Posts filed under “Sentiment”

Do Lower Gasoline Prices Boost Confidence?

Do Lower Gasoline Prices Boost Confidence?
Aditya Aladangady and Claudia R. Sahm
FEDS Notes, March 6, 2015



A gallon of gasoline currently costs one third less than it did last summer, as shown by the solid line in Figure 1. Gasoline makes up 5 percent of average family spending, so lower prices at the pump should increase current purchasing power for many.1 A drop in gasoline prices may also make people more confident about their future resources–the focus of this note.

Figure 1
Figure 1: Gasoline Prices and Consumer Sentiment. See accessible link for data.

Source: Energy Information Administration for motor gasoline regular retail price (including taxes): is an EIA forecast, Univerisity of Michigan Surveys of Consumers for sentiment: Leaving the Board

Accessible version

In fact, recent surveys show a rise in confidence that coincides with the decline in gasoline prices. As one example, we estimate that energy price declines can account for about two-thirds of the increase since last June in the Index of Consumer Sentiment (the dashed line in Figure 1).2

Summary measures of confidence can be difficult to interpret, however, so we turn now to specific questions about expectations and buying conditions to explore the impact of gasoline prices further. First, we see that individuals’ expectations about future changes in gasoline prices are informative of actual prices changes, suggesting that individuals update their views as conditions change. Specifically, the survey asks “… do you think that the price of gasoline will go up during the next twelve months, will gasoline prices go down, or will they stay about the same as they are now?” In Figure 2, the diffusion index of expected gasoline price changes for the next year from individuals (blue line) is quite volatile but appears to align fairly well with the realized year-ahead percent change in gasoline prices (gray line).3 The correlation between the two series of 0.19 is notable given the unpredictability of gasoline prices.

Figure 2
Figure 2: Gasoline Prices: Survey Expectations and Actual Price Changes Over Next Year. See accessible link for data.

Source: Energy Information Administration for motor gasoline regular retail price (including taxes):, Univerisity of Michigan Surveys of Consumers for sentiment: Leaving the Board. Quarterly data from 1982:Q2 to 2014:Q4

Note: The one-year-ahead survey expectations for gas prices were not asked 1993-2006

Accessible version

As actual gasoline prices have fallen recently, expectations about gasoline prices have also softened. This suggests that individuals have a more optimistic view about their future purchasing power. To test this connection, we examine views on expected real income growth conditional on individuals’ expectations about gasoline prices.

The remaining analysis uses the individual survey responses in Michigan survey from July 2014 to February 2015 from about 4000 people. In those eight months, on average, nearly 10 percent of individuals expected gasoline prices to go down, twice the rate in 2013. Figure 3 shows that during this period individuals who expect gas prices to go down in the next twelve months (green bars) are more optimistic about their real income prospects than individuals who expect gas prices to go up (orange bars) or gas prices to stay the same (gray bars).4

Figure 3
Figure 3: Expected Change in Real Income Over Next Few Years by Gasoline Price Expectations. See accessible link for data.

Source: University of Michigan Surveys of Consumers.

Note: In total, 18 percent expect real income to go up, 36 percent expect real income to be unchanged, and 46 percent expect real income to go down. Differences between green and orange bars are statistically significant at 1-percent level.

Accessible version

Second, this inverse relationship between expectations about gasoline prices and expectations about real income appears robust throughout the distribution of current income. Figure 4 expands the left-most grouping in Figure 3–those expecting real income to increase–by income quintiles. At each quintile, those expecting gas prices to go down (green bars) are more likely to expect their real income to increase than those who expect gas prices to go up (orange bars). With the smaller subsample–only those expecting a real income increase–the differences by gasoline price expectations are less precisely estimated. Since last summer, the difference is largest (and statistically significant) in the middle of the income distribution.

Figure 4
Figure 4: Expect an Increase in Real Income by Gasoline Price Expectations Across Income Quintiles. See accessible link for data.

Source: University of Michigan Surveys of Consumers.

Note: In total, 18 percent expect real income to go up, 36 percent expect real income to be unchanged, and 46 percent expect real income to go down. Differences between green and orange bars are statistically significant at 1-percent level.

Accessible version

These data suggest that the lower gasoline prices have made a broad set of individuals more confident. Of course, whether and when this heightened optimism translates to more consumer spending is another question. While the Michigan survey does not ask about spending, it does ask whether it is a good time to buy durables.5 As shown in Figure 5, individuals who expect gasoline prices to decline further are more likely to view buying conditions as favorable.

Figure 5
Figure 5: Good time to Buy Consumer Durables by Gasoline Price Expectations Across Income Quintiles. See accessible link for data.

Source: University of Michigan Surveys of Consumers.

Note: The percent of who say it is a good time to buy generaly rises across the income quintiles: 69, 76, 75, 79, and 81 percent. The differences between the blue and red bars are statistically significant at the 5 percent level for the the first three income quintiles.

Accessible version

Interestingly, the differences in reported buying conditions in the bottom half of the income distribution are even more pronounced than the differences in expected real income. One possible explanation for the stronger responses at lower income levels is that gasoline represents a larger share of family budgets. As shown by the bars in Figure 6, the gasoline expenditure shares from the Consumer Expenditure Survey peak in the middle income quintile and are lowest for top income quintile. Nonetheless, the solid line in Figure 6, indicates that the mean spending on gasoline does rise with income, meaning that while those with low income may see a larger fraction of their income freed up, falling gasoline prices cause the largest real income boost in dollar terms for those with higher income. The extent to which this income boost translates into new spending may also differ across individuals. In particular, those whose spending had previously been limited may exhibit a higher marginal propensity to consume.

Figure 6
Figure 6: Gasoline Expenditures in 2013 by Income Quintile. See accessible link for data.

Source: Consumer Expenditure Survey.

Accessible version

All told, it appears that the decline in gasoline prices since last summer have both made individuals more optimistic about their future real incomes and bolstered their assessment of current buying conditions. Together with the direct savings from lower gasoline prices that are currently being realized, the extra confidence may also boost consumer spending going forward. Importantly, the impact of lower gas prices on confidence appears broad-based across income groups and may therefore lead to increased spending for a wide range of individuals.

1. Expenditure shares for gasoline from Consumer Expenditure Survey, 2013, for example: to text

2. Index of Consumer Sentiment combines the relative score (percent giving favorable replies minus percent giving unfavorable replies plus 100) to five questions about current and expected economic conditions. Our estimate of the energy-price effect is from a monthly regression model of sentiment that also includes CPI inflation, percent change in stock prices, unemployment rate, and change in payroll employment as covariates. The estimation range includes January 1978 to February 2015. Return to text

3. The diffusion index is the percent expecting an increase in gasoline prices minus the percent expecting a decrease in gasoline prices plus 100. Return to text

4. Specifically, the income question asks: “During the next year or two, do you expect that your (family) income will go up more than prices will go up, about the same, or less than prices will go up?” In Figure 2, the bars of the same color sum to 100 percent. The relative heights of the bars of the same color reflect differences in expectations about real income. In general, households are pessimistic about their income prospects. See earlier analysis of income expectations in Sahm (2013): Return to text

5. Specifically, the buying conditions question asks: “About the big things people buy for their homes–such as furniture, a refrigerator, stove, television, and things like that. Generally speaking, do you think now is a good or a bad time for people to buy major household items?” Return to text

Disclaimer: FEDS Notes are articles in which Board economists offer their own views and present analysis on a range of topics in economics and finance. These articles are shorter and less technically oriented than FEDS Working Papers.

Category: Energy, Federal Reserve, Sentiment, Think Tank

Top Issue for Americans? Bad Government

The Gallup polling organization is in the business of taking the pulse of the American public. A new survey detects a surprising shift in perceptions. It points not just to the nation’s sour mood and disappointment with the drawn-out recovery, but also to a sense that something different is now wrong with the U.S. In…Read More

Category: Politics, Sentiment

Rich Countries on Whether Any Day Is a Good One

This is such a strange data point . . .     Source: Pew Research

Category: Data Analysis, Digital Media, Sentiment

Wall St Ends 2014 More Cautious Than it Began

From Bank of America Merrill Lynch: The Sell Side Indicator — our measure of Wall Street’s bullishness on stocks — moved sideways in December, unchanged from November’s reading of 52.1. At the start of the year, it was 53.3. The indicator remains in “Buy” territory, as Wall Street’s bearishness is still more extreme than at…Read More

Category: Data Analysis, Sentiment

Beware Anecdotal Sentiment Readings!

It’s the time of year when predictions are in order. Not by us, but by other people. We have spilled plenty of pixels on why forecasts are folly (see this, this, this, this and this); we won’t revisit that well-trod ground, at least not today. Instead, I wanted to discuss the rather annoying tendency of…Read More

Category: Really, really bad calls, Sentiment

What Worries Investors?

Great graphic from Bloomberg


click for complete graphic
inflation deflation


Read More

Category: Digital Media, Investing, Sentiment

Disbelief in Equities is What is Driving Them Higher

As a fan of investor psychology, I find sentiment intriguing. Measuring it is a challenge. We can’t trust what people say because they become bullish after they buy and bearish after they sell, convincing themselves that past trades were the correct way to go. Humans are notorious liars — especially to themselves. When they are…Read More

Category: Investing, Psychology, Sentiment

Playboy Bunnies, Mila Kunis and the Myth of the Celebrity Stockpicker

Last week, I came across the following headline: “As music sales fall, sax player Kenny G turns to stockpicking.” My immediate reaction: Uh oh. The last thing any bull market needs is for celebrities to be featured in the financial press. As soon as that starts, it means the bull market must be near a…Read More

Category: Contrary Indicators, Investing, Really, really bad calls, Sentiment

Too Bullish or Too Bearish?

One of my favorite pastimes is dissecting accepted Wall Street wisdom to see if it contains any value for investors or traders. Often, upon examination, the widely held beliefs turn out to be closer to magical thinking than financial acumen. One of the more recent examples is the way some analysts use data on sentiment…Read More

Category: Markets, Sentiment, Trading

1958: Public Investing in the Stock Market as Never Before

This week in 1958: LIFE Magazine highlights a strange new phenomenon: The public is investing in the stock market as never before. “On the average,” reports LIFE, “500,000 new customers a year have been getting into the market and 8.6 million Americans now own some kind of common or preferred stock…. To an extent which…Read More

Category: Investing, Markets, Sentiment