Policy makers, the media, and many other people care about small business owners’ perceptions, making the monthly optimism figures put out by the National Federation of Independent Businesses (NFIB) and Discover Small Business Watch (DSBW) noticed statistics. The goal of these measures is to tell us if small business owners are becoming more or less optimistic about prospects for their businesses and the overall economy.
Because these indices get reported and discussed in the media, it’s important to recognize their strengths and weaknesses. First, the overall numbers sometimes mask big differences between groups of entrepreneurs because the surveys are given to very different business owners. Some are male, and some female; some sell products and others provide services; and some serve consumers and others serve businesses. The respondents vary in age, income, number of employees, and years in business.
If the optimism and pessimism of all of these business owners moved in lock step over time, the tendency to focus on the average of all of them wouldn’t be a big deal. Whether optimism was high or low would pretty much be the same for everyone. But when the levels of optimism of different groups don’t all move in the same way over time (they aren’t that highly correlated), then knowing the average but not what’s happening with the different groups hides important information.
I don’t have data on the Optimism Index for different groups of respondents to the NFIB survey, but I do have it for the DSBW from December 2006 through January 2010. So I can talk about those correlations.
While the optimism levels of all of groups are positively correlated, the correlations aren’t super high. For instance, the correlation between the optimism levels of the owners businesses that are one to two years old and those that are six to ten years old is only 0.44 over this time period.
Similarly, the optimism levels of business owners 18 to 29 only correlate 0.64 with those of business owners 65 and older, and optimism levels of owners making under $20,000 per year only correlate 0.66 with those making between $75,000 and $100,000. Thus something common affects the optimism of owners of different ages, those running different aged businesses, and those making different amounts of money, but different factors also drive their levels of optimism.
Second, the responses of the business owners to different questions on the surveys don’t all correlate highly. For instance, there is essentially no relationship (correlation of -0.02) between the share of business owners who say the economy is getting better and the percentage that say they have experienced temporary cash flow problems that caused them to hold off on paying some bills over the past 90 days. And the percentage of small business owners who say that the economy is getting better and the percentage that plan to increase spending on business development correlate only 0.36, while the percentage of owners who say the economy is getting better and the percentage that plan to hire correlates only 0.30 over the August 2006 through January 2010 time period.
What about the question everyone wants to know about right now: are businesses going to hire? Over the August 2006 through January 2010 time period, the share of owners planning to increase spending on business development is a better predictor than the percentage who say that overall economic conditions are improving (a correlation of 0.73 versus 0.36).
But here’s a piece of evidence that shows what a lot of people in Washington are worried about. If the data are split into two time periods – from August 2006 to June 2008 and from July 2008 through January 2010 – the correlation between the share of small business owners planning to spend more on business development and the percentage planning to hire is greater for the first period than for the second. That pattern suggests that the factors driving the spending and hiring plans are more different now than they were in the pre-financial crisis period.
What about the two optimism indices themselves? They’re pretty highly correlated. From December 2006 through January 2010, the NFIB and DSBW optimism indices correlate 0.85. Because the NFIB surveys its members (who tend to run larger businesses than the respondents to the DSBW), that level of correlation suggests that both indices are picking up general trends rather than factors affecting larger versus smaller small businesses or NFIB members versus nonmembers.
The overall measures correlate more highly than specific items. For instance, the NFIB’s measure of the percentage of small business owners who answer “better” minus the percentage that answer “worse” to the question: “About the economy in general, do you think that six months from now general business conditions will be better than they are now, about the same, or worse?” correlates only 0.40 with the percentage of respondents to the DSBW survey who answer “better” minus the percentage who answer “worse” to the question: “Generally speaking, are the economic conditions for your business getting better or worse in the next 6 months?” Unfortunately, we can’t tell whether this low correlation results from the types of businesses surveyed by the two groups or the difference between the NFIB’s focus on general conditions and the DSBW’s focus on the respondent’s business.
None of this says that there is anything wrong with these surveys. They provide us with useful information about what’s going with small business owners’ thinking on an up-to-date basis. We just need to be cautious about how we use them. We can’t assume that the patterns over time are going to be the same for both surveys, between questions on each survey, or between different groups of respondents to the surveys.