An army of census takers has begun knocking on doors all across the country, following up on the 48 million households who didn’t send in very user-friendly forms in spite of a ubiquitous media campaign. There must a better, cheaper and even safer way to do this.
There is. People should be able to respond online.
In 2000, the government quietly experimented with a limited online option and it was very successful, as noted in a 2008 report by ITIF’s Daniel Castro. The online option was not publicized, but a few savvy citizens (.07 percent of respondents) discovered the link on the U.S. Census Bureau’s website and 94 percent of them said they liked it.
Ten years is several generations in Internet time. Less than five percent of Americans had broadband connections at home in 2000. Today, an estimated 68 percent take advantage of broadband access, and that percentage grows by six to eight percent each year. To be sure, there are many without access to computers and still others who would prefer not to respond online. Still, with so many more of us online daily and with the short form the Census Bureau used this year, it is reasonable to think we would have a high participation rate and could end up with a more accurate count. Canada, Australia, Singapore and Norway are using the Internet to collect data. We should, too.
Regrettably, back in 2008, the Census Bureau nixed the idea of an Internet-based data collection for 2010 mainly for three not very compelling reasons -– that it would not necessarily increase the response rate, that the entire undertaking could be wrecked by security problems and that it was too expensive. Given how much of life has migrated online, these objections seem a little flimsy in 2010.
First, no one claims using the Internet-based approach will boost overall participation. The Census Bureau quite correctly cited data from other countries that have adopted an online system. But it might help individuals respond more thoroughly and accurately. Besides, higher participation isn’t the only reason for using the Internet. It is easier and more conducive to the online habits of an increasing number of Americans.
Second, while security concerns such as denial of service attacks, “phishing” (where fraudulent sites pose as legitimate sites to extract sensitive information from people), or the use of spyware to steal data are legitimate, they are not sufficient to completely scrap an online count. For one thing, the risks from a denial of service attack are minimal in a non-time-sensitive project like the census — if the website is unavailable, respondents would simply check back later. The concerns about unscrupulous operators and hackers are part of life online and security experts have gotten pretty good at protecting consumers and citizens. ITIF estimates that 25 years after the first dot-com, the global economy is larger by $1.5 trillion a year thanks to the commercial Internet. The Internal Revenue Service says roughly 100 million of the 140 million tax returns filed this year were done so electronically. That’s a lot of personal data and money floating around and most of it is secure, so long as people are attentive to instructions and wary of possible fraud. There may be no such thing as a 100 percent secure system (on the Internet or in the real world) but that is no reason to put off going online with the census.
Third, but perhaps most important, is cost. The census is not cheap. The cost for the education effort, forms, workers and other activities adds up to $14.5 billion, according to the bureau. In 2008, there was a much-publicized and costly kerfuffle over the bureau’s problems with hand-held devices. The decision to scrap them rather than figure out how to use them effectively may have cost taxpayers $3 billion needlessly.
Aside from that snafu, the agency observed in 2008 that it would be hard to predict online usage and the number of paper forms that would be needed, and that there would be considerable startup costs for an online system. But an analysis by ITIF in February 2008 estimated that even with an online response rate of just 10 percent, the savings could have added up to $28 million — more than enough to cover the $22.5 million the Census Bureau estimated would be needed to implement an Internet-based system.
The bottom line is that there will always be challenges shifting an undertaking as massive as the decennial census from paper to electronic mode. But the Census Bureau, and all other government agencies for that matter, should be creating ways to overcome these challenges and not be fearful of being overcome by them. It is hard to think of anything that could be done more smoothly than the census with the smart use of information technology. That’s 20/20 hindsight for 2010 and foresight for 2020.