Visit Your Local PBS Station PBS Home PBS Home Programs A-Z TV Schedules Watch Video Donate Shop PBS Search PBS
I, Cringely - The Survival of the Nerdiest with Robert X. Cringely
Search I,Cringely:

The Pulpit
The Pulpit

<< [ The Ears Have It ]   |  The Diddy Factor  |   [ Come to Daddy ] >>

Weekly Column

The Diddy Factor: Why We Probably Shouldn't Put Too Much Faith in Presidential Tracking Polls

Status: [CLOSED]
By Robert X. Cringely

Last week, the U.S. Federal Communications Commission (FCC) approved commercial broadband Internet service over electric power lines. A couple years ago, I wrote a whole column about this topic for Inc., (it's among this week's links) and not much has changed from a technical standpoint, but the news is so important and that column was so long ago that I think a few points bear repeating, the most significant being that broadband over power lines (BPL) is going to totally shake up the Internet industry. There's a new sheriff in town.

You might wonder why BPL even matters since we already have cable modems and DSL, and in some places, wireless broadband? It matters because none of those is reliable enough or cheap enough, and I blame that on inadequate competition. Heck, I have BOTH DSL and a cable modem, and there are times when my service has still faltered. The entry of a major new player can only help to improve the quality of service from incumbent providers. At least that's my hope.

What kept this FCC approval from happening earlier were objections from the Amateur Radio Relay League (ARRL), which had vehemently opposed BPL because of possible interference with amateur radio communications. While the ARRL continues to have some reservations about the technology, its own take on this week's vote was generally positive. And the major BPL vendors, too, seem to be taking a conciliatory stance: "Current (Communications -- the only company now selling commercial BPL service) uses HomePlug on the LV (low voltage -- inside your home) side, which 'notches out' the ham radio bands, and Current uses the Medium Voltage (MV) wires with a proprietary approach. The MV frequencies which Current use do NOT overlap with ham radio bands at all (Current does not operate in the Amateurs band on the MV), so the fact that Current use the MV wires will have no impact on Amateur radio. Current has worked hard -- and continues to do so -- to make our system a good spectral citizen," wrote Jim Mollenkopf, an engineer at Current Communications.

One thing to remember about electric utilities is that they are very slow and deliberate. They move like glaciers, so it will take awhile for these services to be available at your house. But like glaciers, they are also impossible to stop.

The appeal here to an electric company isn't that $20-30 per month they'll charge for becoming your ISP. What matters to them and what makes this whole thing so important is that it will lead to your electric meter being monitored 24/7. That means utilities can start to offer true dynamic pricing, with electric costs dropping in low demand time periods and dramatically rising with high demand. While that sounds bad, the end result is actually good, since for the most part, profits from electricity sales will be regulated. The real end result is that demand will be better controlled by dynamic pricing, and the utility may just be able to forego building that $2 billion power plant they've been planning and saving for over the past 20 years. Dropping $2 billion to the bottom line has to appeal to any board of directors and, in a tightly regulated environment, will probably lead to overall power rates going DOWN, not up.

So this BPL stuff is mainly about getting smart electric meters and only partly about offering Internet service. But having made the effort to build the network, offer it they will, generally through unregulated subsidiaries.

Expect better service all around, and downward pressure on prices, all of which is good and all of which will affect you whether you ever use BPL or not.

My other topic this week is what I'm calling the Diddy Factor, named for hip-hop entrepreneur P. Diddy, who is headlining an ad campaign right now on cable TV urging young American music fans to register and vote in the coming Presidential election. I think the impact of this get-out-the-vote campaign is being underestimated. A communication revolution is taking place right now, and simultaneously, creating confounding influences on presidential tracking polls. All the polls right now say close to the same thing, that the race for President is neck and neck, but I'm here to predict that it is not. I'm sticking my neck out a bit, but I'm pretty confident that the polls are wrong. What's most interesting is why we aren't being told that.

The simple story is that nearly every tracking poll is conducted by telephone. A pollster selects a random sample of 500 or so voters, calls them up on the telephone, and asks who the respondent would vote for if the election were being held today.

One confounding influence in all these polls is the fact that people without phones aren't counted at all. Fortunately, for the last 50 years, telephone penetration has been 95-plus percent in this country, so that hasn't been much of a problem...until now.

Just down the street from my place here in Charleston, SC, is a house filled with women students from the College of Charleston. I'm not exactly sure how many girls are living in that house, or whether their boyfriends should be counted or not, but I do know for sure that the house, which has somewhere between five and 10 occupants, doesn't have a telephone. Every kid has a mobile phone they use for everything. And that means no tracking poll is ever going to learn who those girls and boys are going to vote for President because it is literally against the law for pollsters to call mobile phones in this country.

Yeah, but how much of an impact could that have? More than most of us expect, and a LOT more than most pollsters expect.

Anna Greenberg of pollster Greenberg Quinlan Rosner Research told the BBC, or example, that only three percent of Americans use their mobile phones as a sole communication device, but the FCC said two years ago that five percent of U.S. homes have only mobile phone service and that 15 percent of university students have only mobile phone service. And with 77 million U.S. mobile phones owned by people age 18-24, many of those supposedly counted are probably still associated with a parent's hard-wired telephone number but are really mobile. So the numbers of unpolled votes could be huge.

And though pollsters (who after all are generally in business to do this work) deny it, the switch from fixed to mobile communication is already having an impact on the outcome of elections.

In the last presidential election, one might have expected the final tracking polls to pretty closely reflect the actual outcome of the election only a few hours later. But no. Gore was generally two to three points down in most tracking polls conducted on November 6, 2000, but won the popular vote on November 7 by about half a million voters, or half of one percent. True, this is within the statistical range of most polls, but if the deviation from the actual vote count was truly random noise, then half of the tracking polls would have counted high and half counted low. But that's not the way it happened, and the reason isn't noise, but a consistent sampling error.

More recently in the 2003, Philadelphia mayoral election the final tracking polls gave incumbent mayor John Street a slight statistical lead over challenger Sam Katz, yet the actual vote went 59 to 41 for Street. How could those Philadelphia tracking polls be so far off? They missed the extensive effort to register student voters in that city, with its several major universities.

Now how about Diddy and all the others urging young people to register and vote in the upcoming Presidential election? Their stated goal is 20 million new voters (out of a total of perhaps 110-120 million voters) and given the fervent message and extensive advertising on MTV, VH1 and elsewhere, that goal just might be reached, presumably with most of those kids voting for Kerry, the Democratic challenger. If the polls are skewed, then Kerry is actually doing much better and can probably expect a comfortable win.

But if that's the case, why aren't we hearing about it?

The likely answer is simply because Democratic strategists fear any sign of cockiness will result in many of those newly registered young voters not bothering to vote at all, leading to a Bush victory. So nobody says anything, holding their breath and hoping for a particular outcome.

And Diddy, I hear he's planning to sublet the Lincoln bedroom.

Comments from the Tribe

Status: [CLOSED] read all comments (0)