We look to hard data instead of models as much as we can. Two structural components of our research are our surveys of state-level sales and withheld tax receipts, and our indexes of transportation fuel usage.
Both have excellent track records of identifying changes in the economic outlook before they show up in official economic data.
State-level sales and withheld tax collections
Every month we have detailed discussions with senior revenue forecasters in the largest states around the country. In addition to more general discussions of how things are going in revenue world, and special concerns they may have, we track monthly sales tax and withheld tax receipts, both against the forecasts put together by experienced revenuers, and against collections of the same month in the prior year. Four indexes for both sales and withheld receipts flow from this information: an index of the share of states that achieved their forecasted collections, an index of those reporting growth over the year, an average over-the-year percentage change in growth rates, or declines, and the average percentage by which forecasts were exceeded and missed.
Tracking raw tax data is a risky proposition, and gets a lot of people in trouble. Daily collections are very noisy, and trends are highly vulnerable to calendar issues and one-off events, such a tax-payer behavior driven by upcoming changes in tax rates. That’s why our primary index, the one tracking actual receipts against forecasts, is the one we consider the most important: revenue forecasters take the calendar into account when they set their bogeys, they let us know if receipts were affected by a specific event, like accelerated payments, and do their best to figure out the effects of non-economic events. Over the last two decades, our indexes have been highly accurate in catching changes in the economy before they show up in the official statistics and, especially, before they show up in the main stream press. If revenue officials are surprised by the rate at which money is flowing into their coffers, that’s important news.
Each state and the District of Columbia report monthly motor fuel and special fuel taxable sales to the Federal Highway Administration (FHWA). Then, with about a 3-month time lag, the FHWA publishes a report summarizing the state provided information.
The monthly motor fuel series dates from January 1999 and the monthly special fuel series dates from January 2003. Each of these series consists of an aggregation of several grades of fuel. The motor fuel series aggregates regular gasoline, premium gasoline, and a number of ethanol fuel blends. The special fuel series includes primarily diesel fuel, but in addition a small amount of liquefied petroleum gas.
The month designated in the FHWA reports represents the month during which the motor fuel taxes were remitted to the states. Generally, taxes are remitted the month following the distribution of fuels from terminals to retail locations.
Sightlines Bulletin Diesel Fuel Sales Analysis
Each month Sightlines Bulletin presents an analysis of diesel fuel sales. Because almost all taxable diesel fuel is consumed by trucks, it provides a good indication of the direction the economy is moving, with a 3 to 6 month lead time. In the United States, approximately 81% of total freight by weight and 86% of total freight by value moves by truck. Thus, because the movement of intermediate and finished goods precedes household and business purchases in our consumer driven economy, sales of the fuel that powers most trucks provides a strong signal of the overall economy’s future direction.
Sightlines Bulletin analyzes changes in diesel fuel sales both nationally and within different regions of the nation. A 1-month diffusion index measures the extent to which diesel fuel sales increase or decrease across the nation’s 50 states and the District of Columbia. Regional changes in diesel fuel sales are measured over 3-month, 6-month, and 12-month time horizons for eight regions. The relationship between diesel fuel sales and the larger economy is analyzed by tracking year-over-year percentage changes is a 3-month moving average of diesel fuel sales and year-over-year percentage changes in non-seasonally adjusted total non-farm employment and manufacturing employment.
1-Month Diffusion Index
The diffusion index is constructed by assigning values of 0.0, 0.5, or 1.0 to each state and the District of Columbia based on its year-over-year percentage change in diesel fuel sales for the month. If the change is zero or negative the assigned value equals 0.0; if the change is greater than zero but less than 0.05% the assigned value equals 0.5; if the change is equal to or greater than 0.05% the assigned value equals 1.0. All of the values are summed, divided by 51, and multiplied by 100 to yield the month’s index value. The monthly values of the index span the range from 0 to 100.
The U.S. Bureau of Economic Analysis reports a variety of economic statistics for eight regions of the country. The monthly diesel fuel analysis computes 3-month, 6-month, and 12-month diesel sales percentage changes for each of the same eight regions. These regions are defined as follows:
• New England (NE) – Connecticut, Maine, Massachusetts, New Hampshire, Rhode Island, and Vermont
• Mideast (ME) – Delaware, District of Columbia, Maryland, New Jersey, New York, and Pennsylvania
• Great Lakes (GL) – Illinois, Indiana, Michigan, Ohio, and Wisconsin
• Plains (PL) – Iowa, Kansas, Minnesota, Missouri, Nebraska, North Dakota, and South Dakota
• Southeast (SE) – Alabama, Arkansas, Florida, Georgia, Kentucky, Louisiana, Mississippi, North Carolina, South Carolina, Tennessee, Virginia, and West Virginia
• Southwest (SW) – Arizona, New Mexico, Oklahoma, and Texas
• Rocky Mountain (RM) – Colorado, Idaho, Montana, Utah, and Wyoming
• Far West (FW) – Alaska, California, Hawaii, Nevada, Oregon, and Washington.
The relationships among the three time periods indicate whether diesel sales are accelerating or decelerating. A greater year-over-year percentage change during the 3-month period than during the 6-month and 12-month periods indicates diesel sales are growing at an increasing rate. The interregional comparisons provide an indication of variations in expected growth rates for the economies of the different areas of the country.
Diesel Fuel Sales and Employment Trend Comparisons
The U.S. Bureau of Labor Statistics (BLS) generally releases estimates of national non-farm employment and employment by major business sectors the first Friday of each month for the prior month. The BLS releases both non-seasonally adjusted and seasonally adjusted estimates. Revised estimates are released each of the following two months. Once a year the estimates are “benchmarked” based on quarterly unemployment tax returns filed with state employment agencies.
The Year-to-year percentage changes in total non-farm employment and manufacturing employment provide a good indication of turning points and growth rate variations for the overall national economy. The comparison of the non-seasonally adjusted version of these two employment series to a 3-month moving average of national diesel fuel sales provides a basis for gaging the extent to which diesel sales can be used for forecast growth in the national economy. The time lags between the different series are not constant but 3 to 6 month time differences are common.