Benchmarks and the Broadband Ecology.

As folks in broadband policy land know, thanks to the American Recovery and Reinvestment Act (ARRA), the FCC needs to present a National Broadband Plan to Congress by February 2010. The FCC has been holdng a bunch of workshops on various aspects of the plan as part of the inputs. I spoke last week at the Benchmarks Workshop. As I remarked to the sparse crowd, you can tell this is the deep uber-wonk stuff that only a handful of us find terribly interesting and vitally important while the rest of the world zones out and watches videos in their minds. For the short version, I include my latest “Five Minutes With Harold Feld.” Those more interested can watch the entire panel or read my written statement here. My more snarky comments (including my assessment of the panel) below.

More below . . .

What most folks in policy land may not know is that this plan goes far beyond the usual debates about producers, consumers, etc. etc. Yes, all that stuff is important, but, as I first wrote back in February when ARRA passed, the Act forces us to move from the limited view of a broadband market to a broader view of a “broadband ecology.” As I explained in my statement to the FCC:

This conceptualization moves beyond the relatively simple visualization of buyer, seller, producer and recognizes that the ARRA requires the National Broadband Plan to consider the entire universe of stakeholders for the purpose of promoting beneficial impact. As in a natural ecology, impacts in one part of the system influence other parts of the system both directly and indirectly. And, as in an ecology, if we allow a single species to dominate it will naturally try to reshape the ecology to its own advantage.

Happily, that isn’t just me being all touchy-feely. The portion of the statute that requires the FCC to create a National Broadband Plan to detail how it will advance: “consumer welfare, civic participation, public safety and homeland security, community development, health care delivery, energy independence and efficiency, education, worker training, private sector investment, entrepreneurial activity, job creation and economic growth, and other national purposes.” That’s very broad, which gives people a serious incentive to panic and retreat to what they know and define things as narrowly as possible.

The problem with that is it means we will fail. Big time. Epic fail. Because hiding under the mattress because reality is too gosh-darn scary does not make reality go away. As we have spent the last dozen or so years in policyland proving, the effort to reduce policy to a set of pretty equations and reduce metrics to what is easily quantifiable and collectible produces a picture so out of line with reality that it eventually turns on us and bites us in the rear. (Paul Krugman had an good piece on this in academic economics world, and the same logic applies in policy.) Which will bring us back to declaring victory based on metrics that do nothing but help us feel good about ourselves and persuade us we live in the best of all possible worlds.

But it’s worse, because there is a perennial incentive by incumbents to try to push the metrics toward purely economic indicators that have less to do with actual reality and more to do with predetermined outcomes. And after two decades of being hammered by the D.C. Cir., Congress, and various administrations, staff are conditioned to agree with these metrics and assessments. They are also very appealing from an academic standpoint, and from an administrative standpoint — most of them are easy to measure.

An illustration from my panel really brought this home to me. I noted that the statute requires the FCC to include in the plan how it our policy will make broadband affordable for all. So I was asked what the benchmark should be for affordability. I noted that the Department of Housing and Urban Development (HUD) defines affordable in the housing context as a percentage of income paid for housing. This presents some problems for broadband because we don’t collect personal subscription or personal income data. But we can try to develop what percentage of income a person should pay for broadband if we treated broadband as a utility service. We can then look at median income in the census block, look at the median cost of broadband in the census block, and use this as a rough estimate for access to affordable broadband.

Not so fast, responded my fellow panelist, Dr. Richard Clark, Assistant Vice President for Public Policy at AT&T. That doesn’t get you the real value of broadband or its real cost. There are many other factors involved, such as what applications are used/desired, what else is included in the bundle, speed, and so forth. Really, to get the value, you need to develop an index that looks at all these factors and that seperates out the price of broadband from other services.

Ask for a response, I noted that the statute speaks of “affordability.” At the end of the month, someone writes a check for the service. The number on the check is what they pay. If they can only get broadband by also buying other services, such as voice and video, it does not matter that the “price” of the broadband service is lower than what they have to pay t get it, or that the “value” is higher because it comes with additional services they must pay for. If the person cannot afford to write a check of the size required to get the access, then it is not affordable no matter how we might, for other purposes, compute the value or price.

Mind you, I am just as guilty of lapsing into my own frame of reference as anyone else. For example, the first place I go to find out what we need in the broadband plan is the statute. Because I’m a lawyer and I come at this from “what Congress tell the FCC to do.” So I’m not faulting my fellow panelists, or even the FCC analysts, for being who they are and not being me instead.

Rather, as I mentioned in an aside, what we really need to see are more people from other disciplines on both the inside and the out. I’d start with some pure informatix folks, because the first step is figuring out how deal with all the information flowing in and figure out what information you actually need. The next step is for all of us in the process to recognize that we will find wisdom in a number of specialties outside our own. this is much bigger than a legal problem, or an economics problem, or an engineering problem — although all these disciplines must inform the development of the plan and its benchmarks.

Which raises a final point that I made in my testimony but feel the need to make again. Benchmarks are not goals. Benchmarks are measures to tell us whether we are on track to meet our goals. So we may want to look at a lot of different things to measure success — especially as we start to see relationships between things like the relationship between broadband in classrooms and educational outcomes and other benefits. We should also expect to revise benchmarks as we come to understand these complex relationships. What keeps us from redefining benchmarks to change failure into success is that the goals remain both measurable and fixed. The map is not the territory, and the benchmarks are not the goals.

Stay tuned . . . .

This entry was posted in Tales of the Sausage Factory, The Stimulus Package (ARRA) and tagged , , , , . Bookmark the permalink. Both comments and trackbacks are currently closed.


  1. curly says:

    re: the relationship between broadband in classrooms and educational outcomes

    There’s a lot at that link, and searching on the word broadband didn’t give anything. Is there a specific link you can provide?

  2. Harold says:

    And will fix link, thanks.

  • Connect With Us

    Follow Wetmachine on Twitter!


If you do not have an account: Register