Jumpstarting the census

Like everybody else in Ulster County, you and I are getting older every day. Were we residents being replaced in the population as we age by younger people, we might be more sanguine about accepting that vagary of the human condition. But we aren’t. According to the newest data from the 2013-2017 American Community Survey (ACS) conducted by the U.S. Census and released just last week, the Ulster County median age is 43.6, up a year and a half from 42.1 just five years previously. We’re aging fast. The median age in the 2000 census was 38.2.

Fewer of us are working at home. Predictions that digital connections would enable more of us to work at home haven’t panned out — at least not yet. In 2007 6.6 per cent of Ulster County workers worked at home. In 2010, according to the census, that proportion increased to 6.8 percent. But the percentage of home workers in the last two American Community Surveys, including the one released last week, dropped to 6.4 percent.

Other data sources have told us that more Ulster County residents are commuting longer distances to work in other counties rather than holding jobs locally. The latest ACS shows a mean travel time from home to work of 28.2 minutes. That number has been heading slowly but steadily upward from the 26.4 minutes recorded eight years ago.


The data on health coverage, a recent addition to the ACS, has shown dramatic improvement. The percentage of uncovered Ulster County persons was 11.2 in 2010 and 8.4 in 2014. The number released last week for 2013-2017 show another drop to 6.9 percent.

The American Community Survey provides what the professionals call a “rolling sample.” The census contacts 295,000 people a month, or 3.5 million a year. It pursues those folks who don’t respond diligently enough to satisfy the statisticians that the overall sample is a reliable one. I’ve heard the sample gets data on about two percent of the population each year. For a geographic entity with a population like Ulster County’s, two per cent leaves a large margin of statistical error. That’s why the ACS consists of five-year estimates. The numbers released last week combine the surveys of the last five years of counting, 2013 through 2017, into one estimate likely to be statistically reliable.

How the federal census has grown from its modest beginnings!

As most schoolchildren learn, the census is enshrined in Article 1, Section 2 of the national constitution: “Representatives and direct Taxes shall be apportioned among the several States…according to their respective Numbers…The actual Enumeration shall be made within three Years after the first meeting of the Congress of the United States, and within every subsequent Term of ten Years.” The kids also learn of such peculiarities as that each enslaved person counted as three-fifths of a person for the purposes of determining state representation in Congress. 

The founding fathers were aware of the potential of using census data to manipulate political boundaries for partisan purposes, the most famous example of which gave rise to a portmanteau word named in honor of Massachusetts governor Elbridge Gerry’s artful design of a state senate district north of Boston. One only has to read the headlines to verify that the political practice of gerrymandering remains alive and well more than two centuries later. 

When Congress established the first federal census to apportion representation among the states in 1790, James Madison suggested the census take advantage of “the present opportunity” to gather valuable information that went well beyond “the bare enumeration of the inhabitants.” If the census was used to provide a “description of the several classes [occupations] into which the community is divided,” Madison said, that information would prove “extremely useful, when we come to pass laws, affecting any particular description of people.” 

Over time, Congress both added more to the census and gathered information through other means. In 1812 it authorized a study on manufacturing. In 1840 the census asked questions about fisheries, and in 1850 about social issues. Questions about crime and disease were asked between 1850 and 1880. Over time, Questions about among other subjects poverty, race, housing, health insurance, education and commutation have been added.

Particularly in the past few decades, the federal census has developed a whole ecosystem of private and non-profit partnerships. As might have been inevitable in a burgeoning information age, the synthesis of census and private data has become commonplace. The Bureau of the Census, part of the federal Department of Commerce, publishes various surveys of different periodicity and five-year industry censuses.

“Raise awareness to ensure that your constituents, customers and stakeholders are represented in vital census bureau surveys,” the federal census agency recently urged. For better or worse, the digital age has increasingly accelerated information interaction among organizations.

The day after the release, a small Kingston firm, StateBook International, announced that it had already incorporated the new ACS five-year estimates into its platform. New data for the categories of demographics, education, governments, households and incomes, housing and workforce had been posted. 

StateBook explained that the federal government is its primary data provider, and that considerable other data came from private sources and non-profits. On its website, StateBook, headed by part-time Big Indian resident Calandra Cruickshank, claims to be the first nationwide online marketplace for economic development and corporate site selection headquartered in New York State.