This episode of CaseyCast covers all things data with veteran demographer Linda Jacobsen. As vice president of U.S. programs at the Population Reference Bureau, Jacobsen oversees the development and analysis of child well-being statistics included in the KIDS COUNT Data Book.
The Annie E. Casey Foundation has produced the Data Book annually since 1989, and it has grown into a leading source of trusted, nonpartisan data that depicts how America’s families and children are faring.
Listen as the Foundation’s Lisa Hamilton talks with Jacobsen about a wide range of statistics-themed topics, including the importance of disaggregating data by race, the limitations of current surveys and the dangers of agenda-driven information sources.
CaseyCast is a podcast produced by the Annie E. Casey Foundation and hosted by its executive vice president and chief program officer, Lisa Hamilton. Each episode features Hamilton talking with a new expert about how we can build a brighter future for kids, families and communities.
We sure hope so! Go to Apple Podcasts to subscribe to the series or leave a rating or review.
The Casey Foundation released its 2017 KIDS COUNT Data Book, an annual ranking of states on child well-being. The Data Book is an important resource for lawmakers and child advocates who depend on reliable data to advance sound policies that benefit children and families. So, who does the Casey Foundation rely on to analyze the Data Book's data? It's the Population Reference Bureau.
Therefore, I'm delighted to welcome Linda Jacobsen, the vice president of U.S. Programs at PRB, to discuss the state of data on kids and families today. Linda is a demographer with more than 30 years of experience analyzing population trends and their implications. Her research is focused on family and household demography, as well as poverty and inequality.
Welcome, Linda. We're delighted to have you.
Thank you, Lisa. I'm very happy to be with you today.
Great. So, with PRB as a partner, KIDS COUNT uses data to tell the story about child well-being in the United States. Can you tell me where your information comes from?
Well, we're fortunate, Lisa, in the United States to have comprehensive data on a wide variety of child and family well-being indicators from a number of very important government data sources. These include The American Community Survey, data from The National Center for Education Statistics, vital statistics data from The National Center for Health Statistics, and data from The National Survey on Drug Use and Health. But, even though we have wonderful data sources, and they are where we get the 16 indicators that are currently included in the KIDS COUNT Data Book, and in the index, there are still some additional data that we need in order to fully understand or assess children's well-being today.
So. What are the data gaps when it comes to understanding where kids are today and what they need?
One of the gaps that we have is good data on juvenile justice. For example, the best estimate of the number of children and teens who are arrested comes from something known as The FBI Uniform Crime Report. But these data are submitted voluntarily, so they're not available for all states or jurisdictions, and, also importantly, they're often are not provided separately by race and ethnicity, which is very important. Similarly, there are no sources of data that provide estimates of the number of children who've ever been incarcerated, or committed, or detained.
One other area I would mention, that I think is an important data gap, is data on homelessness for children. The data we do have is limited and, sometimes, it's not a very high quality, and we know there's been an increase in homeless families with children, and this was especially true during the Great Recession a few years ago. But, we don't have complete or comparable data across all states.
So, right now, state education agencies, or local education agencies are required to submit information on children who are homeless and who are enrolled in schools. Schools are the only institution legally responsible for identifying and serving children who are experiencing homelessness. But, the school-based approach is a problem because it can, for example, miss homeless children who are younger than school-age, right? So, it's not counting children who are not in school. It also can miss older age homeless children who've dropped out of school as teens. In some cases, the counts may not be accurate because it might double count students if they enroll in more than one school district in a year, and so often times homeless families will move between districts.
They may get temporary housing in one district versus another so that may impact the accuracy of the count. Then, of course, having accurate data requires homeless children to be accurately identified by school personnel, which may not happen in all cases.
One other area I would mention that I think is very important, is understanding the well-being of children in what we usually refer to as mixed-status families. So, these are families in which one or more parents are undocumented, but the child, him, or herself, has actually been born in the United States, and so is a citizen.
But most data sources, and most surveys, including the American community survey, don't ask about legal status. So we have very little information about these children and youth, and obviously that's problematic because they're at risk of being separated from their undocumented parents.
The final important data gap I would mention is that we need more information about health of children below the national level. So, I've mentioned, the ACF as a great resource for socioeconomic data, but unfortunately it lacks detailed information about children's' health. And there are very few national surveys that are large enough to give us reliable sub-national estimates that are updated on an annual basis, which is of course very important for the KIDS COUNT Data Book, which tracks the status of children and changes in their well-being from one year to the next.
Right now, this is especially concerning, because with the proposed potential changes to health care, it's going to be very important for us to be able to track the impact of any of those changes on children's health in state and local areas.
Well, thank you. Those are certainly important data gaps that we need to do our best to see how much information we can get about them.
I'd like to go back to the first that you mentioned, around juvenile justice, which seems to relate to the ways that different state agencies collect data. And I'm sure this applies not to just juvenile justice systems, but probably also education or child welfare systems. Could you say a little about the ways that we depend on state and local agencies to gather data, and what role, if any, the federal government might have in helping us get more uniform data across the states?
I think that's a really important, and a very challenging issue. We are very fortunate in the U.S. to have surveys like the American community survey, which are designed, collected, and data are provided by the U.S. Census Bureau. So by definition, we have comparable questions and comparable collection procedures, and tabulation procedures that ensure that those data are comparable.
But there are other areas, such as the juvenile justice, which I described, where the collection of data is really more for what we would refer to as administrative purposes, and that's for carrying out policies and programs on the state and local level.
So, these are systems that have developed, in many cases, independently in states over many years, and a lot of those procedures and approaches are very institutionalized within different states, and so it's very hard to provide incentives for this wide range of different state and local agencies to change what they're doing, and basically what it requires is resources.
A good example of the way this has been accomplished successfully, is collection of information from birth and death certificates, which are, in fact, collected and maintained individually within each of the states, but the states then compile their information and send it to the National Center for Health Statistics, which puts it all together then and provides us with important data that we use in the Data Book, such as the teen pregnancy rate, low-birthweight babies, and child and teen death rates.
But the way that we are able to get comparable data is that the National Center for Health Statistics has designed what they call a standard birth certificate, and a standard death certificate. And once the states were able to adopt these, and again, it took resources provided by the federal government for them to be able to do this, then they were able to standardize.
But, I would say, the big challenge has to do with the lack of adequate funding or incentive for all these different state and local partners to be able to standardize what they do.
Great. At the Casey Foundation — and in the KID COUNT Data Book — we make it a priority to disaggregate data by race. And we've been hearing a lot over the last several years about the changing demographics in the U.S. Could you talk a bit about what you see in terms of the changing racial composition of, or racial demographics of children in the U.S., and what it looks like, or what we expect may happen over the next several decades.
The most basic fact and the place to start, is that according to Census Bureau projections, by the year 2020, 50% of all children less than 18 in the U.S. will be members of a racial ethnic minority.
I think the important thing to understand is that children of color are the fastest growing group within the overall child population, and there is no reason to expect that to change substantially in the future, and because historically, and, it's been reflected in every Data Book thus far, children of color have lower levels of well-being in general, and, in particular, with specific indicators than non-Hispanic white children.
So, the ability to disaggregate data by race and ethnicity is going to continue to be, if not become, even more important than it's been in the past. And we obviously can't understand the challenges that children of color are facing if we can't disaggregate data, so we can't measure racial inequity, and we can't monitor improvement over time if we don't have data that's disaggregated by race and ethnicity.
When we look at the data, we certainly see disparities for outcomes for children of color, but we also see differences in how different regions of the country are faring: That the south and the southwest tend to have lower levels of child well-being than other parts of the country.
The simple observation is that where you live matters.
States in the South and Southwest have consistently been in the bottom quartile, or amongst the bottom 10 states in the KIDS COUNT Data Book overall index ranking, so that's looking across all four of the different domains from economic well-being to education to health to family and community.
I think the important context for answering this question is for listeners to understand that research has shown that child well-being is influenced not only by conditions within a specific household, but also by conditions in the neighborhood, in the schools, in the city, the county, and the state in which children live.
It's that overall context, not just the measures of what's going on in a particular household, that really influence overall child well-being.
What do the data tell us about growing up in these areas?
What we can tell from looking at these data over time for the South and Southwest, is that the level of resources and state and local policies tend to resolve in children in these states having fewer resources than children in other states. And as a result, they tend to have lower levels of well-being in general.
If we look at parent and family resources that are available to children living in the South and Southwest, what we see is that in these states, there are relatively high rates of household heads who are high school drop outs, and low rates of household heads who have a bachelor's degree or more. So there are fewer educational resources that are available among parents of children who live in these states.
And then also, median family income is much lower in the South and Southwest states. So there are fewer economic resources available within the household.
But I also mentioned that neighborhoods and schools are important, not just what's going on in the household. What we also find is that in Mississippi, New Mexico and Arizona, those three states, have the largest share of children who are living in high-poverty neighborhoods. And high-poverty neighborhoods mean that there is concentrated disadvantage. So it's not only a child living in a household themselves that is below the poverty level, but that a large share of households within that child's community are living below poverty.
Southern states also fall to the bottom of the list when it comes to education outcomes for children, according to the KIDS COUNT Data Book rankings. Can you tell us what the data says about educating kids in the South?
In terms of educational resources in the schools, per-pupil expenditures tend to be the lowest in the South and the Southwest. So they're in the bottom half of expenditures when we look at the country as a whole.
We see the lack of resources in these states reflected in an important way in some key outcomes. So, with respect to education, we see that the 10 states with the highest share of eighth graders who are not proficient in mathematics are in South or southwestern states. These states also have relatively large shares of fourth graders who are not proficient in reading.
I think it's important to round this out by observing that this concentration of fewer resources and worse outcomes in the South and Southwest is problematic, because it creates a self-perpetuating cycle. So fewer resources, and lower levels of well-being in childhood then tend to translate into fewer resources and lower well-being in adulthood, and then these patterns are passed on to the next generation. That's what we have seen going on for a number of years in the South and Southwest, why we see this consistent lower levels of well-being than we see in other states.
I would just conclude that there's clearly room for improvement in terms of access to resources and levels of resources in the south and southwest, and we know that investing in children and families makes a difference, because we see that difference reflected in higher levels of well-being in the other states.
I have a question related to policy changes that are designed to help improve outcomes, and it's a question about the timeliness of data. Often when we put out the Data Book, we are looking at data that might be several years prior to the year of the Data Book. What can you tell us about the timeliness of data and how it could affect the policy solutions that we seek to develop?
That's an excellent point, and I think it really underscores the importance of advocates and those with whom they're communicating… Paying attention to the data, but the timeliness of data is really key when you have large disruptions in terms of policy changes or such, as occurred with the great recession back in 2007 to 2009.
So the real key is having data that are collected on an annual basis, because that gives us the ability as researchers and as advocates to assess the impact of what has happened. And of course, in some cases, impacts are not going to show up in the next year. They may not show up for a year or two.
So just because your data are only a year out or two years out from an event, it may take a while for impacts to show up. I think it's important to always be paying attention to the lag between when certain policy changes have been implemented and when your data measurements are, but obviously the closer in time we can get data collected to when events occur, the better off we are.
And the American Community Survey is a wonderful example of the value and importance of the federal government collecting data in every county in the U.S. every single year. Before we had the American Community Survey, we used to only have the long form data that was collected once a decade with the decennial census. If we had not had the American Community Survey fully implemented in 2006, and we'd only had data from the 2000 Census and the 2020 Census, we wouldn't have had any information about what happened between 2007 and 2009. So, we would have really missed the opportunity to look at the impact on children and families and as I'm sure you know, the poverty rate among children jumped way up during the recession years, and then started to come back down.
But if you don't have data on an annual basis, you're going to miss those year to year changes, and you might make incorrect conclusions about how children are doing, or what the status of well-being is among children and families.
You mentioned the difference between what the ACS provides and the decennial census, and we actually are at this time preparing for the 2020 Census. Can you tell us a little about what the preparations are that are needed, or any changes we're anticipating in the 2020 Census, and maybe even some of the challenges there are when it comes to trying to capture information about children and families?
That's also a great question, Lisa.
I think one of the most important challenges for the 2020 Census is reversing the undercount of children, which has been documented to have been taking place for at least the last four censuses going back to 1980. So, what we know, or what research has shown, is that children under age five are the ones who are most likely to be missed or not to be counted in the census.
For example, in the 1980 Census, research indicates that about 2% of children under 5 were missed in the census, but by the time with got to 2010, this figure jumped up to almost 5%. So, we're missing 5% of children under age 5.
The Census Bureau and others have been conducting research that shows, in terms of trying to understand what are the characteristics of these children who are missed, and what they found is that the children who are missed are more likely to live in lower resourced homes, meaning homes with less income and lower levels of education among parents, and they're also more likely to live in households with complex family relationships.
In particular, the children who are missed are more likely to be non-relatives of the household head, or to be a relative other than a child. So they could be a niece, or a nephew, or a grandchild.
Then also, there are children we know who split their time between multiple households, so they don't have just one usual residence, and they may spend a couple of days a week in grandma's house, they may spend some with mom, and they may spend some with an aunt depending on the adults' work schedules and availability. And those are children who are very likely to be missed. This is important because these are the children with lower levels of well-being, so we don't want to be missing them.
In terms of the current climate, there's also growing concern that undocumented immigrant families with children may be afraid to complete the census, and this could also lead to an increase in the under count of some of the most vulnerable children that we have in the U.S.
Finally, I would just mention that one of the challenges that we have every decade in the census, and it is no less true with the 2020 Census, is measuring race. Race is really a social construct, and our concepts of race and the way that people identify their race and ethnicity is always changing. So, while changing census questions that capture race and ethnicity can improve the measurement and make it fit better with how people actually see themselves, it also can make it harder to make comparisons between racial and ethnic groups over time.
Well it certainly seems like important issues that we need to pay attention to.
The last question I have relates to lots of conversation in the media about what's accurate and what's not accurate. What is fake news. I'm curious, given the work that you do around data, how could you help the average person trust the data that they're getting?
That is a tough one, Lisa, and that's a really important challenge, but I think the basic answer is that consumers need to know and evaluate the sources of data that they're seeing. For example, they need to determine if data are from a reputable government source. That means that they need to look for specific citations for figures or graphs that they see in something that's in print or online.
Equally important, they need to evaluate the organization that's providing the data. Is it a nonpartisan organization, or is it an organization that has a particular well known political point of view? This is important because otherwise very accurate data can always be selectively chosen and presented out of context to further political viewpoint. This is one of the things that makes the KIDS COUNT Data Book so important, and that's the fact that it is a trusted, known brand that's been around for more than 25 years.
In addition, the Casey Foundation's focus on data based advocacy sets it apart from many other advocacy organizations, and certainly from partisan organizations. So, people can trust the Data Book, because it's focused on reporting the numbers, whether they're going up or down. It's not just about good news, or just about bad news, it's about the actual news.
So, I think in what we do at our work in PRB, we very much model what we're doing after what the Casey Foundation does with the KIDS COUNT Data Book, and that is that we make every effort to document and provide citations for all of the data sources that we use, so that people know exactly where a figure comes from, and we don't provide data or information without indicating what the source is. And we also provide links to that source, so that the consumers who are reading something that we publish either online or in print can go directly to that source and they can double check what we've provided themselves if they need to.
But I think that it requires more effort, and part of what we do at PRB is try to help educate consumers about a number of these important data sources like the American Community Survey, so that we try to spread the word about the data sources that are reliable, and promote efforts of organizations such as the Casey Foundation to accurately and reliably provide data from these reputable government sources.
Well thank you so much Linda for your partnership, and the partnership of PRB. You have certainly been at the Foundation of ensuring that the KIDS COUNT Data Book is a powerful resource for those who care about the well-being of children in this country.
It's a very important and valued partnership for us as well.
Well we sure appreciate you joining us today, and I want to thank our listeners for joining us as well. If you've enjoyed today's conversation, rate our show on Apple Podcast to help others find us. You can also ask questions and leave us feedback on Twitter using the Casey cast hashtag. To learn more about Casey, and find notes for today's show, visit us online at aecf.org/podcast and follow the Casey Foundation on Twitter at @aecfnews.
Until next time, I wish all of America's kids, and all of you, a bright future.