Report on Census Test Identifies Concerns Ahead of 2020
Last week, the U.S. Department of Commerce Office of Inspector General’s Office of Audit and Evaluation released a report analyzing the 2018 end-to-end test done to prepare for the 2020 census. The report concluded that the Census Bureau must address several concerns before 2020 in order to complete a successful census.
The end-to-end test is the last major test the Census Bureau runs in preparation for the constitutionally mandated decennial census, the next of which will take place in 2020. Last year’s end-to-end test was particularly important because the Census Bureau is changing their canvassing operation for 2020.
Traditionally, nearly every block of the country is canvassed by temporary employees who compare what they see in the field to data generated by the Bureau. This census will split the operation between in-field canvassers and in-office canvassers, who will compare data to captured aerial imagery to decide if they truly require in-field review.
According to the IG report, the Bureau estimates this will reduce the workload of in-field canvassers by 70 percent.
The end-to-end test took place between August 28, 2017 and October 6, 2017 in three cities. The test focused on passive blocks, which had no additional structures added since the last census, and therefore would be candidates for in-office canvassing.
The IG report found, “OIG analysis of the 433 passive blocks included in the 2018 E2E Test found that in-office address canvassing results differed from in-field results in 61 percent of the blocks tested. This does not mean that all of the housing units within those blocks were missed; rather, it means that at least one housing unit was added to or removed from the address list by a field representative. In total, we found that field representatives added 1,087 housing units and removed 1,287 housing units in the 433 passive blocks. This indicates that in-office address canvassing is yielding incorrect results or, at the very least, results that are inconsistent with in-field address canvassing.”
The report further revealed that the end-to-end test showed a reduced capacity to identify and combat fraud and had severe limitation in the ability to “deliver a complete and accurate address list and spatial database for enumeration and determine the effectiveness of address canvassing operations.”
The IG analysis warned that the limitation to the in-office canvassing capabilities increases the risk of operational problems that will likely remain undiscovered during the census count.
The report offered several recommendations for change to the Bureau before the 2020 count. These included, but were not limited to, evaluating how the number of incorrectly categorized passive blocks will affect census quality, how those errors may affect demographic groups, and assessing the risks to the 2020 census that have arisen as a result of the limitations identified during the 2018 end-to-end test.
Posted in Featured News