This page has been archived and commenting is disabled.
The Relationship Between Gas Prices & Consumer Spending, Part I
This is from Stone Street Advisors
I've spent far too much time over the past few weeks going back to a
hodge-podge of data collected from various government agencies in a
semi-futile effort to really learn how, if at all, the price of gas
affects consumer spending, and if so, to what degree. I say semi-futile
because some of the data is not available in certain frequencies the
rest is, e.g. per capita spending on gas in CPI is annual, while the
rest of the data is mostly quarterly. So, with the caveat that the data
from which the following charts and computations are derived is both
imperfect and noisy**, I present my (simplistic) conclusions:
First
up, let's take a look at Disposable Personal Income (DPI), Personal
Spending on non-gas goods & services (Personal Consumption
Expenditures, or PCE less gas), Retail Gas Prices, and the 4-quarter
moving average thereof:
The most immediate conclusion is that gas prices are MUCH
more volatile than both incomes and consumer spending, even if we use a
smoothed moving average to measure them. Even if we look at a much
shorter (5 years v. 26) time horizon, gas prices are significantly more
volatile (remember the summer of '08?):
What
should be apparent by now is that the shape of the personal spending
curve matches almost exactly the shape of the income curve. In fact,
over the 1984-2009 period, the correlation between disposable income and
personal spending is 0.99, while the correlation between PCE and gas
prices is only 0.54.
Over this same time period, in aggregate,
spending on gas (and motor oil) has accounted for anywhere from 4.01%
(1999q4) to 9.06% (2008q4) of total personal spending, with an average
(arithmetic mean) of 5.23% and standard deviation of 1.23%. This works
to $888 to $2,715 annual per capita spending on gas, averaging $1,312, a
pretty damn wide range considering per capita disposable personal
income ranged from $20,693 to $33,480, averaging $26,784!
Intuitively,
if one year we spend $900 on gas and the next we end up spending almost
$3,000, the difference is going be reflected in some combination of
less spending on other goods and services, dipping into our savings, and
credit card balances. What that combination is (and how it has changed
over time) is the subject of my next post on the subject, so stay
tuned!
** These are unadjusted numbers and include transfer payments. Data from BLS, EIA, DOT, Federal Reserve.
- advertisements -
Those of us who work at home find this article amusing until, of course we fill up once a month. I've had to give up KitKats.
we need to take the analysis beyond even the lyrics of the great American songster John Cougar Mellencamp, into the very heart of science
since i've been told by stone street to only post positive comments or not post at all i will comply: stone street advisors are the kindest, bravest, warmest, most wonderful human beings I've ever known in my life. they post excellent articles at zh, too, and i learn things from them i otherwise wouldn't know.
"everyone lies to Congress"...cant forget that comment..
OK, pretty obvious.
But now do some statistical regression so we know how much of a decrease in PCE can be expected for various levels of gasoline price.
So, if I spend more on gas I'll spend less elsewhere?
Got it.
Who would have foreseen that conclusion?
That's not necessarily true, intuitively or empirically; you could dip into your savings or use credit to keep your spending at the same or an even higher level.
...or you could kick the shit out of some investment advisors and take the money that they rooked their clients out of.
In the follow-up post we get to find out what gets cut.
I can hardly contain my fucking excitement.
This was a great post...
For me to poop on.
Why?
Anyone paying the least attention in ECO 101 already knows this.
This is fodder.
The goal is to try to quantify it beyond "knowing" it, champ.
No one "makes" anybody read any of the articles here.