Consumer Sentiment in Florida and its Regions

• Gene Averkiou
• Mark Girson
• Scott Richards
• Jacob Ball
• Justin Quinn
• UF Survey Research Center
• Bureau of Economic and Business Research
• University of Florida

Since 1985 the Bureau of Economic and Business Research at the University of Florida has released a report on Consumer Sentiment among Floridians on the last Tuesday of the month. This report is often cited in newspaper articles and radio broadcasts. This article explains what Consumer Sentiment is, how it is used, and how Florida’s Consumer Sentiment differs from the rest of the country and by regions of the state.

The concept and measurement of Consumer Sentiment was pioneered in the 1940s by George Katona, a behavioral economist at the University of Michigan who wanted to predict consumer spending. Most economists agree that consumer spending makes up around two thirds of the economic activity in the U.S. each year, so being able to predict it is important. Katona believed that discretionary consumer spending (such as spending for furniture and appliances, rather than a mortgage) was a function of both the ability and willingness of consumers to spend. Typically, that ability has been measured indirectly by looking at things that affect consumers, such as inflation, levels of employment, interest rates, wages and levels of personal savings.

Random Digit Dialing is a method for generating a telephone sample. Instead of dialing phone numbers from a list, such as the white pages in the telephone book, the numbers are made up at random and called. The benefit of this is that both listed and unlisted numbers can be called. The creation of the numbers is based on the practice of phone companies to release them in banks. A bank consists of the Area Code + Prefix + the first two digits of the Suffix. Since it is known which banks are released, by looking in the telephone book, numbers can be created that have a high probability of being real households. Telephone banks can be geo-coded to census data, making it possible to make up phone numbers for certain geographic regions, or other attributes that cluster geographically such as income.

An index is a number that represents change relative to some base. The University of Michigan index—as well as the University of Florida index—use 1966 as their base. In other words, the index was 100 in 1966. Numbers over 100 mean more optimism than 1966. The Conference Board index is based on 1985. An index is not the same thing as a percentage. Its value is derived from comparisons from month to month, and comparing months to the same month in previous years. The Florida index ranges from 2 to 150. Even though the questions used to measure the index may seem dated, it can be a problem to change them because the index may change as the questions change. It is not unlike the idea of using the same thermometer to measure temperature, because you want to know that changes in measurement are due to changes in temperature, not changes in the instrument.

Willingness to spend is a different matter. While willingness to spend is certainly affected by inflation and wages, economists speculated that sometimes people make decisions to spend based on other factors that may or may not be captured by these indictors. Unique events, such as the attacks on the World Trade Center in 2001, are not readily quantifiable, yet may affect choices in consumer spending depending on their interpretation. Consumer choices are seen as dependent not only on government statistics, but also on what consumers read about current events and what their friends say.

Do Not Call lists are maintained by both the State of Florida (go to http://www.800helpfla.com/nosales.html) and the U.S. Government (go to https://www.donotcall.gov). Anyone can volunteer to be on these lists, and telemarketers must check to make sure they do not call anyone on the list, unless they have existing business with them. Survey researchers are among the few organizations that are exempt from this law. So long as the call is for research purposes only, they do not fall under the Do Not Call legislation.

Consumer Sentiment in the United States has been measured nationally by two sources. The University of Michigan produces an index of consumer sentiment (invented by Katona) based on a national Random Digit Dial telephone survey of 500 households. This survey has been conducted since the 1940s and became a monthly survey in 1978. The other index, provided by the Conference Board, is a Consumer Confidence Index that started in 1967 and became monthly in 1977. The Conference Board is a not-for-profit organization that conducts research on business and economics, both nationally and internationally. Its index is based on a mail-out survey where approximately 2,500 responses are tabulated as they come in for a given month.

Both the University of Michigan index of consumer sentiment and the Conference Board’s Consumer Confidence Index are based on five questions that have remained unchanged since their inception. Both surveys ask respondents two questions designed to measure attitudes about current conditions, and three questions that ask respondents about their future expectations(see Table 1).

A comparison of the two indexes over time shows that they tend to move in the same direction (see Figure 1). There are some notable exceptions, such as the mid 1990s when the dot-com boom created a lot of jobs, in 2005 when job growth accelerated, and in 2008 after the onset of the recession. By examining the actual questions it is clear that the survey researchers at the Conference Board believe that perceptions of job security affect consumer spending, while those at the University of Michigan think perceptions of financial security affect spending. Unfortunately this does not appear to be the case for either index.

Several economists have examined the ability of both indexes to predict spending. In the US, neither index does a great job of predicting short term changes in spending. The indexes do seem to capture the concern of respondents when there are sudden events, such as the 2001 attack on the World Trade Center. However, those short term concerns rarely seem to result in changes in consumer spending of similar magnitude. In other words, consumer spending in the U.S. does not seem to be as affected by “non-economic” events as the behavioral economists once thought. There is some evidence that this method may work better in other countries, such as Canada, for predicting short term changes in spending.

In the U.S., short-term consumer spending can be forecast with other indicators that are immediately available, such as the level of the stock market and interest rates. Other indicators are also useful in explaining spending, such as levels of employment and wages, but these indicators lag sufficiently that they are of less value in predicting short term changes.

Figure 1. Comparison of University of Michigan  and Conference Board Indexes

So if Consumer Sentiment is not useful for predicting short term changes in consumer spending, what is it good for? Over the past fifteen years, interpretations of Consumer Sentiment have taken on a different meaning than what was originally intended. Rather than serving as a predictor of consumer spending, it is now used as a barometer of public opinion about the economy, particularly during election cycles. Figure 2 shows that the number of headlines in major newspaper articles citing Consumer Sentiment has grown over time, with a noticeable spike following the attack on the World Trade Center in 2001, and another major spike in 2008 after the onset of the recession.

This interpretation is borne out by a comparison between the index of consumer sentiment nationally (as measured by the University of Michigan), and the index in Florida (as measured by the University of Florida), as depicted in Figure 3. The reader should keep in mind that these two indexes were measured independently by two different survey organizations. With few exceptions, the two indexes track each other almost perfectly. This raises the question as to whether Florida is doing as economically well as the rest of the country.

Although Florida is the fourth largest state in the country, its economy is quite different from most other states.

Although Florida is the fourth largest state in the country, its economy is quite different from most other states. More than the average state, Florida depends on tourism and on retirees, with large medical and financial sectors. It does not experience the same cycle of ups and downs as, for example, the Midwestern states that are dependent on manufacturing. Indeed, the data suggest that over the past twenty years, Florida’s employment and wage growth has differed greatly from the rest of the country. So why do the indexes for Florida and the U.S. track so closely?

One possible answer, which is the source of some debate among economists, is that Floridians as a whole get their economic news from similar sources as people in the rest of the country. In other words, taken as a whole, Floridians react to the same sort of events reported in the media as

Figure 2. Lexis-Nexis Count of Articles in Newspapers with "Consumer Sentiment" in Headline

Figure 3. Comparison of National and Florida Consumer Sentiment

people throughout the rest of the country. This would mean that Floridians form their opinions about the economy more from what happens nationally than locally. Looking at the questions in Table 1, we can imagine that this is true for some of the components, such as opinions about U.S. business conditions. It is surprising, however, that questions about personal finances are less reflective of local differences. Overall, Florida has a large enough population that is representative of the rest of the country; consequently, Florida Consumer Sentiment does not differ much from the nation.

A Metropolitan Statistical Area is a collection of counties defined by the U.S. Census Bureau as “a core area containing a substantial population nucleus, together with adjacent communities having a high degree of social and economic integration with that core.”  Many economic analyses are done at the MSA level to describe the economic infrastructure.

The margin of error is a measure of the reliability of an estimate. It is a function of the variability of a question or measure and the sample size. As respondents are added to a survey, the margin of error decreases at a decreasing rate. Different questions have different margins of error depending on the variability of the responses.

While Florida as a whole may look like the rest of the country in its opinions about the economy, this is not the case when we compare areas within Florida. Figure 4 shows a comparison of the Consumer Sentiment Index by four regions of Florida. Although the overall trends do seem to track each other, we start to see some differences. For example, Southeast Florida, including Miami-Dade, Broward and Palm Beach counties, which declined in confidence relative to other regions after the 2001 recession, and only caught up to the other regions after several years. This reflected the nature of the Southeast Florida economy which is more exposed to changes in the global market, particularly Central and South America, than other regions of the state.

Figure 4. Florida Consumer Sentiment by Region

Table 2 breaks down consumer sentiment by each of Florida’s Metropolitan Statistical Areas. These analyses were done by combining all observations for each MSA for a year. In most cases this resulted in a sample size large enough to reliably estimate Consumer Sentiment. The reader should be aware that the margin of error for some of these estimates is relatively high, meaning that a repeat of the survey may yield different results. We chose to list all values so that readers could observe the trends, even though the reliability of some of the values is suspect.

There are some interesting trends among these MSAs. In most cases, consumer sentiment starts high in 2000 and falls dramatically in 2001. This coincides with the stock market crash associated with the bursting of the dot-com bubble and the recession of 2001. All MSAs show a buildup in sentiment until 2004. After the recession began in 2007, 16 of 20 MSAs registered declines in sentiment of 15 points or more from 2007-2008.  Changes in sentiment have varied across the state since 2008, with 13 MSAs showing improvement, while 7 MSAs have registered further declines.

Table 2. Consumer Sentiment 2001-2011 by Metropolitan Statistical Area

Perhaps the main contribution of Consumer Sentiment, at least here in Florida, is that it serves to focus attention on various economic forces and how those affect ordinary people. We often read articles about one economic indicator or another, and it is difficult to understand how those things affect us. Consumer Sentiment is a regularly recurring measure that captures all of those things into a single number that represents how Floridians are feeling about the economy. Although it may not do a much better job than other indicators for making predictions, it gives us a platform to talk about the various things that really do affect Floridians. Consumer Sentiment doesn’t tell us what will happen; it tells us what has happened, and if it can be expected to continue.

This helps people understand what is happening in the economy around them, and hopefully, to better understand how to adapt to change

Questions about this Florida Focus? Contact Chris McCarty at 352-392-2908 ext. 101 or e-mail chrism@ufl.edu.