An Internet Census and the Digital Divide

by Terri Ann Lowenthal

In my last post, I gave a shout-out to my father, who I fear could be overlooked by a largely electronic census, given dad’s likely nonagenarian status in 2020. Older Americans uncomfortable with today’s gadgets are just the tip of the iceberg when it comes to population groups that the Census Bureau might have difficulty reaching through the Internet.

It does seem like everyone is walking around with a smartphone glued to their ear, or reading their news or the latest Stephen King novel on a tablet. But the hard facts — gleaned from a Census Bureau survey on Internet usage — tell a different story.

In Exploring the Digital Nation: Home Broadband Internet Adoption in the United States, the Commerce Department reported that more than three-fourths (77 percent) of U.S. households own a computer, be it handheld or sitting on a desk or lap. But computer ownership and broadband adoption are not spread evenly across household income levels, race and ethnicity, age, level of education, disability status, and geographic location.

Consider a few of the reports specific findings:

  • Seventy-three percent of urban (metropolitan area) households use the Internet, compared to 62 percent in rural (non-metropolitan area) households. Seventy percent of urban households have broadband access; 57 percent of rural households do.
  • More than four-fifths of Asian households and roughly three-quarters of non-Hispanic White households use the Internet. Less than 60 percent of Black, American Indian and Alaska Native, and Hispanic households can access the Internet at home.
  • Forty-six percent of households with incomes below $25,000 have home Internet access, compared to 84 percent of households in the $50,000 – $75,000 income bracket. There also are significant broadband adoption differences by household income: Nearly 90 percent of households in the $75,000 – $100,000 income range access the Internet using broadband; only 43 percent of households in under-$25,000 group do.
  • Less than half of household heads with a disability use the Internet, compared to three-quarters of those without a disability.

A more fine-grained analysis of the data revealed greater variability by socio-economic characteristic; the department reported, for example, that less than 30 percent of Black rural homes whose head of household lacked a high school diploma use a computer. Commerce Under Secretary (and Deputy Secretary-designate) Rebecca Blank told reporters at a press briefing (11/8/11) that the large gaps in access to broadband and Internet use were “striking and not something we expected to see.”

For census apostles, the most worrisome aspect of the disparate access to computers and reliable Internet is that, to a significant degree, many population groups lagging behind technologically are historically harder to count in the census and prone to disproportionate undercounts. Furthermore, a quarter of households without Internet access cite affordability as a major barrier to this service. Current economic trends do not favor better financial circumstances for lower income households.

Earlier this fall, I mused about the lightening pace of technological change, which will present significant challenges for the Census Bureau as it designs a less-costly enumeration for 2020 over an eight-year span. No doubt, access to computers and the Internet will increase across all demographic subgroups with time. But as new technologies emerge, differential access to those tools is likely to persist.

All of these factors pose significant challenges for the Census Bureau, as it tries to balance the obvious advantages technology offers for ease of participation, operational efficiency and cost containment, with the need to count people who cannot or will not respond electronically. (To complicate the census planning process, U.S. Postal Service budget woes might slow the delivery of first-class mail across the country; Saturday delivery might also be a historical footnote by decade’s end.)

So, yes, I agree with Census Director Robert Groves that 2020 must be a “multi-mode census. … We must move beyond the mailback questionnaire and the personal interview … to ensure that the response options for the census reflect the communication platforms that people are using.” (Testimony before the Senate Subcommittee on Federal Financial Management, Government Information, Federal Services, and International Security, April 6, 2011.) But some of those modes will be more costly and traditional than others, and Congress must be mindful of the digital divide as it decides how much money to spend on planning and execution of the next decennial count.

 # # # 

From the Census Project family to yours:
Happy holidays and best wishes for 2012!
(REMEMBER: The next census is only eight years away!)

Raiding The Census Piggy Bank

by Terri Ann Lowenthal

With the smell of turkey and sweet potato pie in the air, Congress finally approved funding for the U.S. Census Bureau for the fiscal year that started seven weeks earlier. The so-called “mini-bus” appropriations bill — encompassing three of 12 federal appropriations accounts — allocates $943 million for the nation’s largest number-crunching agency (H. Rpt. 112-284).

Well, sort of. The bureau actually will receive $888 million in direct appropriations. Congress decided to dip into the little-known Working Capital Fund (WCF) for the remaining $55 million the Census Bureau needs to pull off the 2012 Economic Census, albeit a scaled-down version. More on that in a moment.

Not familiar with the WCF? For starters, it’s not really a fund. Rather, it’s a revolving account that is used to manage many of the Census Bureau’s core functions. Half of the account represents money from other federal agencies for reimbursable work, such as surveys. In other words, it’s not the Census Bureau’s money. The other half pays for what can loosely be termed “overhead” — that is, basic but essential operations that support all programs. Things like IT systems; the budget, human resources and communications offices; and salaries for the director and other managerial staff.

Appropriators decided that the Census Bureau could spare $55 million from this pot of money, so they wouldn’t have to find more discretionary funding to pay for essential census and survey activities. Last year, Congress permanently torpedoed $50 million of the WCF and pretended it had reduced federal spending by that much. Does anyone else detect a pattern here?

I worked in Congress for 14 years. It is with utmost respect for those who toil in legislative obscurity that I say, “People, the Working Capital Fund is not an appropriator’s piggy bank.” Yes, I am aware of the new Government Accountability Office report (GAO-12-56) suggesting that the Census Bureau allow more sun to shine on the WCF and establish operational performance measures to promote efficiencies. The congressional auditors also noted that dramatic fluctuations in spending on the decennial census require the bureau to save money in the WCF for a rainy day through an operating reserve. Which is now $50 million smaller.

But really, what part of its overhead should the Census Bureau sacrifice to come up with this large sum? The communications office annual budget is less than half that amount. Shut down its congressional liaison activities? Ditch the press releases that inform the media and stakeholders about data products? Congress doesn’t seem to grasp the connection between Census Bureau data and the myriad policy decisions the public and private sectors make on a daily basis, so why bother? Cut back on protecting confidential information from 40,000 daily cyber attacks? Better yet, why not shut down the website entirely, thereby negating the expense of maintaining an Internet presence and defending against hackers — a sort of two-for-one reduction?

Frankly, given the country’s dire economic straits, I think we need to be really creative. Why don’t we furlough the entire senior Census Bureau staff (including the director), and then bring them all back in five years so Congress can blame the agency for not trying hard enough to design a simplified, less costly 2020 Census. Speaking of which…

Have I mentioned that Senate appropriators smartly challenged the Census Bureau to take the 2020 census for the same amount of money it spent on Census 2000, without adjusting for inflation? I’m all for saving money. The Census Bureau must bring the per-household cost of the decennial enumeration under control. In fact, the census director took the unusual step of announcing the closure of half of the bureau’s 12 regional offices, without a nudge from Congress, in a preemptive move to bring costs down.

But to go from spending $13 billion (in current dollars) to take the 2010 census, to counting 10 percent more people for a third of that amount eight years from now? I’m not feeling it yet.

But I digress. Things could be worse for the Census Bureau. It could be languishing under a temporary spending measure (the insufferable Continuing Resolution) with the many agencies that couldn’t get on board a little bus to 2012 funding certainty. House appropriators proposed cutting 21 percent from the bureau’s budget request, potentially dooming the quinquennial detailed measurement of the nation’s economic activity. Cooler congressional heads prevailed in the final hour, offering enough money to proceed with core Economic Census functions. But the Survey of Business Owners is on the chopping block — the only source of data on business ownership by people of color, women and (yes!) veterans.

As for the rest of the bureau’s programs, I suspect managers spent the holiday weekend scouring their budgets for additional expendable activities. The agency can’t cut $55 million from overhead and function effectively, so programs such as 2010 census evaluations and data products, 2020 census planning, the American Community Survey, and other periodic functions must absorb some of the pain.

The real problem is that, in order to yield savings anywhere near the magnitude of those money-green sugarplums dancing in lawmakers’ heads, the Census Bureau must invest modest but consistent resources now to research and test forward-looking methods that will expand response options for increasingly complex household structures. Cutting the agency’s budget to the bare bones won’t generate the level of scientific foresight necessary to tackle the depth of challenges inevitable in a society as culturally, ethnically and politically diverse as ours.

Memo to Census Director Robert Groves: Hold on tight to that piggy bank next year!

Time to Get Down to (Census) Business

by Terri Ann Lowenthal

Is anyone else weary of handicapping the Republican presidential field, or hearing about Amanda Knox (I’m glad she’s home) and Dr. Conrad Murray (MJ and I were born six weeks apart, so you know where my sympathies lie)? Good. Time to start thinking about Census 2020 planning instead.

At a Senate hearing last spring, Census Director Robert Groves laid out the agency’s guiding principles for designing the next decennial count. At the core of all of them is the stark fiscal reality facing the country: the Census Bureau will have to do more with much less. As in far fewer dollars to spend. More people, more housing units, more complex household structures, more language and cultural diversity. All for less money than in 2010. Have I mentioned that Senate appropriators think the Census Bureau could do the job for the price of the 2000 model (without adjusting for inflation)? Good luck with that.

Anyway, over the coming months, I’ll take a look at the eight guideposts Dr. Groves said are based on lessons learned from the 2010 count, offering some historical context and thoughts on key issues the bureau should consider in pursuing each goal. I’ll start today by repeating the underlying point from my post on Sept. 28: No matter how little it is willing to spend on the 2020 census over the long haul, Congress must invest some money upfront for research, testing and design development. The alternative will tie the agency’s hands behind its back until it is too late for meaningful innovation, end-to-end testing to support outcome-based decisions, and timely interaction with community-based partners.

I’ll close for now with another news headline of greater import to the census. As I write this blog post on my iPad and contemplate the untimely passing of Apple’s Steve Jobs, I am reminded of the speed with which technology has evolved and improved in only the last decade. Apple unveiled the iPhone in 2007, a mere four years ago. Is it just me, or does it seem like that gadget has been around forever? Director Groves has rightly highlighted the need for a multiple-mode 2020 census, expanding enumeration methods beyond the traditional (since 1960) “mail, hail, or fail” playbook. His Senate testimony (April 6, 2011) notes that response options must “reflect the communication platforms that people are using.” Well said, but difficult to actualize when you consider that my iPad was overrun by iPad2 within a year. Congress must give the Census Bureau sufficient resources to have technology visionaries in the room as planning for 2020 unfolds.