Behind the Scenes with the FRED Team

Jennifer Beatty:  Welcome. I'm Jenn Beatty and I'm at the Economy Museum at the St. Louis Fed talking with the FRED team. I'm excited to be here today and to learn more about FRED. Who is this little team that runs this big engine called FRED?

Katrina Stierholz: Hi. I'm Katrina Stierholz. I oversee the FRED, Economic Education and the FRASER Digital Library groups, and all three groups work very hard to provide economic and financial data and information to the public, and they work together a lot, too. So there's a nice collaboration that goes on there.

Keith Taylor: My name's Keith Taylor. And I oversee the FRED team, which includes both developing the applications—so FRED, GeoFRED and ALFRED—and then also maintaining all the data in the database.

Yvetta Fortova: Hi, my name is Yvetta Fortova. And I work with the FRED team on organizing, publishing and disseminating data into our FRED platform.

Beatty: So we've learned a little bit about who the FRED team is, but I want to learn a little bit more about what is FRED?

Stierholz: So FRED's an economic database that we provide to the public as a public service. It is a tool that brings economic data together from over 80 sources. We have half a million series. It offers many features. So users can come to the database. They can get data from the Bureau of Labor Statistics or the Bureau of Economic Analysis. They can put data together. They can do transformations of data. They can ask questions of the data and find things out about the economy and learn about the economy using FRED.

Beatty: When you say transformations of data, that sounds fancy. What does that mean?

Stierholz: So transformations of data can be things as simple as, so the CPI, the Consumer Price Index, is an inflation measure. And when it's published, it comes out as an index value, but that's really kind of meaningless to most people. What you can do in FRED is literally with a click of a button change it to something like percent change year over year with a monthly value. That has meaning for people. So they can look at the data and then derive some meaning from it.

Taylor: And really all those tools are there to help people understand the data to make good decisions with the data, which distinguishes us a little bit from other places where you can get a lot of the same data.

Beatty: Why is it important to provide this public service? Why is it important to give people access to data that they can interpret, use, analyze, etc.?

Stierholz: In the modern world, people use data to make decisions. What we're doing is helping people understand their world and make judgments. So when we provide data about the economy or about one's region, people can understand that better. So if you're looking at mortgage rates, you can have a sense of what the mortgage rates are. And we have those in FRED. We have regional data that talks about your local economy, so everything from unemployment at the county level to the percent of people who have subprime credit scores in a particular county, hospital admissions, educational attainment, all kinds of data in FRED that help people kind of understand what's going on and potentially make decisions for themselves.

Taylor: They're going to be consumers who are making decisions like to get a mortgage, businesses making a decision about where they might invest or release a new product, academics using that data to further their research. So, it's a wide audience.

Beatty: I'm curious as to why the St. Louis Fed offers this free database for the world?

Stierholz: It's kind of an interesting thing, because it grew organically out of work that was done in the early 1960s here. We had a research director, Homer Jones, who decided to start sending data out to be an example of what he thought were useful data to look at in making monetary policy. And he says that in the first memo that he puts out. So it was really intended initially to be a public service, to help people understand the economy.

And that public service piece has been a thread that's come through all the way to the present, but of course what has changed is how we deliver that public service. So back then it was paper. It was initially three tables. Now it's a database. It's half a million series. It's got amazing tools and things you can do with that database. You can share it, you can download it, you can do all sorts of great things, but the public service piece is still there. And I think that's a really nice tradition and a great service that we offer to the world.

Beatty: Some would ask why does it matter why the public has free access to data? What is the goal in providing this to the public?

Stierholz: So one thing I'm particularly interested in is data literacy, helping people understand how to use data, how to examine data critically, and make use of it in their life. FRED is really a tool that helps support data literacy. So you provide data, but we also provide ways for people to understand data. And what that does is help people become data informed. Data comfortable, even. Data competent. Where they can look at data not just in FRED, and use it to understand their world at large. So it's great in economic education that they give people these tools that we provide in FRED and lessons or activities that teachers can use. But then when people leave school and they go out in the world, they are able to look at things like credit card offers, and say this is the credit card that's best for my situation. So I think it's terrific in terms of providing people economic data and understanding the economy, but those skills can transfer to many other parts of peoples' lives that actually matter for their well-being.

Taylor: FRED has a wide range of users, from high school students to Nobel Prize winners. Most of them though are novice data users and need a little help to make the most of the FRED tool. And this is really an area where data literacy can help them understand what they can do and ask good questions and be confident that the answers that they are getting from FRED are accurate.

Beatty: Of course FRED has some siblings that you've eluded to, both GeoFRED and ALFRED. Can you tell me a little bit more about them?

Fortova: ALFRED is one of the siblings of FRED, and ALFRED is really archival FRED. And it started back in the days when Bob Rasche was a research director, and he really believed in value of storing series that revise over time. And almost all series we have in FRED do revise, and we set up ALFRED for that reason. So that you can go and access data on a given period in the past and see how the data looked like. So it's not uncommon that we have data revisions, but to some people this kind of data is very important because they can re-evaluate old economic models with the vintage data. And it's also very popular in policymaking.

Taylor: We also have GeoFRED, which allows you to map FRED data at the U.S. level. We have county level, metropolitan statistical area data, state level, and then at the international you can map all the countries. And both GeoFRED and ALFRED are all accessible either as their own domains: geofred.stlouisfed.org or alfred.stlouisfed.org.

Beatty: What about other relatives of FRED?

Stierholz: So there's another tool that we have called FRASER, which is a digital library. Sometimes we call it the cousin of FRED, because it really isn't based on FRED data. Both ALFRED and GeoFRED are really integral parts of FRED, but FRASER is a digital library. It started really as a complement to ALFRED. Over time, it really kind of accelerated during the financial crisis. We realized policy documents were at least as important, and so we started adding policy documents. And we focused for quite a while on financial crises and getting documents in about that. And then also Federal Reserve policy. So now FRASER's a complement to FRED, but it's really a digital library of historic economic and financial information.

Beatty: What's your favorite archive in FRASER?

Stierholz: So my favorite is the Eccles Papers. Marriner Eccles was chair of the Federal Reserve. He oversaw the modern Fed and was really an important person in establishing the Fed as an economic policymaker. Really, that was the beginning of the FOMC—the Federal Open Market Committee—and their decision-making. And he brought that all to fruition.

Jennifer Beatty: So let's dig into a little bit more about how FRED works.

Yvetta Fortova: FRED's magic is very simple. As the originating sources publish their data, we go out, we download the data, we do some quality control checks, and then we upload the data into our FRED database. So really, all time series in FRED live in harmony. And they are displayed on a graph page where they are plotted for over a continuous time. So for example, you can go out on FRED and access gross domestic product, which is available from 1947 to the most recent period.

And with that magic, we actually have five people on our team that really work hard to make sure that data is uploaded as soon as the originating source has published the data sets. In addition, we have three more people who work on our front-end application to make sure that our website and the tools are properly working. But that doesn't end there. FRED is so popular that we have help from inside and outside of the Bank to really contribute to the FRED success.

Beatty: Great. I love your FRED sweatshirt, by the way. So you're taking these credible data sources from all over the world that are publishing statistics. And it's interesting because FRED is this one-stop shop for this data. So tell me a little bit about how that might be helpful for a user.

Fortova: So we recognize that the statistical agencies really do have a mission to disseminate data to users. And we, as FRED, really are sympathizing with that mission as well. So given our collaborations, we are able to build bridges with the respective sources—so, for example, U.S. Census Bureau—to really help us in making sure that users are getting the best data experience to tell their story and understand the data.

Beatty: But what is specific about FRED—either in the data or in the way it's provided—that makes it truly a unique public service for the St. Louis Fed and for the world?

Fortova: Compared to many other countries, U.S. is unique in a sense that there is not a single statistical agency. So each statistical agency specializes in their own data indicators, and they are very good in it. They are the subject matter experts.

What FRED does is putting all the data under one roof. So FRED is a data aggregator. And it's very beneficial for users, because they can then go to FRED and find gross domestic product, which is published by the Bureau of Economic Analysis, and unemployment rate, which is published by Bureau of Labor Statistics, all on one website. But not only that, they can actually put it all on one graph. And they can download the graph, download the data, and pretty much use the data to their lives.

Katrina Stierholz: I would also say one of the other things that's unique is that we have this Economic Education group that helps provide resources. So when you go to a particular series page like unemployment, you can find articles and economic education courses and videos that help explain unemployment and how it's calculated. So these supplemental materials that are unique to FRED that add value and I think provide more for the user, for the public, to understand the data.

Beatty: What’s the audience? Who are you targeting for FRED?

Keith Taylor: So in addition to instructors and students, we also have quite a few people that are using it to make business decisions. So financial professionals, analysts at corporations, things like that. We also have academic style researchers, and they're using it to further their research. And then finally we have a lot of journalists and bloggers who are talking about economics and they're using it to explain that to the public. One of the things that makes it really easy for them to do is you can just grab an image off of our website, tweet to it, put it in your blog, which makes it really easy.

So another project we have is FRED Blog—which are articles written twice weekly—and they dive into some economic topic and then also how you can demonstrate that topic in FRED. And they're a really nice complement to the other products we have with Econ Ed and FREDcast.

FRED Blog targets a broad audience. The articles are often about things that are happening in the news, so they've very topical. Frequently, they'll also get picked up by journalists as a basis for an article.

Beatty: Okay. Any famous people using FRED?

Taylor: Yes. Of course. So I think probably our most famous user is Paul Krugman—who won the Nobel Prize in economics—and he writes a blog for the New York Times.

Stierholz: I would say in addition to Paul Krugman, we also have textbooks that use FRED. So regularly if you have an economics textbook, you will probably be assigned homework that uses FRED. And I think there's sort of a credibility that goes with that and just a validation of our usefulness, which I find terrific.

Taylor: And I would also add, we look at Twitter. A lot of people who want to talk about economics. I'm heartened to see that they're using FRED data to have this discussion. So you can go out on Twitter and you see a lot of FRED graphs, people linking to our website.

Beatty: Our own Twitter feed @stlouisfed is a master user of all things FRED.

Stierholz: And that's another way that people educate themselves. So if they follow St. Louis Fed, that's a way people can kind of keep track and follow what's going on in this economic realm and have data that helps them understand it, which—that's a twofer.

Beatty: There's a big effort at the St. Louis Fed on economic education and economic literacy. And tell me a little bit about how FRED interacts with that and fits in with efforts that you're making to try and reach out to students and teachers.

Stierholz: First of all, our Economic Education group is amazing, and they do great work. They just won an award from Central Banking. It's an international award on financial inclusion. So Economic Education has worked with FRED to do things like provide lessons and activities.

One that comes to mind is something we call FRED Interactives where a teacher can assign homework, and then what happens is the student will perform things in FRED, will be assigned a specific thing to do. And it walks them through how to use FRED, but it also adds economic concepts to it.

And so one of the things that's terrific about economics is it is very data informed. Data really is important, and it is a way of judging economic arguments and understanding economic concepts. So this FRED Interactive on comparative advantage, it gets data from FRED. It also explains things about comparative advantage. And so there's both learning about how to use a database and how to do particular things in a database, but also about an economic concept. So that's sort of one way that they work together.

I think of it as chocolate and peanut butter. They go together really well. They complement each other. They're great individually, but they're also fantastic together.

Taylor: When people study economics or they learn economics, especially students, it can be an abstract thing that they're learning. And data can make that concrete and can make it more real for students. The other thing is a lot of people who take economics or learn economics, they don't go on to become economists, but they are going to go on to probably be using data throughout their lives. And so this combination of Econ Lowdown and FRED allow them to have a foundation of using data to make decisions and to understand that through sort of improved data literacy, which is a skill that they can definitely take with them.

Beatty: In their home life, in their work life, throughout their life.

Stierholz: Yeah. Getting a mortgage is one thing, of course, but also when there are different claims made about policies, you can look at the data. And being able to critically look at data and understand what's in the data, and who produced it, and look at a policy option I think is really important for citizens of the country. So I think it's a really important skill for anyone.

Taylor: The other chart that all high school and college nongraduate students should see is of course the salaries by educational attainment. Lifetime earnings. That's a classic one that really drives home the point of data, and that's something you can get in FRED and talk about.

Stierholz: And unemployment rate by educational attainment.

Beatty: So tell me more about FRED's version of the Braggin' Rights.

Taylor: For those of you who might not be familiar, the Braggin' Rights, it's a basketball game between University of Missouri, Mizzou, and University of Illinois. Takes place always in St. Louis right around Christmas.

We have a product called FREDcast, which allows people to forecast economic variables. The unemployment rate, the jobs number, GDP or gross domestic product growth rate, and inflation. And basically you can create a league and everyone forecasts his values, and the people who are the best have the highest scores.

So students of the University of Illinois and students at the University of Missouri compete with one another. They just finished up their second year. U of I has so far swept the standings, but the second year was very close. And so I think that they just started a new one, so we'll see how it turns out.

Stierholz: So this is the part where I get to say that I went to University of Illinois.

Taylor: And unfortunately, the person who runs the league went to the University of Missouri.

Jennifer Beatty: I want to dig into some of the most popular series that are in FRED. What's the most liked or most often used series in FRED?

Yvetta Fortova: So by far, consumer price index—which is a measure of inflation, and it's published by the Bureau of Labor Statistics—is one of the most popular series in FRED. But we also have many others, such as gross domestic product, which is the amount of goods and services produced in an economy in a given period of time. And that one comes from the Bureau of Economic Analysis. But it's interesting to see our popularity changes over time based on what users are really looking for. So recently we have been observing a lot of yield curve series gaining on popularity. Especially the spread between the short term and long term Treasuries.

Beatty: So FRED is connected to what's going on in the news or in the economy at any given time?

Fortova: That's correct.

Katrina Stierholz: During the financial crisis, the weekly unemployment claims data that came out was the most popular. As unemployment went up and as people were very concerned about what was going on with the economy, that series just became incredibly popular and was the most popular series for quite a while. And I think that reflects peoples' concerns about the economy, what's going on. FRED actually answers peoples' immediate questions about the economy.

Fortova: And one should not forget about times when Greece has been in really financial trouble. So our series on Greek debt has really gained up in popularity and made it into top 10.

Beatty: In addition to the most popular series, what are some of the strangest or most unusual series you might find in FRED?

Fortova: We have many unusual series in FRED. I think one of the most notable ones is our extensive data set on a historical series, which is published by National Bureau of Economic Research. And in that data set, you can find series on shipments of sinks and toilets to the U.S. during the industrialization of indoor plumbing. And you can really see the correlation of the two items during that time.

Moreover, we have a series published by the Bank of England that is for population of England. And that one goes all the way back to 1086. Who would have known data existed back then? And that marks a year of the “Domesday Book.”

Additionally, we also have a quite extensive group of series that cover socioeconomic indicators on a regional basis. So users can go and compare regions for commute time, disconnected youth, burdened households and educational attainment. So there is a lot of interesting indicators that users would find. So the chances are, you go on FRED, you will learn something new.

Beatty: What's really interesting about FRED is that it's not just macroeconomic data series. You can get down to the county level even, and a host of others. What's your most favorite series, your own personal favorite series on FRED?

Fortova: That's a very hard question because it's sort of like, “Do you like Mom or Dad better?” But over time, I pick my favorites. And usually those are series that in some way exhibit some kind of a milestone or interesting fact. So for example, recently industrial production has really gained my attention because that one just recently obtained 100 years of full data.

Keith Taylor: Of course, my favorite series, my favorite graph in FRED, is the toilets and sinks graph that Yvetta eluded to. It just makes you feel so good. You see an almost straight line with a slope of one, which just means every time someone was buying a toilet, they were buying a sink, which just makes me feel good about-

Beatty: Personal hygiene

Stierholz: My current favorite, like Yvetta, they change. I think right now is the percent of—it's a bi-county data, so it's county level data—percent of the population with a subprime credit score. And that's one you can map in GeoFRED. And when you map it—so if you saw this as a list of counties, it would really not tell you much. It would be alphabetical. That's not helpful. When you map it, you can see what's going on across the country. You can see particular areas that have many more families with subprime credit scores. It kind of tells a story. And it was really surprising when I mapped it, and I was like, “Oh, my gosh. That's incredible.” So that's my current favorite.

One thing I like is looking at ALFRED, which is one of the siblings of FRED. And I like looking at data revisions. It is interesting to see how much data can revise, and how the initial understanding of what was happening in the economy can be substantially different than the data that is revised that comes out. The data revisions can be pretty striking at times, particularly when things are happening like the financial crisis.

Beatty: So how do you decide the data series that get put into FRED ultimately?

Fortova: In reality, the FRED team is small. And our bucket list of data sets that we wish to add to FRED is fairly large. And when we try to decide what kind of data set to be added to FRED, we look at many different aspects. We're looking on what kind of institution or organization the data is coming from. Is that a reliable data set? Is that a data set that we do not have in FRED, and how can that benefit the users? So we really are making careful decisions on which is going to be the next data set that will hit the FRED mark.

Jennifer Beatty: So remind me again, how old is FRED? How long has it been around?

Katrina Stierholz: It started in 1991 as an electronic bulletin board. And when it started I think there were about 30 series. I know there was a newsletter that went out that said there were 620 people that accessed it.

In 1995, we moved to the worldwide web. And in the early 2000's, we developed it as more of a charting tool on the internet. And now we have half a million series and 87 sources. I guess the short answer is FRED is 28 years old.

Yvetta Fortova: It just got out of college pretty much and it's entering its productive years in the real world. So it has many years to come.

Keith Taylor: Maybe grad school. It just finished grad school.

Stierholz: I would say its world's gotten bigger. So if you think about it, we started with just national data and just a few series, and over time, we've added regional data and international data. So it's kind of like FRED has traveled the world, gone to other places, lived in a different city.

Beatty: It sounds like FRED has done well for itself. Where will it go next?

Taylor: I would say, in general, we've got the U.S. macroeconomy pretty solid with our data sets. There's some gaps there that we could definitely fill in. But I think the future data acquisitions will be international and probably more socioeconomic, less traditional macroeconomics.

Stierholz: I think one thing that would surprise people is a third of the people who come to FRED are international. They're not from the United States. So we serve a public that is larger than the United States. We serve the world.

Beatty: I want to spend a few moments talking about where is FRED going in the future. What do you see FRED growing to in the future? And how do you see people using FRED?

Taylor: In the immediate future, we're really focused on helping people find economic information or data sets, and then once they've found it to be able to quickly understand it and come up to speed on it. And then of course we're always adding more data to FRED. So that's just kind of an ongoing continuous process.

Longer term, we really want to focus on helping people understand the data. So either pulling in additional content or working more closely with the providers of the data to make sure that we're presenting it in the best way possible, that users have the best experience.

Stierholz: I consider this year, 2019, the year of the user for FRED. We really want to focus on the user experience. Perhaps if there's a busy part of the page, we dial that down. We test and make sure that people can use the page easily, can understand how to make transformations and do some of the sophisticated things you can do in FRED.

Taylor: So there are some things that we could automatically do for the user in terms of harmonizing units, so it would be just a step easier for them. I think also then starting to integrate our other data products. So you could imagine looking at a FRED graph and having a policy paper that is talking about that or an ALFRED graph all on the same page so that you have a richer experience, a better understanding of economics, and it's not just the data in isolation.

Fortova: There's always chances for improvement. So our plans are to devote some time to really look at different search possibilities and see if we would be able to provide user with yet a better experience on how to find series in FRED. And with that, we will be trying to look into ways how we can link our FRED products together.

So perhaps if you're looking on a regional indicator, you will immediately be able to see that data is not only available in a graphical form, but you can access it on a map, too, to see cross-sectional comparisons.

So ideally, we would really like world dominance. In the same aspect like, I don't think people can really imagine their lives without cellphones. We would like them to not imagine their lives without FRED.

Stierholz: I like that.