Public Trust Transparency, Information, and Democratic Engagement

Okay, we are at the top of the hour. Good morning, good afternoon wherever you may be and welcome to our help. I'm an accidental government information librarian web series. I'm Erin Wilson. I am the serial and government information cataloging librarian with the University of Maryland and I serve as one of the co-chairs of the Gold Door Education Committee. We welcome you to our webinar for today. Um before we begin, we want to take care and address a few housekeeping and logistical matters. Um we reminded all participants, we ask that they stay muted during today's presentation. Um and you are welcome to drop in questions for our participants in the Q&A section of the webinar portal. We will be monitoring those questions as they come in and we will be answering those questions in time during present our presentation. So to give our official um welcome and opening, I will invite Julia Ezo to give that. Julia serves as the government information packaging and political science librarian at Michigan State University and chair of God. So Julia, take it away. >> Thank you and good afternoon everyone and welcome. I'm pleased to officially open today's session on public trust, transparency, information, and democratic engagement. Our goal today is to highlight the importance of open government, access to information, and accountability as foundations of a healthy democracy. Before we dive in, I want to extend a huge thank you to our organizers, the Godor Education Committee Sunshine Week planning subcommittee, support from the Godor Education Committee, as well as the Godor help. I'm an accidental government information librarian series, our ALA staff leaison Danielle Pontton, and the Godor social media team for all their hard work on this. I'd like to introduce our event planning chairs, Anna and Patty, who have been instrumental in bringing this together. Anna Jang is head of the Kelly Center for Government Information and Civic Engagement Bond Library at Rice University. Patty Teach is the political science librarian from the University of Florida. Anna, I'll pass the mic over to you and I hope you all enjoy today's wonderful panel. >> Thank you, Julia. Um, very introduction and hi everyone again uh to this uh event today. I am Anashung from Rice University. Uh now it's my pleasure to introduce our uh the our honored panelists today. Uh the first panelist is Dr. Connie Sichra. She is senior scholar in the committee on national statistics at the national academics of science engineering and um medicine. member of the American Statistical Association project to assess and monitor the health of federal statistics. She previously served as CEN stat director and the senior study director. For Zen stat, she has worked on studies of the 1990 to 2020 censuses measurement of poverty and income and the CN stats flagship publication principles and practices for a federal statistical agency. Our second panelist is Patrick Leerty. Uh he is director of research at the Connecticut Connecticut Department of Labor, Connecticut's uh labor market information shop and the data inter uh integration hub for P20 wing. Connecticut is a state longitudinal data system on the P20 wing data governing board and serves as the Connecticut Department of Labor's agency data officer, adjunct faculty at University of Connecticut Labor Commissioner uh designate on the Connecticut Retirement Security Program uh and a New England representative on the local employment dynamics uh steering committee. So he is an expert uh sitting at the intersection between federal and the state data. And our third panelist today is Michelle Ridden Node. Uh she is the executive director of the CTD uh CT data collaborative executives the vision and the strategy of the organization which seeks to democratize access to public data and to facilitate datadriven decision-making. She holds a bachelor's degree in mathematics from Boston College and a master's in public policy from the University of Chicago. Um please join us to welcome our three um panelists. And uh next uh we have two moderators today, Jessica and uh Jennifer. Jessica is a social science librarian for research engagement and government documents librarian from Mark's science and social science library at EO University. Jennifer is business information consultant in Lowinger Library at Georgetown University. Now I will let Jessica take over. Let's get started. >> Okay. Thank you so much. Um and thank you Connie and Patrick and Michelle for being here today. Uh I am going to be asking the first few questions and these are going to be for all the panelists to answer. Um so the first one uh is as follows. Trust in government institutions has fluctuated significantly in recent decades. From your perspectives, whether through government information, statistical systems, or public communication, what do you see as the primary drivers of declining trust today? Um, may I go ahead? Oh, sure. Yes. >> Okay. Thank you. Well, I want to thank God for inviting me to this very important session in this time of yes low trust in the government and many institutions. I thought it'd be useful just to give very quickly some basic overview. The Pew Research Center has followed this issue of trust in government. There is a question, do you trust Washington to do the right thing? almost always or most of the time and that was actually at almost 80% in the early 60s. It fell just monotonically to 30% in 1980. It's been up and down a few times since then, but never more than about 55% which was achieved in uh 1991 the first the Gulf War and in um uh 2001 with 911. But since 2011, where that measure was 15%, it's been in a very tight range of people not trusting the government to do the right thing. And it's currently at about 17 only 17% of the public does have this trust. And again, there are whole variety of factors in this period since 2011. And I think some of them have to do with from President Obama through Trump to I will call it a second administration. The public's in whipsaw big changes in political control including off-year elections that change congressional control and lead to standoffs and inability of administration to carry through on their program. And voters are confused and frustrated. And then we've all read about the polarization among political elites and the public. Wars COVID which really did a number on us as a society and missteps in handling of guidance and information have left mistrust. People want to do their own research rather than looking to expertise. And then economic hardship and inequality have left many people stretched and upset that the government doesn't seem to be able to help. And then there's some media out there that sort of actively fment distrust. So there's a lot going on and even federal statistics which are so necessary for us and should be above the fray. trust in their credibility and willingness to participate in surveys have been affected by a number of factors which I can talk a little bit more about but I don't want to monopolize let some other folks talk. Thank you. Yeah, I mean I would just add on to um Conniey's comments um in particular thinking about the statistical system and what we've seen is um with the first Trump administration threatening the citizenship question on census 2020. I mean that has led to an erosion of uh trust in government data, all government data that's collected. So um and we see it continue today through um as we are leading up to the planning of census 2030. We have huge trust issues again. assistanceship questions coming back in addition to other mechanisms that have taken place over this past year such as sharing of data between agencies um without proper governance um and um all the actions I would say of Doge um have led to severe decline in trust in collection of government data of the public's data sorry >> u yes I I totally agree on that. And then there are some genuine problems that are affecting quality of some federal statistics. Declining response rates being a huge uh huge one. And worryingly response, you know, people just don't like to answer surveys of any kind. But for some key government surveys such as the American Community Survey which is very widely used refu and it's a mandatory survey supposedly refusals have gone up. It's just not that people aren't at home or they don't you know pick up the phone or whatever. Active refusals have have increased um and that's very worrisome. >> Yeah. Patrick, was there anything you wanted to add to that? >> Sure. I I you know obviously we do all the a lot of surveys here and certainly we've seen response rates fall and uh part of it is a lack of trust in general not just in government and some of that unfortunately lack of trust is uh is is correct. There are so many scammers out there so many identity thieves out there. So when we send out a survey or we make a call uh we get so many calls from potential respondents just asking us is this for real? Right? >> Because the uh the fraudsters have gotten so good at faking that they are us. And so some of the lack of trust in us as government uh and I am in government um uh is is really just a lack of trust in general. And they don't even believe we are who we say we are. and and uh fortunately we're able to sometimes convince people that yes, we really are the labor department and yes, we really need your data and yes, they are then willing to cooperate. Um but the the idea that they might be getting uh being a victim of identity theft is not a crazy idea. Um, I mean myself, I almost got scammed by somebody who was claiming to be my uh my internet service provider in my home and um they seemed to be from my the company I am a customer of. They seem to be asking me to do things that were correct and then you know halfway through the call I realized this is not really for real. Unfortunately uh I was not a victim but I was I was on a list. they wanted make to meet me a victim and that experience would make somebody less willing to respond to a government survey. >> I think that's a great point. Um and thank you all for for bringing up yeah that responding to surveys and actually providing information is is a big part of this and it it involves a lot of trust. Um so following up on that um and about talking about federal like statistical agencies um these agencies play a central role in producing trust and public data from census counts to economic indicators. Um what institutional practices help ensure that official statistics remain credible and insulated from political pressure? Well, uh again if I might uh start uh the government statistics really are key infrastructure for the country for our society our economy planning state local federal government media just citizens and they need to be objective and credible and trustworthy. So they need to be insulated and I think that can be harder to do in our system perhaps which is highly decentralized compared to most other countries. There's no statistics USA. There are 13 principal statistical agencies like the Census Bureau, a hundred other biggestish statistical programs around the government and a chief statistician in the office of management and budget. And the chief statistitians position was authorized back in 1986. And there's an inter agency council on statistical policy that was authorized in legislation in 1995. And over the course of the last uh uh the 40 years, the chief stats office has issued a whole bunch of statistical policy directives and standards that cover things like how you release things like the unemployment rate so that nobody gets an early peak, that kind of thing. And standards for how race, ethnicity measured or industry or so on. Also in 2002, some very strong legislation was passed to secure the confidentiality of federal statistics. There are stiff penalties on uh anyone in a federal statistical agency under this act who lets individual data be identified and it also applies to people who do like surveys for them or researchers who are sworn in. And then there is the bipartisan foundations for evidence-based policymaking act of 2018 which is about strengthening the ability of the federal government to uh make policy based on actually evidence and it has a whole title about federal statistical agencies and how their parent agencies like the Census Bureaus and the Department of Commerce are obligated to protect and support them and keep am able to be objective and credible. There are also things like my unit at the National Academy of Sciences, the Committee on National Statistics issues something called principles and practices for our federal statistical agency and we update it every four years. So a new administration or a second term it's available and all this is to perfect the professional independence of the agencies. The problem we face right now is that this administration is uh barely constrained by legislation. It is not constrained by norms or directives or academy publications. In the future, if we get a chance, we might want to have some kind of statistics law doesn't necessarily have meaning consolidate the agencies because just to place them east on each on a stronger legislative footing so that they are better positioned to uphold. Again, they need professional independence to use the best methods and the best data and not having anybody say, "Well, we'd really like you to how about delaying this report till after the election or how about uh you know slightly uh you know downplaying this statistic or something. That is what we do not want. So there are institutional structures but they are not as strong as they need to be. >> Wonderful. U Michelle or Patrick did you want to add to that? >> Well I would also say that you know there's been some degradation in data quality due to uh lack of resources. So you know you know we're partners with the Bureau of Labor Statistics. uh we in state government also have seen you know a decrease of the my staff is much smaller than the staff that was in existence here uh when I started working in this agency you know uh 16 years ago. Um and uh that that the the degradation in quality just because of the lack of resources sometimes gets conflated with political interference and you know we've seen that recently with the big revisions um and you know uh but I I would also argue that you know in some ways it has held I mean the the firing of Erica McIntoer the wonderful Erica McIntoer she really is fantastic person um uh didn't lead to a um uh any sort of political change in the way BLS is run. the person who's now the acting uh uh uh the acting commissioner is a uh longtime BLS employee has served under multiple administrations of both parties and is very objective and the first nominee of the new administration who might have been a more political uh commissioner he he was withdrawn because it was clear that he wasn't going to be confirmed and so at least for now the institutional framework has held Um and then I would say in terms of just very specifically to get into the weeds um here in labor statistics, our main uh statistics come out of the unemployment insurance system which is while it's federally man it's managed under federal rules, it's all state people and so the data we get from our tax system uh and this is true throughout the nation uh would be very hard to manipulate that. you would have to manipulate hundreds and hundreds of people throughout the country and who are state employees uh who are collecting data to help uh administer the unemployment insurance tax unemployment insurance system. Um and so uh you know we have some safeguards in a decentralized system that you might not have if things were more centralized. Maybe just to add on a little bit, Jessica, I think you you asked about, you know, institutional practices and what comes to my mind are um strong data governance. So, I think that's a key piece of um ensuring statistics remain credible. To Patrick's point, stable multi-year funding also ensures that. Um having independent oversight bodies is also critical. um so that you have you know people monitoring what's happening also adding their expertise into um the statistics in the methodologies um although Patrick we haven't seen you know um bad things happen or things have stayed stable right as with the appointment of um the commissioner based on um political loyalty not necessarily merits that also does send a signal to the system of you know um that's different than has been before. So I think that also gets back to the trust you know when you have that kind of um appointment you don't you know it does degradate trust a bit in the system um and so I just I would also add just you know consistency and transparency around methodologies about privacy and confidential confidentiality guidelines are important all to all good practices right to make sure that um the statistical system remains credible and reliable. Yeah, M. Yeah, I don't want to uh be too naive here. When the president of the United States accuses the BLS commissioner of manipulating the data, that is bad. >> Yeah. >> And that erodess trust. >> Yeah. >> Um and he did that and it was not was not good for the system and it was not good for the country and it was not accurate. >> Right. Right. But there was an outpouring of uh press attention. He picked the wrong agency essentially the wrong set of statistics. Uh there was outpouring internationally in the business uh press. The Wall Street Journal kept on this story for a long time and it did uh stiffen the spines in Congress to tell the president that his first nominee was not qualified. The person who is the nominee right now is a credible you very experienced uh very experienced person. But your point Patrick about budgets. Yes. Since the great recession essentially statistical agency budgets at the federal level have been going down in uh inflation adjusted terms and that has just sort of uh accelerated uh recently. But even more troubling is all the effort which wasn't necessarily aimed at fiscal agencies directly but has had really uh made things really tough for them is all the staff losses because of not hiring people who were still in their probationary period giving people these buyout offers outright riffs and uh there are agencies u you spoke Michelle about the 2030 census the uh current deputy director of the census bureau, Ron German, who steered the 2020 census through a very difficult time with COVID, is retiring in May. And uh there's really a lack of just um people at all levels, but including at the top of many of the agencies to handle the the load that they have. And that that's very troubling. Thank you all for Oh, sorry. >> Sure. I was just going to say I did put in the chat uh one to the Pew uh uh data on the trust trends over time and one to the American statistical association project on assessing the health of the federal statistics. There is a lot of material there. We've done reports. We do uh monitoring of what's happening both good and bad. There have been some recent uh hiring opportunities for instance at census and BL um I think BLS which is which is great. There's a lot of resources there for folks who work with with with government data. So just wanted to call that to your attention. Thank you. >> And I'll just add an example to Connie's um uh mentioning the staff firings. um there with the the federal poverty level um the the survey went out went out in the field but when the data came back the entire department had been fired rifted or so there was no one left there to calculate the federal poverty level. So that's just an example of um and you know they have contingency plans but they don't have contingency plans for if the entire department is gone, >> right? Um, so that's just an example of of what Conniey's referencing. >> Thank you. Thank you all so much. I'm so glad that we have um different federal and state perspectives here because it is a really complicated picture um of what is going on and the efforts to um kind of protect the independence of these of these agencies. So I I really appreciate everyone's um takes on this and and the resources. I know I have already benefited from the American Statistical Association uh reports to you know to keep up to date on what is happening um and what the main concerns are. Um and uh and in that vein uh the the next question is actually on uh more uh to do with um people using uh government data. So, what should librarians and others who utilize state and federal data look for when they're evaluating whether to utilize a federal or state data source? That's a great question. It's a tough question for every data user. Many users don't have time to do, you know, become expert in various uh kinds of data. But I think a couple of the questions you ask yourself about first of all um can you find documentation for the data source of that explains how the data are collected uh whether it's from administrative records or surveys or scraping the web or whatever and if you can't find anything very uh I would be um uh question that uh in the federal statistical agencies of course they do provide documentation. It's not always um Census Bureau, which I love, but they do so much that it's sometimes hard to find what you're looking for for a specific data set, but it is there. I think the first question is really um one of relevance. A data set can be accurate, but if it's answering not answering the question that you have, why or if it's answering a question that's different from what you think it's answering. So, a good example of that are the unemployment rates. There are actually half a dozen unemployment measures that BLS reports, uh, maybe even more than that. The flagship one has to do with what people were doing last week and whether they worked even just a little bit or were searching for a job. But there are other measures that take into account underemployment and long-term unemployment and just what among the adult working age population who is who is working and those may be a more relevance. I just saw a piece today about what's becoming uh important in in labor circles is not just like minimum wage but hours because many companies for low-income workers don't give people you know they have this just in time scheduling and you can't work enough hours at any wage to to put together an income. So relevance really is important and then of course given that you have something that's relevant or useful for you accuracy, reliability, margins of error, biases and that's why falling response rates are a real problem. They increase the um they make the estimates less reliable in in statistical terms and they also may introduce biases. Now the statistical agencies don't put out estimates that they think are you know not not good enough but what they may do is say well we can give you a national number but we can no longer give you the geographic uh numbers that you want because we have this problem of of uh again of of of falling response rates or perhaps not enough budget just to collect the kind of sample size that we used to collect. So first make sure the data are actually answering the question you have and then look for accuracy and reliability and um and if and your role as librarians if the documentation's not clear or doesn't seem adequate fire off a note you know let let agencies know because they do want to serve the public but sometimes they don't they are not themselves a data user from the outside and so they may not have explained things as clearly as they could, that sort of thing. >> Michelle or Patrick, is there anything you want to add to that? I do have a brief follow-up question, but I want to see if there's anything you want to add to that. >> Well, you know, I I would just agree with Connie and just say that, you know, the data often does not speak for itself. I mean, it really pro needs context and, you know, there's, you know, You can lose credibility in a lot of different ways. And and to me, there's a big difference between, you know, a president accusing political manipulation and say the Fed chair criticizing the BLS's birth death model. Well, that's a legitimate discussion. And the birth death model problem really comes back to the resources again where I think they would love to modernize it and just don't have the ability to do that. um uh and they kind of have to defend their output even though they kind of know there's some problems with the birth death model. Uh and this is when businesses are born and and died. The other thing is to also understand that there's a big difference between administrative data and self-report and you know one of the issue one of the realities of the past decades or so is the number of people who selfidentify as biracial has increased dramatically. doesn't necessarily mean that the population has changed that much. It's just that people are self-identifying. I mean, I get asked all the time, how do you decide who's Hispanic? And I answer, I don't >> people people anyone who calls themselves Hispanic on our survey, we say they are Hispanic. And we aren't in the business of DNA testing or anything else. We're not ancestry.com. if we we deal with self-identified data and you know I selfidentify as a white man that's what I'm counted as >> um >> yeah yeah that that's a really good point uh there is a new race and ethnicity standard from the office of management and budget uh and the revised uh standard that was last revised in 1997 seven and it's going to be used in the American Community Survey I think in another year and then in the census and but administrative records like birth and death records which the census bureau uses to update population in between uh censuses they won't change to this new uh setup overnight. So there's going to be a lot of issues there. The Census Bureau is planning to help users by doing what they call crosswalks and but it is yes it is a um uh that's a particular area where how people are identified by race and ethnicity varies a lot depending on um a whole bunch of factors such as are they being asked are their parents being asked are somebody in a hospital say you know being asked uh etc. Thank you. Um, so my my quick follow-up question to that, and this is more kind of at the state uh level, but what kinds of state data do you feel um are most overlooked or but that may be useful as uh institutions seek reliable data? >> I'm interested in the answer to this one myself from hopefully from Patrick. So, >> well, I mean, we produce a lot of statistics and um, you know, I would just say that, you know, from whether it's someone in a local economic development board to the governor's office, they don't look at them, they don't read them, they call me up because and I, you know, they'll even ask me for things that are I', I've re is hilarious nonsequator, but I've received Freedom of Information Act requests for things that are on our website. So like you don't need to do an FOI request. Just ask us. We'll tell you the answer. Um so um you know I think this is where the librarians could really help. I mean there's there's there is so much data that is being produced. Um I do believe that at least from the state level our state statistical agencies and and this is goes red states blue states uh our state statistical agencies are producing very good data. I I do think we're do we're working hard and this is one of the questions that's coming. We are working hard to make it more accessible. Um perhaps some of it's a little obscure. Some of the our website itself is very dense and maybe it's hard to find the right thing. Um uh so I think there's a lot that's overlooked. Um and um all I can say is we're trying trying to get it out there. Um >> I just um would just add that um at the state level and I think one of the advantages that you have in terms of state data is administrative data. So we're you know talking about federal statistical data which is would be different than the administrative data but you know thinking about Medicaid data you know a portion of that does go to the federal government um in a standardized format so that it can be compared across the 50 states but Medicaid offices you know within states have a rich have rich information that goes beyond what's available at the at the federal level. So that's just one example. I mean, if you you can think of a myriad of programs that are administered by states that would have, you know, this administrative data that that can be, you know, depending on um the type of data you're looking for, can be tapped into for um research purposes, program administration purposes, you know. So there's so I think that's a a piece that as we start to see and the degradation of the federal statistical and this administration trying to um you know I don't know how to say this in nice terms but you know destroy the um federal statistical system. Um, I think it's going to require many in the states to figure out like what do what do we have what are the assets that we have and and start funding them and start building them up and using them too >> and working collaboratively across states and with um um you know relevant academics and and so on. And there are a number of uh state consortia on in various subject matters I know that are doing doing great things. Um >> thank you so much. Um yeah I I definitely think there is a tendency at least that I've experienced of people to kind of and researchers to jump to FOYA when that's not necessarily that's not really the fastest or most effective way to get information. Um, and there is really a wealth of information particularly at the state level that that a lot of people miss out on. Um, all right. So, I am I'm going to hand over the next set of questions uh to Jennifer. >> Oh, you're muted. Thank you. And this is going so well and and it is just so good here that I'm not going crazy that the you you know the response rate is going lower and the statistical and it's like what could we be doing? So we've been talking about uh what the federal government has been doing and the state governments are doing. Let's now talk a little bit more about what the users um and so uh we've been talking about users but many citizens interact with the government primarily through data dashboards their statistics and reports. How can statistical agencies and public institutions communicate complex data in ways that both are transparent and understandable to the public? That's that's the that's the big challenge. In fact, I see something in the chat about redesign of federal websites being not helpful and not quite sure which agency the commenter is referring to, but uh I certainly know examples. It it is that is a um a challenge and I think that public institutions, agencies, groups of users, libraries I think could be absolutely critical. I would love to see projects that focus on uh students and the general public's education about how to read statistics, say in the in the in the in the news media. Um for example when economic indicators come out there it is essential that for trends that things be put in inflationadjusted amounts because otherwise it is misleading when particularly if inflation is a bit on the high side that doesn't always happen. I think I I sort of imagine walking into a public library or a museum where among other things they have one of these um you know computer enabled thing was what would you like to know about your neighborhood and you give topics like unemployment or crime or or uh something and it would then take you through uh okay here's some data here's how to the when I was talking about relevant you know which unemployment rate might be the most meaningful etc. And it would in also explain things like inflation adjusted real dollars versus nominal dollars and things like that. The um there are some programs the Census Bureau has a statistics in schools program that has some very helpful uh materials. The New York Federal Reserve Bank of all things has a comic book series explaining the history of money and of inflation and all the terms that are used and it is really quite quite good. Now, somebody has to have some funding for some of these ideas. And I do think it would be wonderful if a group like God, the Association of Public Data Users, all kinds of groups like that were to get somebody to fund a real kick the tires for Federal Statistical Agency websites. And how could they be made more user friendly for users of varying levels of expertise? because there's a huge amount of variation in how useful they are across, you know, from the Bureau of Justice Statistics to the Census Bureau to the National Center for Health Statistics, etc. Um, I would love to see something like that. Meanwhile, I think statiscal agencies need to be encouraged to um be as clear as they can to have things like glosseries and cross references. So, you want to learn more about, you know, why we use this inflation indicator, how we do the population, update the population estimates, you know, go here, that kind of thing. But it is it is a real challenge to figure out priorities for what would be most helpful in this regard. But again, I have this vision of that there would be in my world there would be a lot of userdriven okay fix your website this way and you'll find usage exploding because people will be more able to understand where to go on how to get the data. And that, by the way, should help you with Congress. You know, the more users you have, the more um the support hopefully you can build. >> And I I guess I would um push back a little on um when I think of the government's role, I think of it, they're primarily responsible for making public data available. Yes. And so I would never want to see resources diverted from um >> making public data available um to making it more accessible um because I think you know that is the role of government is to make it available and it could be I could be a little biased because my organization in Connecticut data collaborative we are an intermediary. We consider ourselves the data intermediary. So our role and part of our mission is to make data accessible. So we focus on taking the government data that's made available and through visualization tools and you know different um ways trying to make that data accessible to a broad audience. So and increase and we work on increasing data capacity skills, data literacy skills. So, and I see librarians as a key piece of that. Like librarians are data intermediaries and your role is to be the translator, to be the facilitator of here's the data that's available and now let me help you put it into context. Let me help you find the metadata, let me teach you how to use it. Um, and so if data is not made available by the government, we then we don't have that accessibility. So I I go back to just I think we need to the primary role of government should be making p you know public data is a public good. So keep focusing on the public good. >> I to I totally agree with that. I mean if the data are not available then you know if it's only or put it this way if it's only available to uh say researchers who have money and time to go through the very arduous process to work with uh which is now possible to work like with confidential data if if agencies were to pull more of their data into that kind of environment that that would on the other hand there are things like I don't know how many of you use or try to use data.sensus.gov. >> It is not it is not great and it is um and it's never been great since it was factfinder and so on and um and I think part of that is the Census Bureau doesn't have the resources and they don't really kick the tires enough with users. Um and but many users will go, you know, they want something, you know, the the somebody needs a number for the mayor, you know, in the next 20 minutes or 10 minutes. And so I think you need both. I think the intermediaries are very important, but I also think that agencies with some structured feedback could improve um their own um u their own things like websites and and and so on. Patrick, do you have any comment? >> Yeah. Well, you know, I mean, I guess I'm on both sides here. We are I got the comment about I don't like the redesign. We're going through a modernization of our website. Now, one uh criteria though is that we are doing everything we can to make sure that it uh that in the near future our website will be compliant with the ADA requirements. And the person who's in charge of that whole project for state government at as a whole, not just with regard to data, but everything about state government websites uh really has a very strong philosophy that it is about accessibility for all citizens and it's not just focused on folks that might have a visual impairment, although that's obviously an important population, but for everybody. So, you know, they've been pushing us to try to make our language more accessible in terms of lowering the grade level. And sometimes it gets tough because we had a lot of jargon, you know, in economic land here. Uh, but we're we're uh we're at least hearing that and we're going to try to respond. There's a there's a certain limit that we can't if we we have to accurately describe how the unemployment rate is calculated. uh and so if we don't do you know it's hard to do that without getting into some jargon uh so uh it's a balance but it is something that at least we're working on and I know other states are as well uh to try to uh make make our situation more accessible knowing that we'll probably always need the med intermediaries Michelle don't worry we're not going to put you out of business uh uh but we are going to try to make it easier uh for some of our users And I still think that in the end people are going to still go to their librarian, go to Michelle, call me up and just say, "What's the unemployment rate in Waterberry?" when they could easily find it on our website. But um somehow have difficulty with that. Um and then what one thing we don't know about the future is, you know, to totally go off topic for a second here is the whole, you know, what is AI going to mean? because AI will make it so much easier for some people to be able to scrape our website, get all the data, format it any way they want. And then there may be this may be a new digital divide where where some users are just able to customize everything and get it all. and others, you know, are who don't have that skill or access to those tools will be uh needing maybe even more handholding. We'll see. >> I I think that would be a great question for towards the end because I think everybody here has an opinion about how AI will save us or condemn us. Um but let me get to another question which is inst which is kind of related institutions like libraries, universities, journalists and civil so uh society organizations often act as that inter between the government data and the public. What role should we the institutions play in strengthening public understanding and trust trust in official information? Well, I think you know it does take some time to learn what's going on. So like when we have a big revision in our data for example, some people would want to like have some conspiracy theory around that. And then you know the true story is response rates are falling. With falling response rate there's a higher standard of error. That's just a statistical property. Um and then when we get the administrative data where we can benchmark it to uh suddenly we discover that this the the the we were within that very large unfortunately margin of error uh but that the estimate was was off and that the administrative data is less timely. I mean that's the thing there's always this trade-off and that the big revision isn't necessarily a sign of some big conspiracy. It's a sign of the fact that we actually are trying to do it right. Like, you know, similar to, you know, when the newspaper that publishes a lot of corrections, that's actually a sign that they're responsible journalists. That's not a sign of problems. And the news organization that never some some news organizations brag about the fact that they've never published a a correction, well, that tells me they're terrible, you know? uh we make our best estimates and then when we get better data and more information we revise and that's actually something that should increase trust not decrease trust and unfortunately for the public it's often the opposite. So the intermediaries can be part of that education process that a big revision is actually a sign that they're really trying to get it right. It's not a sign that they don't know what they're doing. That's that's just such an excellent point and a fact is that mo most people are not comfortable with uncertainty and with changes and that's of course what happened uh in spades in the covid epidemic. I mean it was called originally the novel corona virus. You know it was nobody knew its properties at the beginning. So there was a lot of back and forth and uh and understandably people were frustrated by that. But as Patrick was saying, uh revisions of economic statistics are necessary uh given that the markets want things fast and the tradeoff is that again there will be there will be revisions. There are a couple of smaller things that I think could help, which is libraries and other groups um could push for clear and adequate citations to source data. Yeah. >> You know, because you'll see in the paper, oh, you know, just well, or it might say it might say Census Bureau, but it would be very helpful if it would say American Community Survey or whatever it was. And the Census Bureau has actually recently put forth some citation guidelines for their stuff. And I don't know whether you all have looked at those or think they're uh on the right track, but that's something where I think educating everybody who cites uh government's data for, you know, how to how to do that. On the other end, I'd love to see you often see now on government websites, this is a official government agency website, something like that, right? And it takes you to a little something. I'd love to see with our decentralized system. This is a uh principal federal statistical agency. And then that would take you to a bunch of things about how we protect confidentiality and how we're objective and role of revisions and so on. Um there in other words some branding that gets all of the major statistical agencies uh with a brand and I'd also like to see that on their when they send out like advanced letters for surveys. Often it comes from the survey organization and there are who's their contractor and they're very good organizations but it really ought to come from the agency and it ought to be we're you know we're an official fiscal agency we're collecting data for the public good. people do respond. The research has shown if you can say these data are going to help your community or they're going to help uh make better policy in a bunch of areas people do respond to that. So some of these things and groups such as again association of public data users goort could be encouraging agencies to um you know not hide their light under a bushel in that way but we uh but to uh try to get a brand identity more than they have now. People know the census bureau they don't necessarily know any other agency uh but there are a lot of important ones um fiscal agencies. Michelle, what else could we as users be doing to support the efforts? Um, I mean, I I I think librarians are are trusted individuals in the community. So I I think just you know librarians in general have facilitate that trust in statistical um data because because you are a trusted source of information right that's that's where you go to find out um um information and understanding sources and primary sources for secondary sources you know so I think um yeah I think librarians and libraries are great places for building trust in communities. Um I think universities are also critical because they help contextualize a lot of this data um and can provide independent verification and accountability through their research. So, um, those are just, you know, two thoughts that weren't already mentioned by Patrick or Connie or >> and and we were all, just let me add, we were all hopefully as are registered voters and when we talk to our officials when we run into somebody on the plane or if we had a town hall, ask about statistical relevancy and and what they're going to be doing to promote trust in the government. Now we have uh there we do have a question from the chat and so uh let's see what we got. Erin, did you have a you want to voice that question? >> Certainly. Um when talking about the reduction of survey responses, are there ways for people to seek out the survey themselves or does someone from the agency have to reach out to each individual? For instance, if someone was aware that the survey was live, could they go to a website or call a number to receive access to the survey? >> That's a that's an interesting question. Um, in general, you want surveys to be representative. You want them to be probability surveys. So that means that uh people are selected from various ways to be in a survey and the survey organization say you know the bureau of labor statistics or somebody wants responses from that population. So that means that uh that you would they would not want to have this sort of opportunity. But for the census where everyone is supposed to be counted, they have tried over the decades to um and they did a lot in 2020 uh to have a procedure where if you didn't get your letter that had your ID code to enter in on the web on the internet or you misplaced it, you could just there was a lot of advertising. Um, it was called nonID. It meant that you might whip out your cell phone and answer the questions and they would include, you know, your your name and address and so on. So, Census Bureau could verify whether you'd been counted already by somebody else in your household. But that's the one thing where they have always been seeking ways to let people know that they they can and should get in contact uh with a phone number or website. But otherwise, for a probability survey, the idea is you've picked a sample and that's what you're those folks are the ones you're going with. And I think Connie to your point earlier about having um Census Bureau agencies state that this survey is coming from the Census Bure have some kind of citation is important because when the household poll survey was going on >> um I had many people ask me is this text message legitimate? >> Yes. >> So it's so it it goes back to what we talked about earlier in the session of Patrick. You mentioned scammers. They're trying to distribute surveys through different modes and mediums, but um when you get a text message and you don't know, is this a legitimate? Am I gonna be taken, you know, to a nefarious link? Um the problem. Yeah, >> I I I agree. I have personally gotten, you know, I try to answer every legitimate survey given what field I'm in and but I've gotten things that are again from the survey organization that doesn't really make it clear enough that they're doing this on behalf of a trustworthy, you know, federal statistical agency and that is just not um yeah, make a lot of sense. We we don't have much time today to to talk about some of the other questions we have for you and this is very fascinating and I'm thrilled to have so many people here. You will be getting a survey about this event and we will be asking what topic and and and go education committee would be happy if we see an outpouring of a question to do something else to put something together but at this point I think we have to go to start wrapping it up. Thank you Jennifer. Thank you Jessica for your moderation of this um this wonderful discussion um to um Patrick to Michelle to Connie thank you so much for the wealth of information that you have shared with us um in today's webinar um again to Anna and to Patty thank you for facilitating the work of this um this group for this um second sunshine week um webinar that we've done as a education committee and of course to our friends with go door um both Julia and to Danielle and to all those attending. Thank you so much for um just facilitating this. This has just been a great discussion. I'm looking forward to to more of these going forward. Um these these webinars particularly for Sunshine Week have just been a a delight to see just the turnout and the appreciation um of information that we are learning of how to engage from a library perspective but also from our partners in other um aspects of this um work and making sure information is promoted. So thank you everyone for joining us. Until next time, thank you for um being with us here at for our go to our webinar series and we look forward to seeing you at our next webinar. Have a good one. >> Thank you. Try to