Last summer I had the chance to help different companies apply for grants under the first round of the Broadband Stimulus program. One of my clients was The Knight Center of Digital Excellence (http://www.knightfoundation.org/home/), they contracted with me to map demographic data in various cities where they were applying for funding under the adoption portion of the stimulus program. My job was to create static map images by census block group showing data points such as age, sex, education level, income and internet adoption rates. The raw data covered various cities for the project. One of these was Philadelphia, a city of which I have an interesting level of experience. I worked for EarthLink as a Chief Wireless Engineer for their Municipal Wi-Fi projects and Philadelphia was my responsibility. I had overall charge for the wireless design and implementation, supervising the junior engineers and approving their work. This gave me a unique perspective on the data that I was mapping for the Knight Center. As I created these maps, I noticed something strange when looking at the levels of broadband adoption. The Knight Center wanted me to clearly show the areas of the city that had adoption rates below 40% of the households. As I placed the raw data in map form, a very alarming pattern showed up. The less than 40% rate census block groups were in exactly the same places as the live EarthLink network! I am but one of a handful of people who could have ever noticed this. The odds of someone mapping an adoption rate for this city who would have also known the actual deployment area of the EarthLink live network are very slim.
If you examine the areas of red and orange on the map they correlate almost exactly to the areas of the city that EarthLink had the actual live Wi-Fi network. What makes this even more interesting is that when EarthLink sold the system to NAC (Network Acquisition Company) in June of 2008, the network was left wide open with no restrictions for access. In many locations you could pull a 6 meg connection over Wi-Fi. I had spoken to NAC after they pruchased the system as they were discussing the option of hiring me as a consultant. They admitted they had not done anything to secure the network from free use nor limit the connectivity speeds. So the very next year, the reported broadband adoption rate in these areas where there was free Wi-Fi high speed internet access, was still below 40%!
This raises a couple of key questions. The first one I considered, was to how the data gathering was conducted. Did they survey people and ask them if they subscribed (meaning paid) to any high speed internet service? If that was indeed the question, I can see how this might show low adoption. These people did not have to subscribe to high speed internet service because they could get it for free. The second question to consider is, if you have high speed internet available FOR FREE, without the need for any special equipment, why aren’t more people on the internet?
Now these areas that did get built by EarthLink were high on the priority list for Wireless Philadelphia. They were identified early on as low areas of broadband internet adoption. So after almost 2 years of access to free internet service, why is the adoption rate still so low? One must consider the possibility that it is not lack of access to the technology, but some other factor(s). Obviously cost of the service was not a issue, it was free. These neighborhoods are not out in rural America where you cannot get high speed internet (outside of satellite) no matter what. They have access to cable internet and DSL along with cellular data networks. There are even other Wireless Internet Service Providers (WISP’s) offering service in these same neighborhoods. We had to deal with them and EarthLink went to Federal Court with one over interference issues. There may very well be other issues to adoption. Do these people just not want to be on line? Can they not afford a computer? Are they of the age where they are intimidated by computers and therefore won’t try to use them? Is it an education or literacy related issue? Is there no clear cut answer?
These are key questions that need to be answered. The stimulus program laid out a large amount of money for projects that were specifically for adoption in cities and areas that already had access to broadband services. Many people are still guessing at these issues and will use those monies as an experiment to find out why. Personally I think they should have looked at this specific situation first before going after all these large scale adoption projects in large cities. Philadelphia offers a good place to start looking as you could not get a better experiment to see what would happen. Heck they gave it away and still we have low adoption. Lets put the efforts in to looking at this area first. Once the answers are known, money can be better spent on adoption.
These issues and answers are very important especially as the FCC is considering changing the rules for the Universal Service Fund (USF). They are working on opening up the fund for broadband internet services and part of that is being considered for adoption. If we don’t understand why there are low adoption rates in areas with access, then why are we inviting rule changes that don’t offer proper solutions and will just become more waste for those funds?
My next blog – “Where should USF be allocated when considering broadband service?” I have hard and fast data that can determine this today. I know the threshold private industry will support without subsidies based on household densities and can break it down state by state.