Tag: National Content Test

It’s a Horse Race

Census Project Co-Director Terri Ann LowenthalBy Terri Ann Lowenthal

And… they’re off!

First out of the gate: President Obama, who unveiled his Fiscal Year 2016 budget request for the federal government last week. The commander-in-chief took office a mere year before the start of the last census and promptly asked Congress for an extra $1 billion to shore up hiring and promotion for the decennial count, after plans to automate door-to-door interviewing in the 2010 census went down the drain. Lawmakers, ever mindful that they really really really should pay attention to the census in the home stretch, ponied up.

It’s sort of a perennial decennial habit of the legislative branch. Congress Rip-Van-Winkles through much of each decade, finding better uses for money that the Census Bureau says it needs to get a head start on the nation’s largest peacetime activity. This time around, the bureau is trying hard to save a lot of money (as much as $5 billion over the census lifecycle) by investing early in research, testing and development of new methods and technologies. Lawmakers, especially in the House, haven’t quite grasped the “pay now or pay later” concept. Or the “test drive before you buy” principle.

Anyway, the Obama Administration is pressing ahead, requesting $1.5 billion for the Census Bureau in FY 2016, including $663 million (+$317M over FY 2015) for the 2020 Census and $257 million (+$15M) for the American Community Survey (ACS). Lawmakers no doubt will cast a skeptical eye on the 91 percent increase for 2020 Census planning. The next census isn’t even on their radar screen, and fiscal austerity is a badge of honor for many legislators. It’s like asking your boss to double your salary, while trying to convince her that burning the midnight oil now to revamp the company website, marketing materials and customer service protocols will bring in huge profits down the road. Gotta be brave to make that case.

But, time for our annual cataloguing of what the Census Bureau plans to do with all that extra 2020 Census money. For starters, it has to get moving on the technology front. Remember, you heard it here first: the “cyber-census” is coming. That is, if Congress forks over enough funding to build production systems in 2016 (and 2017) to execute the 2020 Census design, in preparation for a big operations-readiness test in 2018. Major tests this spring (as well as the one last fall)—of high-tech, streamlined field work; use of “big data” to update the address list and government records to count unresponsive households; targeted digital advertising; and flexible Internet response options—will determine which sweeping new initiatives are worth pursuing. The fall 2015 National Content Test will inform final questionnaire changes—including the closely watched decision on whether to combine the race and Hispanic origin questions—which the agency must nail down by 2017.

The Census Bureau also will be playing catch-up next year. Congress cut $124 million from its budget request for the current year (FY 2015); the agency shaved $100 million of that amount from census planning and another $15 million from the ACS. For 2020, that means development of partnership activities, language translations, research on how the public views data privacy (think that one is getting any easier?), and other vital components of a successful census are again on the back burner. For the ACS, the so-called 3-year estimates are toast. That’s right: no more data averaged over 3-years (for example, 2012-2014, originally scheduled for release this fall) for places with populations of 20,000 or more. Going forward, jurisdictions under 65,000 population, which includes more than three-quarters of all counties, will have to rely solely on the 5-year estimates (e.g. 2010-2014). A lot can happen economically, socially and demographically in five years, possibly making this dataset less precise for many uses. The Office of Management and Budget has to approve the Census Bureau’s final 2015 spending plan, and chatter I’ve heard among data users suggests they may not give up these estimates without a fight. Budget-cutters in Congress, many of whom represent areas that will lose a valuable data source, might want to think twice before wielding the ax again in 2016.

I know I sound a little like Paul Revere: “The census is coming; the census is coming.” But, people, it is. If Congress doesn’t adopt the sense of urgency the census requires now, it will find its hoped-for reforms fading fast in the home stretch. And anyway, I needed a final horse analogy.

Limping Along the Road to 2020

Terri Ann Lowenthalby Terri Ann Lowenthal

So where were we, census fans?

As we rub the sleep out of our post-government shutdown eyes, and revel in data on the re-opened Census Bureau website, we are faced with the prospect that the uphill climb to Census 2020 just got a little steeper. Okay, maybe a whole lot steeper.

Let’s consider the good news first. The federal government is open for business. Surveys are in the field again. Economic indicators are, well, lagging, but the Census Bureau will publish them in due time. Small-city, county, town and neighborhood data from the American Community Survey will be late, but we’ll have them before year’s end.

Now the gloomy news. The federal government is running on last year’s gas tank. That ramp-up in funding to complete decennial census research and testing? Fuhgeddaboudit. With sequestration and across-the-board cuts, the Census Bureau’s budget was already 11 percent below the request for FY2013. That fiscal year has come and gone, but the ghost of budgets past will linger for at least a few more months in the form of a Continuing Resolution that expires on January 15. Not only is there no modest, yet vital, funding boost into the second quarter of FY2014, there is no way to dig out of the 2013 hole yet. Budget uncertainty plus inadequate funding equals increased risk of cut-corners and missed savings opportunities.

What’s at risk? The bureau has already pushed back a critical milestone — to select the 2020 Census design framework — by a year, to the end of FY2015. The National Content Test will be a year late, too, in 2016. That doesn’t leave much time to scrub the results before the legally-required content submission to Congress by April 1, 2017.

Now comes word that the Census Bureau is suspending work on several 2020 research and testing projects. It has “temporarily” reassigned 86 employees to other divisions because the money just isn’t there, and won’t fill a similar number of vacancies for 2020 Census planning unless it gets a sufficient funding boost for FY2014. People, this is not a promising sign of robust progress towards fundamental census reform. Congress can twiddle its thumbs while the fiscal clock ticks away, but the Census Bureau cannot defer the day of reckoning — April 1, 2020.

The Census Director says the agency will be able to complete the most critical 2020 Census activities planned for the current fiscal year. So why am I still tossing and turning at night? The Government Accountability Office (GAO) has identified IT management as a “long-standing” decennial census shortcoming. But thanks to sequestration and other budget cuts, acquisitions and systems development for 2020 will start a year later than planned. Let’s hope the bureau doesn’t have to cut corners when it comes to integrating and perfecting IT and field infrastructure.

Then there are promising strategies for maximizing electronic self-response on the front end and increasing follow-up efficiency on the back end. But is cyber-response the antidote to Congress’s census sticker-shock nightmare? Consider phishing scams, NSA-fueled privacy jitters, and finicky rural Wi-Fi connections before you answer. Can you imagine inviting scores of millions of households to participate online and not fully load-testing the system? Can you say healthcare.gov?

Maybe lawmakers should repeat Cyclical Program Ramp-up 101, before the 2020 Census is back to paper, postage and pencils — for twice the price.