AS THE ECONOMY TURNS
MEASURING THE MOUNTAINS AND THE MOLEHILLS
Dr. William Shingleton
May 22, 2006
In ECONOMICS, the study of employment, inflation, and growth is
called MACROECONOMICS; it is the study of economic SYSTEMS, of how the pieces
fit together. [NOTE: The other set to the bookends is called MICROECONOMICS,
the study of the economic decisions of individuals.] We have always thought
that one of the most remarkable features of MACROECONOMICS is that so many
people so fiercely defend so many different opinions, even when they have
little or no actual knowledge of the subject.
Unlike our dear readers, members of the general public believe certain
ideas are true because they think they should be true or because they want them
to be true, not because of anything more substantial than anecdotal evidence. Of
course, economists, who supposedly have such well-developed knowledge of the
subject matter, will fiercely defend even more contradictory opinions because we
think they should be true or because we want them to be true, but that’s a sad,
sad story for another day. MACROECONOMICS
is the dimension of economics where never have so many said so much while
knowing so little, with apologies to Mr. Churchill for the ineloquent paraphrase.
What is the first step to take if one wants to be able either to evaluate
the health of our economy or to understand the way our economic systems
function? We would argue that, in order
to be able to engage in an informed discussion of either subject, you should
begin by acquainting yourself with some of the basic statistical measures,
their meanings, and their shortfalls.
Once you have done that, you’ll be well on your way to deciding for
yourself whether the system is working well or not, even if it takes a good
deal more to actually develop your own theory of how the pieces fit together. Moreover, since the array of economic reports
that we can get on a periodic basis is so broad, unless you really want to waste
time on these myriad (and often contradictory) reports, what you want to do is
to try to focus on a few that might mean something to a normal person. That’s
why we would suggest that there are only a handful of these reports that are actually
worth bothering with, although the relative importance of the various reports
can change over time as the texture of the economy itself changes. We’ll pick three
for you to keep an eye on in just a moment.
In order to decide which ones you want to watch for, you probably
want to consider the four dimensions or each report. First, you want to consider the FREQUENCY, how
often does it come out? We doubt that
you want to check online every morning for the latest information on the
FEDERAL FUNDS RATE (the short-term inter-bank interest rate). It’s really up to you, but for most people
once each month is usually more than enough. Second, you want to watch
something that has some IMPORTANCE, you probably don’t care about the FUTURES
PRICE of wheat for August delivery, unless your personal occupation or
investments are somehow related to the adventures of the wheat market. Third, you probably want to stick to
something that is readily MEASURABLE. As important as the DISTRIBUTION OF
INCOME might seem, it is actually measured in a number of different ways and leads
to a number of contradictory conclusions. Finally, unless you are a bit of a
geek (Or is it a nerd?), you don’t want to bother with anything that is overly
technical. Ideally, you probably even want something written in some
approximation of understandable English. [NOTE: It is not true that most
economists are terrified of the Senate’s proposal to require the ability to
read and to write English as a condition of citizenship. Rumors like that are
started by mean-spirited sociologists.]
If we accept the once a month standard for frequency, we still
have a ton of reports to choose from, although we have already trimmed the number
of choices by close to ninety-seven point three percent, or thereabouts (Just
kidding, we think!). The importance
factor is a little more difficult because, to some extent, it is a matter of
taste. Most people want to know
something about the LABOR MARKET because the labor market determines the number
of jobs, the UNEMPLOYMENT RATE, and the level of wages. Here, the Bureau of
Labor Statistics (BLS) puts out the comprehensive “Employment Situation Report”,
which comes out on the first Friday of each month and is available at <http://www.bls.gov/news.release/empsit.nr0.htm>. Then, since every sane person in the
We have a couple of general points to consider before we look at
any of these reports. First, they are always
assembled by the professional, rather than the political, staffs at the BLS and
the BEA. [NOTE: The BLS is part of the
DEPARTMENT OF LABOR and the BEA is in the DEPARTMENT OF COMMERCE.] While a president, or his/her Secretary of
Labor or Secretary of Commerce, may run out onto the White House lawn to claim
that everything that is good in the report is a direct result of the
president’s wise economic policies, the reports steer clear of any political
baloney. They represent the economic equivalent of Jack Webb’s “Just the facts,
ma’am.” [Another sad, sad NOTE: At least
according to the bibliography in IMBD, Jack Webb never said, “Just the facts,
ma’am.” Apparently, somebody’s memory is
a little shoddy when it comes to Jack Webb. See: http://www.imdb.com/name/nm0916131/bio] Also, you should know that each report
containing the latest data usually presents those data as ESTIMATES, and those
estimates are often subject to rather substantial REVISIONS, especially when
they are marked as PRELIMINARY. For instance, in the labor market report we got
earlier this month, the BLS states, quite explicitly, “Nonfarm employment
increased by 138,000 in April, and the unemployment rate was unchanged at 4.7
percent.” What could be less complicated
than that? In this case, we don’t even
find out in the fine print of the footnotes that the data may be subject to
(sometimes substantial) revision, when more or better data become
available. The data are informed
guesses; they do not represent actual counts of the numbers that are reported
and sometimes information comes in after the deadline that causes corrections
in the official numbers. As we teach in our statistics classes, when we pay
enough for a guess we call it an estimate, but it’s still a guess.
Why don’t they get it right the first time? There are a couple of
reasons here but mostly it’s because the statistical agencies in the government
are dealing with such a complex economy. To get the numbers out sooner would
mean settling for even less accurate reports and to wait until the data were
completely solid would mean waiting until they were almost totally
useless. It’s our old friend, the
OPPORTUNITY COST problem again. To sort
out the data sooner and more accurately would be possible only if the
government were willing to spend much more money on the effort and then
Congress would not have as much money to spend on bridges to nowhere, other
legislative playthings, and tax cuts for their friends [NOTE: That’s the kind
of political comment you will never find in these reports we are talking
about.].
How do you read this stuff?
Your best bet is to find an article about the report online or in your
morning newspaper. Articles for the
general public generally cut out all of the junk and tend to be written in
English, giving them two major advantages over the reports themselves. However, it has been quite noticeable over
the years that some of the authors of these articles don’t know nearly as much
about economics as they do about journalism, particularly when it comes to
talking about prices and inflation. [NOTE: A PRICE INDEX attempts to measure
the level of prices while the RATE OF INFLATION attempts to measure how fast
prices are rising.] If you want to bring it down to an understandable level,
just pick one or two numbers to watch from each report. When a reporter blows
the explanation of a number you have been following, then it is time to get
another source.
From the employment report, the best number to watch is the number
of JOBS from the ESTABLISHMENT SURVEY, rather than the number of PEOPLE
EMPLOYED, from the HOUSEHOLD SURVEY, or even the UNEMPLOYMENT RATE, also from
the HOUSEHOLD SURVEY. [NOTE: For reasons best known to the government, which
loves to do things this way, the “Employment Situation Report” actually reports
the results of two completely different surveys and the numbers are thrown
together. Reasonable people can get
completely confused by the sometimes inconsistent numbers.] The number of jobs
is a better measure of how we are doing because the survey that it comes from
is more comprehensive than the one that reports the number of people employed
(although it leaves out non-farm workers) and the unemployment rate is driven
by two factors, the number of jobs and the number of people in the labor force.
In the “Consumer Price Index Summary”, we like to watch the “Consumer Price Index for All Urban Consumers (CPI-U)”,
particularly how it has changed over the most recent 12-month period. To use the technical term, the financial
markets tend to GO NUTS over every little monthly wiggle in the numbers but the
pattern is usually one of over-reaction over a few hours, followed by a return
to what passes for normalcy until the next statistical jiggle is felt. By the time that most people find out about
the news the markets have already reached their artificial highs (or lows).
Finally, in the “News Release: Gross Domestic Product”, the only
thing that really matters is the percentage change in REAL, as opposed to
NOMINAL, GDP. REAL GDP attempts to
measure, after washing out much of the price distortion, the volume of goods
and services that were produced in the
If you are trying to gain some insight from the data itself, the
best shorthand method is just to see how it compares with the previous report. For instance, in the April report, prices
were riding merrily along at a little less than 3.4 percent growth rate, which was
the official inflation rate. But the May
report showed a substantial change, to a little more than 3.5 percent, that
could have policy implications. Generally, a lower rate is better than a higher
rate, as long as it doesn’t get too low, and a rising rate makes the markets
nervous. That’s largely because a rising
inflation rate is seen as a sign that the FEDERAL RESERVE will continue to
raise interest rates. 3.5 percent means they are going to continue to raise
interest rates and the market reacted by dropping 214 points. In the employment report for May, the number
of jobs rose, but only by 138,000. That
stinks, especially at this stage in the business cycle. Because the number of new jobs was so low (About
200,000 to 225,000 would be a decent target range. The April number was 200,000)
the financial markets reacted to the labor report by jumping almost 139 points.
The interpretation was that, with job growth slowing down, the FEDERAL RESERVE
might stop raising interest rates in order to protect the growth of the
economy. For real GDP, the most recent
report was that the economy was growing at an annual rate of 4.8 percent in the
first three months of 2006, well above the 20 year average of 3.3 percent, but
also about what everyone expected. The
market didn’t pay any attention, dropping 15 points for the day, largely
because the April 26th report on the same three months had given us
about the same numbers.
After comparing the numbers with the previous report, the next
step is to focus on the TREND, rather than on the current measure. The TREND tells us how much, and in which
direction, the numbers are changing; it represents the pattern that a number
traces over time. If a number normally increases by 10,000 each month, then
we’ll be surprised about any month in which we only get a 4,000 increase and
the financial markets may even react to it as if it were a decrease. However, in the real-world economy, very few
numbers follow nice, smooth trends; there is always some sort of statistical
noise in the data. Some of the distortion can be filtered out, particularly is
the distortion itself has a pattern, such as SEASONAL VARIATION. If we notice
that a number tends to rise by three percent every November because of the
holiday seasons, a two percent increase will actually be seen as a slowdown,
because the number did not hit its seasonal mark. Most of the numbers,
including the labor report, provide two sets of numbers. One set is seasonally adjusted and the other
is the raw data itself. It is up to you
which one you want to swallow. The major advantage of using the 12-month data
is that you are always using all of the months, so you don’t have to worry
about the seasonal variation.
The bottom line would be that all of your data bounce around a
little bit. If you follow one or two
numbers to read the economy and a number comes in outside of its normal (or
expected) range, then sometimes it is a signal that there is something afoot in
the economy. However, most of the time a
one-month or one-quarter signal means nothing at all. Changes in patterns represent real change and
to get a change in a pattern you need more than one outlier. The fun part about
economics is that the first one to notice a change in a pattern can get very
rich. For most of us it’s just a sideline. Enjoy the show, but keep your day
job.
Final notes: If you would
like to be removed from the distribution list just send a reply on email. Back issues are available on my website
<http://www.oocities.org/wsirius30/2cents.html>. The opinions expressed
in these newsletters are those of the author.
Comments, including suggestions for future newsletters, are always
welcome.