digital wireless TV channels. I find, in the particular case of White-Fi, that people usually talk about fairly targeted applications, such as providing broadband access for rural areas or filling in the gaps in city-wide (regular) Wi-Fi networks; This is due to certain limitations of the technology, such as speed and complexity. This is still great, of course. And, perhaps other spectrum might be freed up that wouldn’t tend to be viewed in this somewhat limited fashion. But, in any case, I wonder how powerful a solution this represents, in terms of really creating what has been called a Third Pipe: a truly competitive and wide-spread wireless broadband offering. Plus, how long it will take for such an offering to be implemented and widely adopted? Don’t forget, consumers would need to buy millions of new devices (or perhaps add-in cards) that support the new wireless technologies. Julius Genachowski recently said that by 2013 U.S. consumers will use nearly 400 petabytes per month of wireless data compared with 6 petabytes per month in 2008. If true, that is a fairly near term explosion in bandwidth requirements. Large-scale Wi-Fi implementations, if wide-spread, could go a long way toward satisfying this type of need. The Wi-Fi ecosystem is here now, and there are forces that will keep it growing, as I mention in my blog article. Wi-Fi is also the common denominator among all types of wireless devices, since everything from smart phones to laptops tends to be Wi-Fi enabled.

    Well, this leads naturally to the most controversial idea in my article, what I call the “über venture capitalist” approach, where the Federal government would INVEST – rather than grant or loan – a portion of the $7.2B that Congress has allocated, and seek a return on investment. My suggestion is that the weightiness of the FCC and Congress could give the necessary momentum to the large-scale Wi-Fi industry, and encourage private investment, while it ultimately nurtures this industry to the point of self-sustaining profitability and rapid growth. I certainly can understand skepticism about this idea. But, I hope to encourage people to let it simmer a bit, and mull it over. Professor Greenstein says “I am quite skeptical that anyone in a government job could match the best VCs in such activity.” I absolutely 100% agree!! That’s not what I’m trying to say. I am suggesting that joint venture public-private partnerships could be formed, where the private sector partners are primarily responsible for the business analysis and due diligence that precedes a funding cycle. The Federal government would provide broad vision in support of a “startup industry”, and would make strategic investments. But, this effort would be done in partnership with private companies that have a vested interest in the outcomes of specific investments, and who are primarily responsible for them. I honestly don’t know how difficult it might be to work out this type of a shared-responsibility investor model. But, at least understand, I mean for the private sector to do what it does best, and for the public sector to do what it does best.

    My full blog article, once again, is at:

    Thanks again to Shane Greenstein for taking the time to comment on my ideas!

    Comment by Stan Stringfellow — October 12, 2009 @ 7:07 am | Reply

  2. [in passing] On the other hand, local moderate-bandwidth, moderate-latency wireless service could be treated as a public service like roads. It costs a lot less than roads, after all.

    Comment by The Raven — October 12, 2009 @ 8:15 am | Reply

  3. I referenced this blog entry on the MuniWireless LinkedIn group in the discussion Ken Bilba, Managing Director at Novarum, added the following interesting comments:

    I think experimentation here is very important. The changes in the last few years in muni WiFi are really quite profound … and ultimately change the ground of the discussion.

    1. The dramatic shift to small mobile devices (iPhone is the archtype- but we will likely see more) that all have WiFi … and are increasingly bandwidth hungry.

    2. The clear inadequacy of the classic telcom wireless networks to meet the demand. In looking at the Cisco statistics on network usage (an iPhone user generates 30x the network traffic of a classic smartphone), matched with the clear overload of the primary example of their use (the ATT HSDPA network) – a bit of mathematics shows that the overwhelming move to this new generation of device completely swamps not only ATT, but every other wireless carrier in short periods of time. As the Novarum study with PC Magazine demonstrated, ATT is poor today … Android phone success with Verizon will swamp it as well next year.

    3. The new generation of WiFi based on .11n uses the same technology TODAY (MIMO, beamforming) that WiMax and LTE will deploy in several years and clearly is on a faster product track. Novarum measurements of real networks clearly demonstrate that .11n materially changes the performance in outdoor environments as well as indoor environments – coverage, performance AND capacity.

    4. WiFi rides the Moore’s Law pricing/performance curve … Novarum’s analysis suggests that muni WiFi using .11n technology can deliver MORE capacity and performance than either WiMax or LTE – at 20% of the CapEx.

    These are all disruptive issues … it is time to take advantage of them.

    Comment by Stan Stringfellow — October 22, 2009 @ 3:28 pm | Reply

  4. government jobs are still the best when it comes to job security :-;

    Comment by Network Switch : — October 28, 2010 @ 5:53 pm | Reply

  5. Well, six years later, there is evidence that this is actually happening, but on the level of city government. And this is not just moderate bandwidth access. New York is creating a world class “cloud of gigabit-speed Wi-Fi access” to help the city compete globally. Today, one wonders if this might play into Smart City IoT initiatives. …and if this might evolve from many hotspots to become true ubiquitous access in the future?

    Comment by Stan Stringfellow — August 13, 2015 @ 10:04 pm | Reply

RSS feed for comments on this post.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

Blog at

%d bloggers like this: