Census needs review, panel says

doi:10.1126/science.370.6515.391-b

GSTDTAP > 气候变化

DOI	10.1126/science.370.6515.391-b
	Census needs review, panel says
	Jeffrey Mervis
	2020-10-23
发表期刊	Science
出版年	2020
英文摘要	An expert panel last week called for an independent group of researchers to pore over the billions of bits of information collected by a truncated 2020 U.S. census and report publicly on whether the Census Bureau has met its goal of “counting everyone once, and only once, and in the right place.” The report, from a task force of the American Statistical Association (ASA), was prompted by what ASA regards as unprecedented political interference by President Donald Trump's administration in the $15 billion head count conducted by the nation's top statistical agency. “We are doing our best to support the Census Bureau because they have been put in a very difficult situation,” says ASA President Rob Santos, who co-chaired the task force, which includes three former Census directors. “They don't have full control of their operations.” Recent reports from two government watchdog agencies have voiced similar concerns. The twist is the latest in the troubled 2020 census, which went live on 1 April after 10 years of planning. The COVID-19 pandemic delayed field operations, originally set to end on 31 July, and prompted the Trump administration to request additional time. But the White House later reversed course and announced it would seek an earlier end to field operations; it also said it needed only three rather than 5 months to crunch the data. Civil rights groups and local officials sued, fearing that truncated fieldwork would miss a disproportionate number of residents from minority, immigrant, and low-income communities. But last week a Supreme Court ruling enabled the Census Bureau to end operations on 15 October. The groups had also pushed to extend by 4 months, to April 2021, the deadline for sending the president the final tally, which is used to determine how many seats each state gets in the 435-seat House of Representatives. But the administration has stuck to a 31 December deadline, creating what the task force fears is a formula for disaster. The Census Bureau has “eliminated many quality-control steps,” it asserts, and the agency's “current plan for quality assessment is unknown.” The census “is a chain of many, many operations, and it is only as strong as its weakest link,” says Nancy Potok, co-chair of the task force and a former deputy Census director. Statisticians are also concerned about political meddling. Trump has ordered the agency to find ways to subtract undocumented residents from the count—a task that many legal scholars say is illegal and most data experts say is impossible ( Science , 7 August, p. [611][1]). And the recent arrival of three high-level political appointees at the traditionally nonpartisan agency has raised concerns that they will try to influence the analysis in order to give Republicans an advantage in House apportionment. Commerce Secretary Wilbur Ross, whose department includes the Census Bureau, has downplayed those concerns. Last week, he said the census has met its goals because “99.9% of housing units have been accounted for.” But Potok calls that “a meaningless number” that can't be used to measure data quality. For example, fieldworkers might have marked an address as counted even if they repeatedly failed to obtain any information from residents, or gleaned only incomplete information from “proxies”—for example, a neighbor or landlord. “Ross says it's possible [to have a quality census despite a truncated schedule],” Santos says. But internal government emails that have been made public as part of ongoing litigation “say that he's living in a fantasy world,” he adds. “It's sophistry.” The task force report lists many indicators that the Census Bureau—and outside evaluators—could use to determine how close it has come to a complete count. They include the percentage of addresses enumerated by proxy, for example, or how much it relied on data already in government files to fill out a resident's demographic profile, which are also less reliable than self-responses. Statisticians could look, as well, at the percentage of records that lack a full name or date of birth, the number of duplicate enumerations, and how much information is being imputed. (Imputation means making an educated guess about the demographic characteristics of occupants based on indirect information, such as the type of housing unit or characteristics of the neighborhood.) Such data are readily available and accessible because this year's census is the first in which field operations were conducted electronically. This week, Census officials released data on the percentage of addresses enumerated by proxy and by using administrative data. Potok calls it “a good first step” but says researchers still need to assess the results at the smallest unit, covering a few city blocks. What the outside experts learn could affect how Census Director Steven Dillingham presents the final tally. “He could tell the president he cannot submit a count, or that he doesn't believe the numbers are accurate,” Potok says. “Or he could say that we've identified problems with the count that Congress might want to look into.” Santos doesn't think those scenarios are likely because they would require Ross's support. But he says the task force wanted to “send a message” about the value of using an independent assessment “to restore public confidence in the 2020 census.” [1]: http://www.sciencemag.org/content/369/6504/611
领域	气候变化 ; 资源环境
URL	查看原文
引用统计
文献类型	期刊论文
条目标识符	http://119.78.100.173/C666/handle/2XK7JSWQ/300266
专题	气候变化资源环境科学
推荐引用方式 GB/T 7714	Jeffrey Mervis. Census needs review, panel says[J]. Science,2020.
APA	Jeffrey Mervis.(2020).Census needs review, panel says.Science.
MLA	Jeffrey Mervis."Census needs review, panel says".Science (2020).