Transcript Advanced FRED - St. Louis Fed

Advanced FRED:
Doing more with the data
Katrina Stierholz
June 7, 2010
Using the scenarios…

You’ve learned how to
◦
◦
◦
◦
◦

Add a series
Change units
Set dates
View and download data
Save as pdf, capture the link
Now we’ll go deeper
◦
◦
◦
◦
◦
Review/repeat some from first session
Find data series with different data types
Transform series
Set up user accounts
And, practice working with FRED
FRED data
One nifty feature: if you are ever unsure
of the measure to use for data, this FRED
At a Glance will give you the generally
accepted way of presenting data. E.g., for
CPI, it indicates that the measure is
% change from year ago. That can come in
handy at times.
http://research.stlouisfed.org
On FRED, users can:
Search for data
by category
Check updates
Regional data has
its own category
Discover popular series
Practice FRED



I want to understand inflation a little better
Many economists use CPI (Consumer Price
Index) to evaluate inflation
There are two ways of looking at inflation:
◦ Headline (that’s everything)
◦ Core (takes out food and energy)


So, let’s use FRED to compare the two
Please do this on your computer with me
FYI—PPI is the
price index for
wholesale prices.
This is the “headline”
CPI graph, as an index.
To make changes,
click on “Edit” Graph
To put the “core CPI” on
this graph, we add a
data series
You can either type
in significant words,
or use the Browse
feature to find data.
We have both index
numbers on the graph,
but I’d like to see the
percent inflation, year
over year.
This is where we change the
units displayed to percent
change from a year ago.
Must be changed on both lines,
and then click on Redraw Graph
Core CPI and
Headline CPI.
You can see how
much more volatile
the Headline CPI is,
compared to the
Core CPI (but that
they also move in
tandem).
This one on your own

First, find Real GDP
◦ What is GDP, you ask? The most commonly used
measure of our national economy
◦ What are the elements that measure GDP?
 personal consumption expenditures;
 private domestic investment (investment by companies
and households in big things, like factories or houses);
 net exports (exports minus imports);
 government spending
◦ What’s “Real” about it? (See Economic Snapshot
from Inside the Vault).
The news
release from
BEA
Once you’ve found real GDP…
We’ll look at Real personal consumption
expenditures and add it to the graph
 Then we’ll do a little work to show how
to put them on the same line, one as a
percent of the other.

Using the data
category option
Basic options for graphs
Information about the data
Clicking on any of the
options here will bring
up the graph in its
edit form (so you
don’t have to just pick
“Edit”)
Once you’ve found real GDP…
Add Real personal consumption
expenditures
 Click on “Add a Data Series”

◦ Add it to Line 1

Then we’ll do a little work to show how
to put them on the same line, one as a
percent of the other.
Clearly, personal consumption
is a large part of our GDP.
Hmmm… how large?
Here’s where we do a
transformation
First, we actually have to remove the
consumption line
 Then, we add the consumption data to
the gdp line (I’ll show you)
 Then, we do the math.

Now that we know
about it, we need to remove
it as a separate line and put it
on the same line as GDP.
Click on Add Data Series, and
Choose to put it on Line 1. Once you
indicate that, it will only allow
a series with the same frequency
Series choices are now limited
to quarterly series.
FRED immediately
bounces you down
to the bottom, so you
can fill in the formula,
which is based on the
“a” and “b” designation
above.
So, I’m dividing real PCE/GDP
to show the percent of our GDP
that consists of consumption.
Real Personal
Consumption Expenditures
as a percent of Real GDP
More transformations
What percent of the population that is
over 16 is employed?
 What’s the percent of unemployed
persons who’ve been unemployed for a
long time?

Basic payroll
employment
data
Some changes to
help improve
readability
Percent of our civilian
>16 population that is
employed.
To save your
work, use the
“save” feature
on nearly any
screen. This
will save what
you’ve done so
you only have to
do it once.
Now, a transformation to look
at the long-term unemployed
I have peeked, and know that the
series we want are in the household
survey section. We could have used
the search feature… but this works, too
This one will require adding together
all of the unemployed, and then figuring out
the proportion over 27 weeks.
I’m starting with the most
recently unemployed. I think
that will be easier for me to
add the others in order
(I can be forgetful).
To get started, I click Edit
And, then I just kept
adding the series for
the various weeks
unemployed to Line 1
Now I have them all, so I
need to scroll down to do
the transformation.
So, we can now
see how many
long term unemployed
there are as a percent
of all unemployed
(over 40%, and way
more than we’ve seen
in past recessions).
Data download
This is the
information,
downloaded from
the FRED page
 Notice that I have
downloaded the data
in the graph (not the
original data, but the
transformation)

One issue: you can’t have different
time periods or frequencies



Great article, great
idea
But… even though
we have annual
GDP and federal
debt data, the
federal debt data is
fiscal year and GDP
is calendar year
These would need
to be downloaded
in Excel and then
transformed there
Powerpoint

To capture screen shots
◦ For now, use <alt> print screen
◦ Good news! According to web reports, there
will be a screen capture for powerpoint and
MS Word in their 2010 versions.
Practice more transformations

GDP per capita (use nominal GDP and
divide by population)
◦ Compare that to some states’ GDP per capita
(those numbers are already in FRED).

Percent of population in the labor force
that has been unemployed more than 26
weeks (UEMP27OV / CLF16OV)
◦ You could add some other durations (so
compare that to the percent that have been
unemployed less than 5 weeks)
Keep on…
We’d like you to continue practicing with FRED
First, if you can (have access to your email), set up a
user account and save your work.
 Then, here are some possibilities (if you are looking
for ideas)


◦ Compare unemployment across a couple (or more)
regions
◦ Compare short-term unemployed (less than 5 weeks) to
long-term unemployed (more than 27 weeks)
◦ Compare overall retail sales with e-commerce sales (this
can’t be transformed… they are two different frequencies)
◦ Or, check out the St. Louis Fed Stress Index—a new
measure of the stress on financial markets.