« »

Friday, April 11, 2014

2014 economic forecast: Small to mid-size business insights vs. economic statistics

Economic data that claims to measure the same thing often differs. This variance is evident across numerous forecasts from various industrial demographics and between government and private organizations. For instance, when compared against large multi-national or federal organizations, the economic projections of small to mid-size businesses demonstrates considerable variance in GDP growth, forecasted employment trends, payroll rises and revenue growth. This is apparent in the attached slide share research published by Pepperdine University.

To illustrate the differences in future economic expectations, consider U.S. gross domestic product or GDP. According to the International Monetary Fund, U.S. GDP is forecast to grow 2.8 percent in 2014. Yet, the Pepperdine University survey of small to mid-size businesses yields very different GDP estimates ranging from .6 percent to 1.2 percent. How numbers such as GDP are calculated reflect variable selection bias between data sets as otherwise the numbers would be the same assuming the input variables are correct. Thus, how GDP is calculated is just as important to the accuracy of what it predicts.

2014 Economic Forecast: Insights from Small and Mid-Sized Business Owners from Pepperdine University Graziadio School of Business and Management


Business confidence is another area where conflicting data sets raises questions of validity. For instance, in the above slide show, business confidence improves from 55 percent of businesses being less confident or neither confidence or unconfident about future business in 2013 to 45 percent in 2014. However, if one consults the National Federation of Independent Business' Small Business Optimism Index, it appears year-over-year confidence about the future is largely unchanged on average between 2013 and 2014. Taken together however, the two measurements do seem to indicate a sentiment that is neither here nor there in regard to future commercial successes.

Furthermore, the Federal Reserve Bank has itself published statements that attest to the inaccuracy of government statistics and their use as a political tool.  This is clear in another Moneycation post from March 16, 2012 titled "Federal Reserve accuses BLS and media of misleading the public via economic statistics". According to the post, the Federal Reserve Board has stated misinterpretation and "measurement problems" reduce the effectiveness and interpretive capacity of economic statistics. 

An additional example of differences between economic data and potentially misleading economic data forom the Bureau of Labor Statistics is in labor force metrics. The Moneycation post titled, "Why U.S. unemployment data is skewed" illustrates this point. Moreover, in terms of data produced by the BLS, labor force data such as size of workforce and unemployment rate is vulnerable to being skewed. A part of the reason why unemployment data is skewed is stated to be the following per Moneycation:
"What skews the data is that unemployment statistics are based on the total labor force of 153.4 million per the Bureau of Labor Statistics and not the labor force participation rate of roughly 100 million. Such being the case, unemployment statistics include non-participatory civilian labor when calculating unemployment percentages."
When evaluating the strength of data such as economic statistics, Richard F. Taflinger of Washington State University suggests one must evaluate who or what organization conducted the study, the sampling group utilized in the study and what is being measured. Another important attribute of statistics is the method of analysis. There are multiple ways of aggregating, comparing and testing the strength of statistical data. For example, California State University notes the statistical accuracy of data is measured by statistical significance of information and strength of association between variables. The former determines whether or not a data is meaningful and the latter evaluates how strong the data is in representing a sample such as a group of people surveyed.

When viewing media reports, business studies or non-profit research that cites economic statistics, there is some chance the information is either incorrect or inaccurate. When data serves a hidden purpose such as advancing policy agendas or product sales, the probability of imprecise information is also a possibility. In light of this, comparing economic data and weighing it against other data for the purpose of evaluating validity or quantitative reasonableness is wise practice.