WASHINGTON, DC – This afternoon, U.S. Senator Rob Portman (R-OH), Ranking Member of the Senate Homeland Security and Governmental Affairs Committee, pressed the Acting Director of the United States Census Bureau Ron Jarmin on the ongoing 2020 Census data delay that will impact legislative redistricting in Ohio and across the country.
Excerpts of the hearing can be found below and a video can be found here.
Portman: “Thank you, Mr. Chairman. And I know that Senator Lankford has already asked you guys some pertinent questions because he has a strong interest in this given the situation in his state and also his expertise in this issue. Ohio is one of those states that’s negatively impacted, as you know. We have Constitutional requirements that may lead us to not be able to use your data. Which is just unheard of. So I guess, Dr. Jarmin, my first question would be, I know that due to COVID-19, you announced this six-month delay. The apportionment data only needed a four-month delay. My question for you, I guess, would be by law redistricting data should be available three months after the apportionment data, but this is five months later. Why that delay?"
Ron Jarmin, Acting Director of the United States Census Bureau: “I think part of that, I was just explaining to Senator Lankford, was while we were making an effort to make the statutory deadline for the apportionment data, which is December 31, we crashed the schedule, put all hands on deck to accomplish that, and that meant that some of the work that we would’ve started for the redistricting data was set aside until later and so I think that’s added some time. I think the other thing is we’re trying to make sure that we have this all right and make sure that the timeline that we’ve specified is one that we can meet. So I hope we can beat the timeline a little bit but it’s definitely, as you saw last fall when we had said, you know, timing on apportionment data was going to be this or going to be that, you know, we want to give the public and the states a little more certainty about when we can actually deliver.”
Portman: “Certainty is nice but having the data available is even nicer. We have a September 1 deadline, as you know. Let me ask you this, why can’t you provide states like Ohio, that have these Constitutional deadlines, data on a rolling basis? In other words, provide us hard data. You don’t need to provide it to every state, because some states have more flexibility than we do.”
Mr. Jarmin: “No and we understand that. The original plan was to put it out on a rolling basis but I think as we’ve gone back and tried to find the fastest way that we can deliver data to the states it is to produce it all at once. The reviews are done sort of across all the states. I think if we were to prioritize, you know, a small number of states that would actually increase the amount of time it would take us to get the data out. So we’re trying to optimize and get it out as quickly as possible.”
Portman: “I hope you understand the negative consequences here because we’re in the middle of a redistricting change in Ohio, not just because of the ten-year data but also because we’re looking at redistricting reform and this is really putting us back and causing tremendous problems. GAO has indicated that September 17 is your target release date with a two-week security buffer. Will you commit to notifying this committee as soon as you become aware of any issue that could delay the data past September 17?”
Mr. Jarmin: “Absolutely.”
Portman: “That’s at a minimum what we really need to do. I’m told that you don’t want more money to be able to speed this up because you think that won’t help. How could that be? Wouldn’t more resources be able to give you the ability to put more people against this and to get this done more quickly?”
Mr. Jarmin: “So that’s a great question, Senator, and so I think what you need to remember is that right now we’re in a process where data are processed and then they are reviewed and that review is done by experienced expert staff and trying to increase the number of people looking at the data would require those expert staff to then train new folks and I think that would just add time to this. This is one of the things, you know, when we talk about looking ahead to 2030, if there is something that happens in data collection that puts us behind for data processing, the data processing was not a focus of a lot of the efforts for the 2020 census. And so it was all done with thinking there were five months to get this all done, or at least to get the apportionment data out, so I think thinking about what the backend looks like for 2030 could help us in this, but you know obviously for 2020 with the pandemic, we didn’t have that luxury. And so it’s a situation where the processing of these data is done by a relatively small - I mean by small I mean 100 or so - expert staff that process and review these data and increasing that staff right now would actually slow us down.”
Portman: “That seems nonsensical to me, to be frank with you. That you can’t find some way to bolster your expertise, between now and August, which gives you a few months to do it.”
Mr. Jarmin: “We’ve already brought in some internal staff that, you know, have similar types of skills but if we were to bring in new people from outside, I’d have to take people who are currently working on this and have them teach these new people, you know, what to do.”
Portman: “Well, give me a commitment today that you will put your thinking cap on and see if there is any creative way that you could at least get us this data at least a couple of weeks early. I mean literally, we are going to have no time in Ohio to look at the raw data much less the final data and if we had another two weeks, it would make a huge difference.”
Mr. Jarmin: “We’re already looking for every way that we can speed this up and we will continue to do so.”
Portman: “I see Senator Romney is here and I want to go to him next but ask you a question that actually relates to some of his local communities. This most recent data set, I’m told that some small towns in Utah saw a 20 to 30 percent population change as a result of differential privacy. Do you expect similar results for small towns in the final data?”
Mr. Jarmin: “I do not sir, those test data were done with a higher privacy parameter so that protected the data more than we will in the final data and also we’ve worked out many of the other processing issues that caused some discrepancies and so we’ll be releasing additional data in April that folks can look at. That’ll be a lot closer to but even probably still not quite as accurate as the final data that we put out.”
Portman: “Do you believe the final data will be accurate as to these small towns despite the use of this deferential privacy?”
Mr. Jarmin: “So, accuracy and privacy are conflicting goals and we are trying our best to balance the accuracy with protecting the respondents’ information.”
Portman: “We’ve got to be sure the data is usable, one, and be sure that it can be upheld in a court because it is going to likely be subject to adjudication. Well, thank you.”