estimation and forecasting retail sales in florida

ESTIMATION AND FORECASTING OF RETAIL SALES IN FLORIDA

By:

The Washington Economics Group, Inc.

July 29, 2008

2655 LeJeune Road, Suite 608Coral Gables, Florida 33134

Tel: 305.461.3811 – Fax: 305.461.3822 [email protected] www.weg.com

i

TABLE OF CONTENTS

Page

I. EXECUTIVE SUMMARY ............................................................................................................ 1

II. BACKGROUND AND SUMMARY FINDINGS OF RETAIL SALES TRENDS IN FLORIDA............. 2

A. Retail Sales Data by Major Category ...........................................................................3

B. Retail Sales – Seasonal Patterns...................................................................................4

C. Retail Sales – An Analysis of Trends over Time .........................................................10

D. Retail Sales – Population Growth is Important in Forecasting Retail Sales..............12

E. Income Changes are Also a Key Driver of Retail Sales .............................................12

F. Retail Sales – Cumulative Growth ..............................................................................14

II. ECONOMETRIC ANALYSIS: FORECASTING EXPECTED CHANGES IN RETAIL SALES ........ 15

A. Results of the Regression Analysis..............................................................................17

III. FORECASTING EXPECTED CHANGES IN RETAIL SALES – SECTOR ANALYSIS ................... 20 APPENDIX I: INFORMATION REGARDING FLORIDA TAXABLE SALES DATA .............................. 30

APPENDIX II: THE WASHINGTON ECONOMICS GROUP, INC. PROJECT TEAM AND QUALIFICATIONS ...................................................................................................... 35

ii

List of Tables

Table ES-1. Retail Sales Forecast 2009 (by Product and Services Sectors ............................................ 1 Table 1. Average Retail Sales by Type and Month 1990-2007 (Current Million $) ........................ 7 Table 2. Average Retail Sales by Type and Month 1990-2007 (Constant 1983 Million $) ............. 8 Table 3. Retail Sales by Type and Month as Share of Yearly Sales 1990-2007............................... 9 Table 4. Retail Sales Shares by Type and Year 1990-2007............................................................ 11 Table 5. Retail Sales Growth 1990-2007 ........................................................................................ 13 Table 6. Relative Sales Levels 1990=100....................................................................................... 14 Table 7. Forecast Estimation Equation Coefficients....................................................................... 18 Table 8. Estimated Long-Run Marginal Effects on Retail Sales .................................................... 19 Table 9. Taxable Sales Forecasts Current Dollars by Sector-2008................................................. 20 Table 10. Taxable Sales Forecasts Current Dollars by Sector-2009................................................. 21 Table 11a. Estimated Coefficients—Durables.................................................................................... 23 Table 11b. Taxable Sales Forecast—Consumer Durables.................................................................. 23 Table 12a. Estimated Coefficients—Non-durables ............................................................................ 24 Table 12b. Taxable Sales Forecast—Consumer Non-durables........................................................... 24 Table 13a. Estimated Coefficients—Tourism and Recreation ........................................................... 25 Table 13b. Taxable Sales Forecast—Tourism and Recreation ........................................................... 25 Table 14a. Estimated Coefficients—Building Investment ................................................................. 26 Table 14b. Taxable Sales Forecast—Building Investment ................................................................. 26 Table 15a. Estimated Coefficients—Business Investment ................................................................. 27 Table 15b. Taxable Sales Forecast—Business Investment................................................................. 27 Table 16a. Estimated Coefficients—Autos and Accessories.............................................................. 28 Table 16b. Taxable Sales Forecast—Autos and Accessories ............................................................. 28

List of Figures

Figure 1. Level of Sales by Month ..................................................................................................... 5 Figure 2. Durables, Non-durables and Business Peak in December .................................................. 6 Figure 3. Sales by Month as Share of Yearly Sales – Major Categories............................................ 6 Figure 4. The Share of Sales for Automobiles Has Increased.......................................................... 10 Figure 5. The Share of Sales for Tourism Has Declined.................................................................. 10 Figure 6. Total Forecasted 2008 Taxable Sales Month Change From 2007 .................................... 21 Figure 7. Total Forecasted 2009 Taxable Sales Month Change From 2008 .................................... 22 Figure 8. Factors Explaining Sales Variability Across Time ........................................................... 29

The Washington Economics Group, Inc. Page 1

I. EXECUTIVE SUMMARY

The forecasting model of retail sales contained in this Study was developed exclusively for Florida Retail Federation members by The Washington Economics Group, Inc.

The overall results of statistical testing indicate the following:

Population and per-capita income growth accounts for a significant 87 percent of the expected change in overall retail sales

Consumer expectations and month of the year account for a 12 percent change in retail sales.

The remaining 1 percent change in retail sales can be attributed to random factors, not accounted by the model.

The forecast model estimates a moderate recovery in overall retail sales for 2009, after a decline of 3.0 percent in 2008. The expected recovery of retail sales is estimated at 2.75 percent from the low levels of 2008.

The cumulative effect of 2007 and 2008 loses, with the moderate gain forecasted for 2009, would still leave the total retail sales level close to where it was in 2004, after adjusting for inflation.

In essence, the expected 2009 performance is the start of a recovery from the declines suffered in prior years. As Per-Capita Personal Income, Population and Consumer Confidence pick up by 2010, a stronger recovery and eventual expansion of overall retail sales can be expected.

Table ES-1 highlights the forecast for retail sales in 2009 adjusted for inflation and broken down by principal product and services sectors:

Table ES-1. Retail Sales Forecast 2009 (By Product and Services Sectors)

Sector Total Retail Sales

$ Million Change From 2008

(Inflation Adjusted %) Non-Durables $100,522 1.50% Durables 29,373 8.76% Tourism/Recreation 68,596 1.32% Business Related 70,822 4.95% Automobiles & Accessories 65,633 12.31% Building Related 23,972 6.00% Totals (Weight Adjusted) $358,918 2.75% Source: Forecast Model.


II. BACKGROUND AND SUMMARY FINDINGS OF RETAIL SALES TRENDS IN FLORIDA

The goal of this project is to create an economic forecasting model to estimate retail sales by major retail establishment classification sector and by geographical regions--Metropolitan Statistical Area (MSA).

Based on the stated goal, the Study examines retail sales figures for the period 1990-2006 and develops a model to forecast retail sales by major sectors on a statewide basis. The Study also provides some preliminary findings for an extension to forecast retail sales at the MSA level. Findings of the Study include:

• Population and per capita income account for 87 percent of the change in overall retail sales. Month of the year and consumer expectations account for 12 percent. The remaining 1 percent is unexplained by the model and can be attributed to random factors.

• A 1 percent change in population increases retail sales by 4/10th of 1 percent in the short run and 9/10th of 1 percent in the long run.

• A 1 percent change in real per capita income changes retail sales on the short run by ½ of 1 percent, and by 1.07 percent in the long run. This means that sales respond more than proportionally to changes in per capita income adjusted for price changes (inflation).

• Consumer expectations are important. A 1-point change in consumer expectations, changes sales by 1/10th of 1 percent.

• December is the best month for retail sales. After adjusting for income, population and consumer expectations, December sales are 30 percent higher than the lowest month (January), and from 17 percent to 22 percent higher than any other month except March. March is the second best month, only 8 percent below December, on average.

• December sales are highest for durables, non-durables, and business. Tourism and automobile retail sales are highest in March. Construction is highest in October.

• Since 1990, the share of the retail dollar has increased for Autos and Accessories (+6.5%), Consumer Durables (+17.8%) and Business Investment (+5.2%). The share of the retail dollar has dropped for Tourism and Recreation (-6.7%) and Consumer Non-durables (-4.7%).


• In real terms, total taxable retail sales have grown at 2.6 percent per year from 1990 through 2007.

• About two-thirds sales growth is driven by Florida’s population growth. After adjusting for population growth, the sales grew at a rate of only ½ of 1 percent per year.

• Since 1990, the years 2004 and 2005 marked the best growth years. Inflation and population-adjusted growth was in excess of 5 percent each of these two years. However 2006 and 2007 marked the worst back-to-back sales performance. Population adjusted sales growth was slightly negative in 2006 (-0.6 percent) and a record setting -7.7 percent in 2007.

• The negative growth rates in 2006 and 2007 are the most notable because they occurred even as real per capita incomes grew—albeit at a low 1 percent rate—for each of these two years. Previous declines in per-capita sales had been concurrent with income declines. The evidence is highly suggestive of consumers retrenching to devote a larger share of their income to pay for debt or for savings.

A. Retail Sales Data by Major Category

For the purposes of this analysis, 17 complete years of data (1990-2007) on retail sales have been aggregated into 6 major categories. These categories, which correspond to available data from the Florida Department of Revenue (FDOR), are: Autos and Accessories, Consumer Durables, Tourism and Recreation, Consumer Non-durables, Building Investment and Business Investment.

The data for the Study comes from FDOR taxable sales data reports. Sales taxes are levied on sales of goods, but not services, although there are some taxable services. Major categories of exempt sales are food not prepared for immediate consumption, medical and legal services, residential utilities, items purchased for resale, intangible personal property, and rentals over six months. It is estimated that taxable sales comprise 40-45 percent of all retail sales.

Taxable sales are classified according to the sales tax license main line of business of the establishment reporting the sales, not by the type of goods being sold. For example, wine sold at a grocery store would be classified as a “Non-durable” while wine sold at a liquor store would be classified under “Tourism and Recreation.” Likewise an appliance sold at a home improvement store would be classified as “Construction” rather than a “Durable.” With these caveats, taxable sales are classified according to the following list of business activities:


Autos and Accessories The category of "automobiles and accessories" taxable sales includes the sale of new and used cars, repair shops, auto supply stores and taxable sales at gasoline stations.

Consumer Durables The category of "consumer durables" taxable sales includes the sale of appliances, furniture, home electronics, aircraft, boat dealers, hardware and decorating stores.

Tourism and Recreation

The category of "tourism and recreation" taxable sales includes hotels and motels, bar and restaurant sales, liquor stores, photo and art stores, gift shops, admissions, sporting goods, rentals and jewelry stores.

Consumer Non-durables

The category of "consumer non-durables" taxable sales includes food and convenience stores, department and clothing stores, drug stores, antique dealers, bookstores, florists, pet dealers and suppliers, social organizations, storage, communications firms, print shops, nurseries, vending machines, utilities and any "kind" that doesn't fit into the other categories.

Building Investment

The category of "building investment" taxable sales includes sales by building contractors, heating and air conditioning contractors, insulation, well drilling, electrical contractors, interior decorating, paint and wallpaper shops, cabinet and woodworking shops, soil, lumber and building suppliers and roofing contractors. Services provided by these businesses are not generally taxable.

Business Investment

The category of "business investment" taxable sales includes farm equipment, feed and seed suppliers, store and office equipment, computer shops, machine shops, industrial machinery, hotel and restaurant suppliers, transportation equipment, manufacturing and refining equipment, industrial suppliers, paper and packaging materials, medical and optical supplies, commercial rentals and wholesale dealers. Transactions reported as subject to the "use" tax are also included here, regardless of the kind of code of the business reporting the "use" tax.

B. Retail Sales--Seasonal Patterns

Tables 1 and 2 on pages 6 and 7 show average retail sales by month in current and constant dollars for the period between January of 1990 and December of 2007, for example for the month of March, the sales average for Autos and Accessories was $3,567 million in current


dollars and $1,953 million in constant (1983) dollars1. Table 3 shows the shares of sales of each type by month. For example, 11.8 percent of the sales for Consumer Non-durables were made in December alone.

An analysis of Tables 1, 2 (pages 6 and 7) and 3 (page 8), reveals some important patterns. The more obvious one being that December is the best retailing month. Sales of consumer durables and non-durables are particularly high during this month. On the other hand, March is the best month for automobiles and accessories and tourism and recreation. While December is the best month with 10 percent of all sales for the year, March is the second best with 9.1 percent. January is the worst retailing month with 7.8 percent of all sales.

Relative to December, January’s sales are lowest (78 percent as high). All other months have sales that are between 79 percent and 83 percent of December’s level, except for March at 90 percent.

1As a point of reference a 1990-dollar is equal to about $2.02 in 2006. Therefore, the constant dollar figure in 2006 dollars would have been $3,945 million.

Level of Sales by Month (December=100)1990-2007

70%

75%

80%

85%

90%

95%

100%

105%

Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec

Figure 1


Figure 2

Durables, Non-durables and Business Peak in December

7.0%7.5%8.0%8.5%9.0%9.5%

10.0%10.5%11.0%11.5%12.0%


DURABLESNONDURABLESBUSINESS

Sales by Month As Share of Yearly Sales--Major Categories

7.0%

7.5%

8.0%

8.5%

9.0%

9.5%

10.0%

10.5%

11.0%

11.5%

12.0%


AUTOMOBILEDURABLESTOURISMNONDURABLESBUILDINGBUSINESSTOTAL

Figure 3


Table 1. Average Retail Sales by Type and Month 1990-2007 (Current Million $) Month Automobile Durables Tourism Non-durables Building Business Total

January

$3,177.71

$1,423.48 $3,676.64 $5,167.27 $1,191.59 $3,232.54 $17,869.23

February

3,178.28

1,373.55 3,994.17 5,410.26 1,155.43 3,249.70 18,361.40

March

3,566.88

1,552.12 4,475.01 6,187.22 1,279.98 3,600.44 20,661.64

April

3,372.62

1,425.72 3,936.75 5,594.56 1,283.79 3,348.02 18,961.47

May

3,430.79

1,422.09 3,593.70 5,519.22 1,272.96 3,382.17 18,620.92

June

3,249.61

1,475.54 3,636.67 5,682.16 1,259.01 3,560.29 18,863.28

July

3,312.64

1,397.90 3,691.14 5,247.87 1,231.45 3,267.88 18,148.89

August

3,388.90

1,371.35 3,389.29 5,396.48 1,202.98 3,318.56 18,067.55

September

3,153.53

1,386.97 3,179.81 5,472.77 1,178.22 3,471.69 17,842.98

October

3,326.60

1,387.88 3,474.90 5,267.06 1,294.03 3,359.11 18,109.59

November

3,162.76

1,514.69 3,512.96 5,919.14 1,181.94 3,349.87 18,641.36

December

3,257.87

1,956.13 4,191.31 8,173.48 1,275.45 3,795.73 22,649.97

Monthly Average

$3,298.18

$1,473.95 $3,729.36 $5,753.12 $1,233.90 $3,411.33 $18,899.86


Table 2. Average Retail Sales by Type and Month 1990-2007 (Constant 1983 Million $)

Month Automobile Durables Tourism Non-durables Building Business Total

January

$1,758.49

$778.06 $2,874.49 $2,054.07 $648.48 $1,773.29 $9,886.87

February

1,749.35

747.18 3,013.82 2,225.83 635.46 1,781.92

10,153.56

March

1,952.52

844.26 3,407.51 2,471.04 696.61 1,957.54

11,329.49

April

1,853.78

775.57 3,097.40 2,167.66 697.64 1,819.70

10,411.76

May

1,883.52

769.21 3,046.77 1,978.58 691.33 1,833.64

10,203.04

June

1,779.51

799.33 3,122.03 1,992.01 685.82 1,928.82

10,307.52

July

1,814.75

758.86 2,891.12 2,028.63 666.35 1,769.19 9,928.90

August

1,847.67

743.37 2,980.44 1,864.75 651.28 1,794.75 9,882.26

September

1,723.00

749.64 3,000.11 1,743.40 638.07 1,877.08 9,731.29

October

1,818.42

752.88 2,882.91 1,896.50 700.47 1,814.96 9,866.15

November

1,729.22

817.87 3,236.40 1,919.60 638.95 1,813.42

10,155.45

December

1,777.51

1,059.21 4,484.99 2,305.57 694.70 2,064.37

12,386.35

Monthly Average

$1,807.31

$799.62 $3,169.83 $2,053.97 $670.43 $1,852.39

$10,353.55


Table 3. Retail Sales by Type and Month as Share of Yearly Sales 1990-2007 Month Automobile Durables Tourism Non-durables Building Business Total

January 8.0% 8.0% 8.2% 7.5% 8.0% 7.9% 7.9%

February 8.0% 7.8% 8.9% 7.8% 7.8% 7.9% 8.1%

March 9.0% 8.8% 10.0% 9.0% 8.6% 8.8% 9.1%

April 8.5% 8.1% 8.8% 8.1% 8.7% 8.2% 8.4%

May 8.7% 8.0% 8.0% 8.0% 8.6% 8.3% 8.2%

June 8.2% 8.3% 8.1% 8.2% 8.5% 8.7% 8.3%

July 8.4% 7.9% 8.2% 7.6% 8.3% 8.0% 8.0%

August 8.6% 7.8% 7.6% 7.8% 8.1% 8.1% 8.0%

September 8.0% 7.8% 7.1% 7.9% 8.0% 8.5% 7.9%

October 8.4% 7.8% 7.8% 7.6% 8.7% 8.2% 8.0%

November 8.0% 8.6% 7.8% 8.6% 8.0% 8.2% 8.2%

December 8.2% 11.1% 9.4% 11.8% 8.6% 9.3% 10.0%

Total 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0%


C. Retail Sales – An Analysis of Trends over Time

Analyzing time trends is also useful for estimating future retail sales. Table 4 in the following page shows the changes in the share of the retail dollar between 1990 and 2007. During this period the share of the retail dollar increased for autos and accessories from 15.3 percent to 17.1 percent. Durables also increased going from 6.6 percent to 8.2 percent. Tourism and recreation declined from 21 percent to 18.7 percent and consumer non-durables dropped from 32.1 percent to 28.4 percent. All other categories did not see much change in dollar share during this period.

The Share of Sales for Tourism Has Declined

18.0%

20.0%

22.0%

1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007

Figure 5

The Share of Sales for Automobiles Has Increased

13.0%

15.0%

17.0%

19.0%

1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007

Figure 4


Table 4. Retail Sales Shares by Type and Year 1990-2007 Year Automobile Durables Tourism Non-durables Building Business Total

1990 15.3% 6.6% 21.0% 32.1% 6.6% 18.3% 100.0%1991 15.1% 6.6% 21.7% 32.8% 5.7% 18.2% 100.0%1992 16.0% 6.5% 21.3% 32.7% 5.7% 17.7% 100.0%1993 16.7% 6.9% 20.5% 32.3% 6.1% 17.5% 100.0%1994 16.9% 7.0% 20.0% 32.4% 6.1% 17.6% 100.0%1995 17.3% 7.0% 19.9% 32.3% 6.0% 17.5% 100.0%1996 17.1% 7.3% 20.0% 32.4% 5.9% 17.3% 100.0%1997 17.0% 7.5% 20.1% 32.3% 5.8% 17.2% 100.0%1998 17.3% 8.0% 19.8% 31.4% 6.2% 17.3% 100.0%1999 17.8% 8.3% 19.7% 30.8% 6.3% 17.1% 100.0%2000 17.7% 8.3% 19.8% 30.8% 6.4% 17.0% 100.0%2001 18.5% 8.2% 19.4% 30.7% 6.2% 17.0% 100.0%2002 19.2% 8.2% 19.9% 28.7% 6.6% 17.4% 100.0%2003 19.1% 8.2% 19.3% 28.8% 6.5% 18.2% 100.0%2004 18.3% 8.3% 19.3% 28.1% 7.2% 18.7% 100.0%2005 18.0% 8.4% 18.7% 28.0% 7.6% 19.4% 100.0%2006 17.1% 8.2% 18.7% 28.4% 7.7% 19.8% 100.0%2007 16.3% 7.8% 19.6% 30.6% 6.4% 19.2% 100.0%

Average Share 17.3% 7.6% 19.9% 30.9% 6.4% 17.9% 100.0%

Share change 1990-2007 6.51% 17.79% -6.77% -4.73% -2.76% 5.18% 0.00%


D. Retail Sales -- Population Growth is Important in Forecasting Retail Sales

Table 5 on the following page shows the growth of total retail sales during the 18-year period. In real terms, total taxable retail sales (in 2007 dollars) increased from $207.2 billion in 1990 to $344 billion in 2006 before dropping to $321 billion in 2007, for a compounded growth rate of 2.6 percent per year – however about two-thirds growth was driven by Florida’s population growth. After adjusting for population growth, the real rate of growth was only ½ of 1 percent per year. While the overall rate of growth for the period was positive, there were years of decline or very small growth.

Since 1990 and 1991 were years of decline. During this period, sales decreased by 4.7 percent and when adjusted for population growth, the decline was a significant 6.7 percent. However, growth between 1991 and 1999 was very positive. During this 9-year period, retail sales grew at rates that averaged 4.9 percent—2.6 percent after adjusting for population. Between 2000 and 2001 retail sales cooled considerably and went into decline; the level of sales observed in 2000 was not seen again until 2005. After adjusting for population growth, the average growth rate was 4/10ths of 1 percent per year for the period between 2000 and 2007. The years 2001 and 2002 were particularly difficult for retail sales; during those two years, total sales declined by $11 billion in constant 2006 dollars.

Growth resumed in 2003. The years 2004 and 2005 marked the best growth period. The total inflation-adjusted growth rates were 7.5 percent and 7.8 percent respectively. Population- adjusted growth was in excess of 5 percent each of these two years. Growth slowed considerably in 2006 and further in 2007. Population-adjusted sales growth was slightly negative in 2006 (-0.6 percent) and a record setting -7.7 percent in 2007.

The negative growth rates in 2006 and 2007 are most notable because they occurred even as per capita incomes grew—albeit at a low 1 percent—for each of these 2 years. Previously, declines in per-capita sales had been concurrent with income declines. The evidence is highly suggestive of consumers retrenching to devote a larger share of their income to pay for debt or for savings.

E. Income Changes are Also a Key Driver of Retail Sales

Table 5 also illustrates the relationship between income and retail sales. On average between 50 to 55 cents of every dollar in per-capita income is spent on taxable items at the retail level. Generally speaking, during economic expansion periods, a higher share of income is spent on taxable retail sales.


Table 5. Retail Sales Growth 1990-2007 (Current Million $)

Inflation-Adjusted Growth Rates Year Total Sales

Current $ Total Sales

2006 $ Per Capita

Sales 2006 $ Per Capita

Income 2006 $ Sales Per

Capita Sales

Per Capita Income

1990 $ 134,013 $ 207,242 $ 15,840 $ 30,281 1991 $ 133,231 $ 197,593 $ 14,780 $ 29,677 -4.7% -6.7% -2.0%1992 $ 145,264 $ 209,056 $ 15,364 $ 29,773 5.8% 4.0% 0.3%1993 $ 156,773 $ 219,200 $ 15,799 $ 29,892 4.9% 2.8% 0.4%1994 $ 167,412 $ 228,120 $ 16,089 $ 30,148 4.1% 1.8% 0.9%1995 $ 177,450 $ 235,209 $ 16,256 $ 30,659 3.1% 1.0% 1.7%1996 $ 191,026 $ 245,997 $ 16,657 $ 30,982 4.6% 2.5% 1.1%1997 $ 203,063 $ 255,515 $ 16,951 $ 31,681 3.9% 1.8% 2.3%1998 $ 217,207 $ 269,143 $ 17,485 $ 32,657 5.3% 3.2% 3.1%1999 $ 236,048 $ 286,221 $ 18,154 $ 33,269 6.3% 3.8% 1.9%2000 $ 252,148 $ 295,779 $ 18,324 $ 33,479 3.3% 0.9% 0.6%2001 $ 255,124 $ 291,100 $ 17,657 $ 33,346 -1.6% -3.6% -0.4%2002 $ 252,906 $ 284,011 $ 16,848 $ 33,230 -2.4% -4.6% -0.3%2003 $ 266,336 $ 292,424 $ 16,923 $ 33,506 3.0% 0.4% 0.8%2004 $ 294,110 $ 314,478 $ 17,763 $ 34,743 7.5% 5.0% 3.7%2005 $ 327,788 $ 339,104 $ 18,718 $ 35,790 7.8% 5.4% 3.0%2006 $ 343,300 $ 344,063 $ 18,598 $ 36,337 1.5% -0.6% 1.5%2007 $ 329,916 $ 320,781 $ 17,172 $ 36,744 -6.8% -7.7% 1.1%

Average $ 220,776 $ 265,544 $ 16,953 $ 32,321 2.6% 0.5% 1.1%


Also, as income grows or contracts, retail sales adjust by a larger proportion. For example, per capita income dropped by 2 percent from 1990 to 1991 yet retail sales dropped by 6.7 percent on a per-capita basis. The following year per-capita income rose by .3 percent and per-capita sales rose by 4 percent. On average, sales changed three times as much as income. Nevertheless, in spite of sales year-to-year larger volatility over the 17-year period, overall per capita sales increased by 1 percent in real terms while income grew by 1.1 percent. Compounding the year-to-year difference in growth-rate yield a net difference of 1.01 percent between per capita income and sales. In the long term, growth rates for sales and per-capita income tend to converge.

F. Retail Sales—Cumulative Growth

Table 6 shows the cumulative effects of income on growth rates at the aggregate and per-capita (population-adjusted) levels. Setting the year 1990 as the baseline (100 percent), we see that through December of 2007 sales grew by a cumulative 54.8 percent, while per-capita sales grew by 8.4 percent. The difference (46.2 percentage points) is due to population growth (about 2.2 percent per year). This Table also shows that cumulative per-capita income growth was 21.3 percent. Thus the ratio of sales growth to income growth is .39.

Table 6. Relative Sales Levels 1990=100 Inflation Adjusted Relative Level Year

Sales Per Capita Sales Income 1990 100.0% 100.0% 100.0% 1991 95.3% 93.3% 98.0% 1992 100.9% 97.0% 98.3% 1993 105.8% 99.7% 98.7% 1994 110.1% 101.6% 99.6% 1995 113.5% 102.6% 101.2% 1996 118.7% 105.2% 102.3% 1997 123.3% 107.0% 104.6% 1998 129.9% 110.4% 107.8% 1999 138.1% 114.6% 109.9% 2000 142.7% 115.7% 110.6% 2001 140.5% 111.5% 110.1% 2002 137.0% 106.4% 109.7% 2003 141.1% 106.8% 110.6% 2004 151.7% 112.1% 114.7% 2005 163.6% 118.2% 118.2% 2006 166.0% 117.4% 120.0% 2007 154.8% 108.4% 121.3%


II. ECONOMETRIC ANALYSIS: FORECASTING EXPECTED CHANGES IN RETAIL SALES

The analysis of the trends in the data previously analyzed so far suggests that seasonal patterns, income and population changes are important determinants of retail level demand. Economic theory implies that, in addition, relative prices and consumer expectations are also important components of changes in retail sales.

Relative prices are important because consumers adjust consumption in a manner such that they will buy more of the goods that provide more value for the dollar to substitute for those that have become relatively more expensive. Consumers will buy more chicken as its price drops relative to beef—holding other things such as income and taste constant. Consumer expectations and habits also play a major role in their lives. For instance consumers may spend less in anticipation of tougher economic times, or consumers may continue to purchase goods because the short-run substitution is neither feasible nor desirable. Finally, other factors such as age distribution in a population will affect retail demand. For example, an area with a higher concentration of older individuals may have a lower demand for durables and a higher demand for medical services relative to an area with a younger population. In essence, the demographic and economic structure of a region is important for determining changes in the types of retail sales categories.

While incorporating consumer expectations, habits and demographics to the forecasting model is quite straightforward, incorporating prices adds a large degree of complexity to the model. This added complexity may not be warranted if we consider that while relative prices are important, they are more so when the purpose is estimating changes in demand for closely related items, either substitutes or complements2.

For the purposes of this Study we have aggregated retail sales into large groups such that items within the groups are better substitutes or complements than items between groups. This alleviates to a large extent the problem of omitting prices. Further, the main objective of this Study is to produce a forecasting model. Since relative price changes happen either over time or seasonally, the effect of prices will be embedded in the model’s variables that represent time or seasonality. This makes the coefficients for the variables a mix of the price

2Goods are said to be substitutes when the demand for one good changes in the same direction as the price of the other good. For example, the amount of Coke sold would increase if the price of Pepsi increases. Goods are complements when the demand for one good changes in opposite direction as the price of the other good. For example, as the price of gas increases we would expect the demand for gas consuming autos to decrease.


and other effects. However, to the extent that the level of substitutability or complementarities is limited, these biases will be limited.

To allow for consumer expectations, habits and demographic characteristics that may influence demand, the Study develops two proxy variables: The Index of Consumer Sentiment and Lagged Retail Sales (LRS). The Index of Consumer Sentiment is published monthly by the University of Michigan and focuses on three areas: How consumers view prospects for their own financial situation, the general economy over the near term and for the economy over the long term. The Consumer Sentiment Index has proven to be a reliable leading indicator of consumer behavior3. The use of LRS serves the purpose of reflecting that past demand is a good predictor of current demand because habits, tastes and demographics change slowly when aggregated over populations.

When each of the time series comprising a retail sales level category is analyzed in isolation, the statistical quality of the estimates is quite good, and the signs and magnitude of the variable coefficients are highly plausible.

The estimation procedure used for this analysis used linear regression utilizing the log-log specification and categorical variables for the seasonal effects. An advantage of the log-log specification is that the coefficients of the continuous explanatory variables in the model can be interpreted as elasticities, i.e. the percent change in sales due to a one percent change in the variable.

The seven-estimated sector specific equations are thus of the following form:

3 According to its publisher “The Surveys of Consumers have proven to be an accurate indicator of the future course of the national economy. The Index of Consumer Expectations, produced by the Surveys of Consumers, is included in the Leading Indicator Composite Index published by the U.S. Department of Commerce, Bureau of Economic Analysis. The inclusion of data from the Surveys of Consumers by the Commerce Department is a significant confirmation of its capabilities for understanding and forecasting changes in the national economy. Each series included in the composite Index of Leading Indicators is selected because of its performance on six important characteristics: economic significance, statistical adequacy, consistency of timing at business cycle peaks and troughs, conformity to business expansions and contractions, smoothness, and prompt availability. No other consumer survey meets these rigorous criteria.” http://www.sca.isr.umich.edu/main.php


A. Results of the Regression Analysis

The results of the estimation for each of the retail sales categories are presented in Table 7 on the following page. The coefficients for the month variables represent percent difference in the month’s sales with respect to December sales—holding all other variable levels constant (December ceteris paribus.) For example, total sales were 30 percent lower, on average, in January than in December ceteris paribus i.e., assuming that there were no differences in any other factor.


Real per capita income has larger effects on goods that are often bought on medium or long-term credit. For example, the effect of income on consumer non-durables is .437, meaning that a one percent change in income results in a change of 4.3 tenths of one percent in sales. The coefficient for durables is .546 and for construction 1.849. The effect of population is larger on autos than for any other category. Consumer Sentiment seems to be more important for consumer durables and autos than for other categories. These are items whose purchase can often be easily postponed in anticipation of tough times ahead.

The use of a Lagged Retail Sales variable allows us to compute long-run effects. The rationale is that a change in income in, say January 1990, will affect sales in January 1990

Table 7. Forecast Estimation Equation Coefficients Sector Forecast Equation

Variable Total Auto

Non-durables Durables Tourism Building Business

Constant intercept -1.530 -7.433 4.297 -5.388 2.746 -5.196 -1.110 Real per capita income 0.496 N.S. 0.437 0.546 0.600 1.849 1.004 Consumer sentiment 0.106 0.261 0.171 0.203 0.088 N.S. N.S. Population 0.401 1.377 0.200 0.618 0.272 N.S. 0.263 January -0.304 N.S. -0.591 -0.402 -0.183 -0.069 -0.185 February -0.170 N.S. -0.326 -0.292 -0.057 -0.059 -0.118 March -0.078 0.110 -0.232 -0.154 N.S. N.S. -0.029 April -0.219 N.S. -0.387 -0.304 -0.171 N.S. -0.145 May -0.197 0.052 -0.357 -0.266 -0.203 N.S. -0.105 June -0.180 N.S. -0.328 -0.232 -0.157 N.S. -0.062 July -0.220 N.S. -0.413 -0.298 -0.141 N.S. -0.170 August -0.205 0.034 -0.344 -0.291 -0.231 -0.058 -0.118 September -0.217 -0.036 -0.351 -0.268 -0.260 -0.072 -0.080 October -0.194 0.042 -0.390 -0.262 -0.147 N.S. -0.136 November -0.178 -0.034 -0.261 -0.195 -0.177 -0.110 -0.126

Lagged sales 0.536 0.397 0.518 0.575 0.464 0.409 0.517

Regression DiagnosticsR-Square 0.985 0.955 0.952 0.981 0.980 0.927 0.973 Durbin-Watson 2.327 2.223 2.278 2.361 2.203 2.289 2.297 Forecast Error 95 percentile 0.045 0.098 0.069 0.076 0.054 0.140 0.063 * Coefficients in shaded cells were not statistically different from zero (p>.05) The models have the log-log specification. Explanatory variables are natural logarithms, except for the categorical variables for the month and the Consumer Sentiment Index which is a percent. The coefficients for the month variables are differences relative to December. The "forecast error 95 percentile" refers to the size of the percentile of the distribution of the absolute value of the forecast errors. For example for the "Total" retail sales equation, 95 percent of the times the forecast error was less than 4.5 percent


(short-run effect) and also in successive months. This long run effect is embedded in the lagged variable. For example, the model for February 1990 sales uses income for that month as an explanatory variable, and it also uses the sales for January of 1990 as an explanatory variable (the lagged variable). The January 1990 sales are explained by income in January 1990 and in turn, January’s income gets carried as an explanatory variable for sales in February.

Table 8 has the long run effects of the continuous variables showing that the long-run effect (cumulative) of income on sales is 1.069 percent for a percent change in income. Also, the long-run effect of population is .863 and consumer sentiment is .227.

Table 8. Estimated Long-Run Marginal Effects on Retail Sales*

Sector Forecast Equation Variable Total Auto

Non-durables Durables Tourism Building Business

Real per capita income 1.069 N.S. 0.906 1.286 1.120 3.126 2.076 Consumer sentiment 0.227 0.433 0.354 0.478 0.165 N.S. N.S. Population 0.863 2.285 0.414 1.456 0.509 N.S. 0.545 * The long-run effect is derived by dividing the estimated coefficient for the variable by (1-coefficient of lagged sales)


III. FORECASTING EXPECTED CHANGES IN RETAIL SALES—SECTOR ANALYSIS

Equations 1-7 can be used to forecast the level of taxable sales on a month-by-month basis. The forecasts are dependent on the assumptions we make about population and income growth as well as consumer sentiment. Further, as with any forecast, its reliability declines the further out we go in time. For the following forecasts the following assumptions are made:

a) Inflation growing at 4 percent per year.

b) Population growing at 1.5 percent per year.

c) Real income growing at 1.5 percent per year.

d) Consumer Sentiment remains at 70.8 for the forecasting period.

Assumptions a) through c) are in keeping with the Florida Legislature's forecasts; assumption d) carries forward the last observed value for this variable.

Table 9 summarizes the taxable sales forecast for all sectors for 2008. Overall total sales are forecast to decline by 3.0 percent, following the 6.8 percent decline of 2007 relative to 2006. Therefore, the cumulative decline for the 2-year period totals close to 10 percent. Since 1990, there was another instance of back-to-back taxable sales declines (2001 and 2002), but the decline then was 4 percent over the 2-year less than half the size of the projected 2-year decline for 2007 and 2008.

Table 9. Taxable Sales Forecasts Current Dollars by Sector- 2008

Sector Sales Forecast ($ Million)

Change from 2007 (Inflation Adjusted)

Non-durables $ 98,055 -5.97% Durables 25,857 -2.97% Tourism and Recreation 66,871 0.20% Business Investment 65,414 -0.05% Auto 54,731 -1.58% Building Investment 22,119 -1.0% Total (Weight Adjusted by Categories) $333,057 -3.00%

The forecast predicts declines for 4 of the 6 sectors (Durables, Non-durables, Auto sectors and Business Investment). It shows modest gains for Tourism and Building Investment.


On a month-by-month basis the model forecasts declines in 2008 relative to the same month in 2007 through the middle of the year and only slight increases for the second half of 2008.

Table 10 shows that for 2009, the model forecasts only a modest recovery with overall sales growth of 2.75 percent relative to 2008. The cumulative effect of 2007 and 2008 losses with the gain of 2009 would leave the taxable sales level where it was in 2004, after adjusting for inflation.

Table 10. Taxable Sales Forecasts Current Dollars by Sector- 2009

Sector Sales Forecast ($ Million)

Change from 2008 (Inflation Adjusted)

Non-durables $ 100,522 1.46% Durables 29,373 8.76% Tourism and Recreation 68,596 -1.32% Business Investment 70,822 4.95% Auto 65,633 12.31% Building Investment 23,972 6.15% Total (Weight Adjusted by Categories) $358,919 2.75%

According to the forecast model Non-durables see further declines in 2009, Tourism shows a small decline, and Durables, Business, Auto and Building Investment show gains, but from the low base of 2008 (i.e., a recovery, not expansion).

Total Forecasted 2008 Taxable Sales Month Change From 2007

-15%

-10%

-5%

0%

5%

10%

15%

Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Year

Figure 6


Figure 7 shows that the month over previous year month growth in sales is larger earlier in the year. This is because the months from January through June of 2009 are compared to the same months from 2008 which showed dismal levels of taxable retail sales.

Table 10b. Taxable Sales Forecast—Total ($ Million) 2008 2009

Month Taxable

Sales Change from

Year Ago1 Taxable

Sales Change from

Year Ago1 January $ 26,112 -10.77% $ 29,827 9.79%February $ 26,326 -7.42% $ 28,857 4.43%March $ 28,957 -11.34% $ 33,939 11.90%April $ 26,799 -7.58% $ 29,655 5.93%May $ 27,705 -2.34% $ 29,702 2.76%June $ 28,286 -0.32% $ 30,024 1.11%July $ 26,941 -0.20% $ 28,487 0.25%August $ 26,672 1.34% $ 28,159 -0.20%September $ 26,455 2.89% $ 27,862 -0.43%October $ 27,312 3.74% $ 28,738 -0.53%November $ 28,043 4.24% $ 29,467 -0.58%December $ 32,602 4.32% $ 34,202 -0.59%

Total $ 332,213 -2.13% $ 358,919 2.75%1 Inflation adjusted

Total Forecasted 2009 Taxable Sales Month Change From 2008

-15%

-10%

-5%

0%

5%

10%

15%

Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Year

Figure 7.


Durables: The estimated coefficients for the forecasting equation are:

Table 11a. Estimated Coefficients—Durables

Variable Coefficient Estimate

Constant intercept -5.388 Real monthly per capita income 0.546 Consumer sentiment percent 0.203 Population 0.618 January -0.402 February -0.292 March -0.154 April -0.304 May -0.266 June -0.232 July -0.298 August -0.291 September -0.268 October -0.262 November -0.195 Lagged sales 0.575

The month-by-month forecast is:

Table 11b. Taxable Sales Forecast—Consumer Durables 2008 2009

Month Taxable

Sales Change from

Year Ago1 Taxable

Sales

Change from Year

Ago1 January $ 1,909 -19.33% $ 2,573 29.72%February 1,892 -17.63% $ 2,373 19.91%March 2,026 -16.12% $ 2,505 18.28%April 1,907 -11.56% $ 2,337 17.23%May 2,163 -3.62% $ 2,479 9.63%June $ 2,148 -2.60% $ 2,384 6.19%July $ 2,081 0.84% $ 2,266 4.17%August $ 2,089 4.35% $ 2,247 2.93%September $ 2,134 5.77% $ 2,278 2.14%October $ 2,103 7.10% $ 2,232 1.60%November $ 2,478 11.35% $ 2,619 1.20%December $ 2,922 8.50% $ 3,079 0.88%



Non- Durables: The estimated coefficients for the forecasting equation are:

Table 12a. Estimated Coefficients—Non-durables Variable Coefficient

Estimate Constant intercept 4.297 Real monthly per capita income 0.437 Consumer sentiment percent 0.171 Population 0.200 January -0.591 February -0.326 March -0.232 April -0.387 May -0.357 June -0.328 July -0.413 August -0.344 September -0.351 October -0.390 November -0.261 Lagged sales 0.518


Table 12b. Taxable Sales Forecast—Consumer Non-durables 2008 2009

Month Taxable Sales

Change from Year Ago1

Taxable Sales


January $ 7,722 -7.44% $ 7,513 -6.36%February $ 7,894 -4.59% $ 7,669 -7.08%March $ 8,904 -10.45% $ 9,558 2.67%April $ 7,687 -8.01% $ 7,804 -2.89%May $ 7,885 -4.86% $ 8,073 -2.06%June $ 8,138 -5.75% $ 8,417 -1.05%July $ 7,359 -8.05% $ 7,649 -0.55%August $ 7,139 -7.18% $ 7,436 -0.32%September $ 7,773 -3.72% $ 8,103 -0.23%October $ 7,537 -5.22% $ 7,858 -0.21%November $ 8,521 -5.17% $ 8,881 -0.22%December $ 11,097 -5.60% $ 11,560 -0.26%

Total $ 97,655 -6.36% $ 100,522 -1.46%1 Inflation adjusted


Tourism & Recreation: The estimated coefficients for the forecasting equation are:

Table 13a. Estimated Coefficients—Tourism and Recreation


Constant intercept 2.746 Real monthly per capita income 0.600 Consumer sentiment percent 0.088 Population 0.272 January -0.183 February -0.057 March 0.00N.S.

April -0.171 May -0.203 June -0.157 July -0.141 August -0.231 September -0.260 October -0.147 November -0.177 Lagged sales 0.464


Table 13b. Taxable Sales Forecast—Tourism and Recreation 2008 2009

Month Taxable Sales


Taxable Sales


January $ 5,947 7.04% $ 5,808 -6.02%February $ 5,965 0.40% $ 6,036 -3.22%March $ 6,840 -0.78% $ 7,098 -0.75%April $ 5,923 -1.59% $ 6,072 -1.94%May $ 5,291 -1.33% $ 5,443 -1.59%June $ 5,519 -0.83% $ 5,732 -0.64%July $ 5,275 -3.26% $ 5,500 -0.23%August $ 4,926 -1.24% $ 5,144 -0.08%September $ 4,604 -0.83% $ 4,808 -0.05%October $ 5,143 -0.32% $ 5,369 -0.07%November $ 5,111 -0.65% $ 5,332 -0.12%December $ 5,997 -0.15% $ 6,252 -0.18%

Total $ 66,542 -0.29% $ 68,596 -1.32%1 Inflation adjusted


Building Investment: The estimated coefficients for the forecasting equation are:

Table 14a. Estimated Coefficients—Building Investment


Constant intercept -5.196 Real monthly per capita income 1.849 Consumer sentiment percent -0.040N.S

Population 0.242 N.S January -0.069 February -0.059 March 0.037 N.S April -0.001 N.S May -0.019 N.S June -0.103 N.S July -0.046 N.S August -0.058 September -0.072 October 0.030 N.S November -0.110 Lagged sales 0.409


Table 14b. Taxable Sales Forecast—Building Investment 2008 2009

Month Taxable Sales


Taxable Sales


January $ 1,716 -16.24% $ 2,243 25.77%February $ 1,552 -11.67% $ 1,797 10.73%March $ 1,677 -16.92% $ 2,217 26.42%April $ 1,804 -7.55% $ 2,118 12.30%May $ 1,903 3.74% $ 2,033 2.24%June $ 1,922 6.76% $ 2,027 0.93%July $ 1,940 3.89% $ 2,035 0.37%August $ 1,823 2.30% $ 1,907 0.10%September $ 1,730 4.28% $ 1,807 -0.04%October $ 1,959 6.62% $ 2,045 -0.13%November $ 1,808 6.85% $ 1,885 -0.20%December $ 1,784 5.48% $ 1,859 -0.26%



Business Investment: The estimated coefficients for the forecasting equation are:

Table 15a. Estimated Coefficients—Business Investment


Constant intercept -1.110 Real monthly per capita income 1.004 Consumer sentiment percent 0.00 N.S Population 0.263 January -0.185 February -0.118 March -0.029 April -0.145 May -0.105 June -0.062 July -0.170 August -0.118 September -0.080 October -0.136 November -0.126 Lagged sales 0.517


Table 15b. Taxable Sales Forecast—Business Investment 2008 2009

Month Taxable Sales


Taxable Sales


January $ 4,921 -12.07% $ 5,867 14.75%February $ 4,880 -8.71% $ 5,574 9.24%March $ 5,180 -13.60% $ 6,462 19.32%April $ 5,107 -7.42% $ 5,906 10.62%May $ 5,539 -0.33% $ 6,025 4.06%June $ 5,766 1.34% $ 6,160 2.20%July $ 5,524 3.27% $ 5,844 1.21%August $ 5,368 2.24% $ 5,648 0.67%September $ 5,549 4.08% $ 5,819 0.36%October $ 5,537 5.80% $ 5,794 0.16%November $ 5,344 5.53% $ 5,583 0.02%December $ 5,886 6.77% $ 6,142 -0.09%



Autos & Accessories: The estimated coefficients for the forecasting equation are:

Table 16a. Estimated Coefficients—Autos and Accessories


Constant intercept -7.433 Real monthly per capita income -0.383 N.S

Consumer sentiment percent 0.261 Population 1.377 N.S January 0.016 N.S February 0.004 March 0.110 April 0.017 N.S May 0.052 June -0.021 N.S July 0.021 N.S August 0.034 September -0.036 October 0.042 November -0.034 Lagged sales 0.397


Table 16b. Taxable Sales Forecast—Autos and Accessories 2008 2009

Month Taxable Sales


Taxable Sales


January $ 3,897 -19.37% $ 5,823 43.79%February $ 4,142 -14.05% $ 5,408 24.85%March $ 4,331 -19.75% $ 6,100 34.72%April $ 4,371 -8.71% $ 5,418 18.56%May $ 4,924 1.17% $ 5,649 9.74%June $ 4,793 5.98% $ 5,303 5.86%July $ 4,761 6.74% $ 5,193 4.36%August $ 5,327 11.09% $ 5,776 3.77%September $ 4,665 9.31% $ 5,047 3.54%October $ 5,034 12.82% $ 5,440 3.45%November $ 4,782 14.87% $ 5,166 3.42%December $ 4,917 16.18% $ 5,310 3.40%

Total $ 55,944 0.60% $ 65,633 12.31%1 Inflation adjusted


Figure 8 Source: Econometric Modeling, The Washington Economics Group, Inc.

Factors Explaining Sales Variability Across Time

Other2%

Month and Consumer Sentiment

11%

Population and Income

87%


APPENDIX I

INFORMATION REGARDING FLORIDA TAXABLE SALES DATA

SOURCE OF DATA- These data come to the state on sales tax returns filed monthly by retail establishments with the Florida Department of Revenue. Taxable sales are those sales subject to Chapter 212, Florida Statutes. Generally speaking, the sales tax is levied on sales of goods, but not services, although there are some taxable services. Major categories of exempt sales are food not prepared for immediate consumption, medical and legal services, residential utilities, items purchased for resale, intangible personal property, and rentals over six months. It is estimated that taxable sales comprise 40-45% of all retail sales. In 2006 taxable sales comprised 54% of the state's GDP.

TIMING- Sales taxes are due on the 20th of each month for the preceding calendar month. Taxable sales are reported directly on the return and these data are entered into the Department's computer system, a process which consumes several weeks. Thus, as an example, data on taxable sales transactions which took place in January would be sent in on the return due in February, and available as data toward the latter part of March. The data in these tables are dated according to the month of transaction.

REVISIONS- After the initial release of the data by the Department of Revenue, it undergoes an audit process to check for errors. Error checking involves making sure the data from the return is internally consistent. This process usually takes six to eight weeks, when the revised data is released. A notation indicating whether the data is preliminary or revised is at the bottom of the monthly tables.

USING THE DATA- Care should be taken when using un-audited data, especially for MSAs that are not large. Users may expect a typical revision between the un-audited and audited data on the order of 4-5%. Even after going through the audit process, data for small MSAs can be quite volatile from month to month, with a high degree of seasonality as well as what appears to be random noise. Users are cautioned against making inferences based on less than three or four months of data, even when allowing for seasonality.

CATEGORIES OF TAXABLE SALES- When a business applies for a sales tax license, it is classified according to its main type of business. There are 90+ "kinds" of business in the Department of Revenue's classification scheme. All taxable sales for a business are attributed to its "kind", regardless of what actually is sold in the transaction. The categories of taxable sales in the tables are groupings of data from related "kinds" of businesses. For example, grocery stores are included under the "non-durables" category while liquor stores are included under the "tourism and recreation" category. A grocery store that sells liquor and remits a single tax return for all its business activity would have its liquor sales classified as non-durables sales in the data. Another example would be an appliance sold at a building supply center. The sale would be counted in the "construction" category rather than the "consumer durables" category. The following paragraphs describe the main kinds of business activity that are included within each category in the taxable sales tables:

Autos & Accessories- The category of "automobiles and accessories" taxable sales includes the sale of new and used cars, repair shops, auto supply stores, and taxable sales at gasoline stations.

Consumer Durables- The category of "consumer durables" taxable sales includes the sale of appliances, furniture, home electronics, aircraft, boat dealers, hardware and decorating stores.

Tourism & Recreation- The category of "tourism and recreation" taxable sales includes hotels and motels, bar and restaurant sales, liquor stores, photo and art stores, gift shops, admissions, sporting goods, rentals and jewelry stores.

Consumer Non-durables- The category of "consumer non-durables" taxable sales includes food and convenience stores, department and clothing stores, drug stores, antique dealers, bookstores, florists, pet dealers and suppliers, social organizations, storage, communications firms, print shops, nurseries, vending machines, utilities, and any "kind" that does not fit into the other categories.

Building Investment- The category of "building investment" taxable sales includes sales by building contractors, heating and air conditioning contractors, insulation, well drilling, electrical contractors, interior decorating, paint and wallpaper shops, cabinet and woodworking shops, soil, lumber and building suppliers, and roofing contractors. Services provided by these businesses are not generally taxable.

Business Investment- The category of "business investment" taxable sales includes farm equipment, feed and seed suppliers, store and office equipment, computer shops, machine shops, industrial machinery, hotel and restaurant suppliers, transportation equipment, manufacturing and refining equipment, industrial suppliers, paper and packaging materials, medical and optical supplies, commercial rentals, and wholesale dealers. Transactions reported as subject to the "use" tax are also included here, regardless of the kind code of the business reporting the "use" tax.

Retail Index- The "index of retail activity" is designed to provide a measure of retail activity for an area and allow comparisons with other areas over time. The index is constructed as an attempt to smooth the volatility in the taxable sales data and thereby allow unaided comparisons of one MSA to another on a monthly basis. The index is constructed by aggregating the categories of autos and accessories, other durables, tourism and recreation, and consumer non-durables. These categories represent the bulk of non-investment spending and is analogous to personal consumption. The sum of these four categories is seasonally adjusted and a four-month-moving average is taken.

The resulting data series is indexed such that the base period of December 1988 equals 100. Each MSA is measured against itself, that is, each MSAs index equals 100 in the base period and is calculated independently of activity in other MSAs. The index values can be directly read as percentages from the base period. An MSA with an index of 200 would have taxable sales in the four categories measured equal to twice the base period, for example, or a 100% increase (200 - 100 = 100%) from the base period. Likewise, an MSA with an index of 300 would have taxable sales equal to three times the base period, for a 200% increase (300 - 100 = 200%). The second MSA could be said to have grown twice as fast as the first MSA (200% / 100% = 2) from the base period, at least as measured by the index.

MSA DESCRIPTIONS

The MSAs in the tables are comprised of the following counties:

Daytona Beach: Flagler, Volusia Fort Lauderdale: Broward Fort Myers: Lee Fort Pierce: Martin, St. Lucie Fort Walton: Okaloosa Gainesville: Alachua Jacksonville: Clay, Duval, Nassau, St. Johns Lakeland: Polk Melbourne: Brevard Miami: Dade Naples: Collier Ocala: Marion Orlando: Lake, Orange, Osceola, Seminole Panama City: Bay Pensacola: Escambia, Santa Rosa Punta Gorda: Charlotte Sarasota: Manatee, Sarasota


APPENDIX II

THE WASHINGTON ECONOMICS GROUP, INC. PROJECT TEAM

J. ANTONIO “TONY” VILLAMIL

Dean, School of Business of St. Thomas University of Miami and Principal Advisor

The Washington Economics Group, Inc.

Tony Villamil has over thirty years of successful experience as a business economist, university educator and high-level policymaker at both federal and state governments. He has served as a Presidential appointed U.S. Undersecretary of Commerce for Economic Affairs, and is the founder of a successful economic consulting practice, The Washington Economics Group, Inc. (WEG). Since August 2008, Tony is the Dean of the School of Business of St. Thomas University of Miami, while continuing to serve as senior advisor to the clients of WEG. Tony is a member of the President’s Advisory Committee on Trade Policy and Negotiations in Washington, D.C. He is the immediate past Chairman of the Governor’s Council of Economic Advisors of Florida, and during 1999-2000, he directed the Tourism, Trade and Economic Development activities of the State in the Office of Governor Jeb Bush. Presently, he is on the Board of Directors of the Spanish Broadcasting System (NASDAQ), Mercantil Commercebank, N.A. and Enterprise Florida – the State’s principal economic development organization. Among other leadership positions, he served in 2008 as the economist of the Constitutionally mandated Tax and Budget Reform Commission of Florida (TBRC), and is currently Chairman of the Economic Roundtable of the Beacon Council – Miami-Dade County’s official economic development organization. He is also a Senior Research Fellow of Florida TaxWatch, an established fiscal and policy research organization of the State. After winning the gubernatorial election in November 2006, then Governor-elect Charlie Crist appointed him as his Economic Advisor during the transition period. Tony earned Bachelor and advanced degrees in Economics from Louisiana State University (LSU), where he also completed coursework for the Ph.D. degree. In 1991, Florida International University (FIU) awarded him a doctoral degree in Economics (hc), for “distinguished contributions to the Nation in the field of economics.” He speaks frequently to business, government and university audiences on economic topics, and was until the summer of 2008 a member of the Business Faculty of Florida International University (FIU).

HORACIO SOBERON-FERRER, PH.D. Associate Consultant for Economics

The Washington Economics Group, Inc. Horacio Soberon-Ferrer, Ph.D. has over twenty years of experience as a professional economist. His expertise is in applied microeconomics, consumer policy, demand forecasting, health care systems analysis and the economic of aging. He has held the positions of Director of Planning and Evaluation, Florida Department of Elder Affairs; Director of the State Infrastructure Bank at the Florida Department of Transportation and Senior Analyst for Forecasting and Environmental Scanning, AARP. He has also held full-time faculty positions at the University of Florida and University of Maryland where he taught Consumer Economics, Public Policy Analysis, Finance, and Statistical Methods. He received his Ph.D. in Applied Economics from Clemson University specializing in quantitative policy analysis and consumer behavior. He also has a Licentiate degree in Actuarial Science and a M.S. in Management. Horacio has published widely on the topics of consumer expenditures estimation, economics of energy demand and the economics of aging.

The Washington Economics Group, Inc. (WEG) has been successfully meeting client objectives since 1993 through economic consulting services for corporations, institutions and governments of the Americas. We have the expertise, high-level contacts, and business alliances to strengthen your competitive positioning in the growing marketplaces of Florida and Latin America. Our roster of satisfied clients, over the past fourteen years, includes multinational corporations, financial institutions, public entities, and non-profit associations expanding their operations in the Americas.

EXCLUSIVE CONSULTING APPROACH:

Each client is unique to us. We spend considerable time and effort in understanding the operations, goals, and objectives of clients as they seek our consulting and strategic advice. We are not a mass-production consulting entity nor do we accept every project that comes to us. We engage a limited number of clients each year that require customized consulting services in our premier areas of specialization. These premier and exclusive services are headed by former U.S. Under Secretary of Commerce, Dr. J. Antonio Villamil, with over twenty-five years of experience as a business executive and as a senior public official of the U.S. and most recently of Florida.

PREMIER CONSULTING SERVICES:

Comprehensive Corporate Expansion Services. Our seamless and customized service includes site selection analysis, development of incentive strategies and community and governmental relations. Economic Impact Studies highlight the importance of a client's activities in the generation of income, output and employment in the market area serviced by the entity. These studies are also utilized to analyze the impact of public policies on key factors that may affect a client's activities such as tax changes, zoning, environmental permits and others. Strategic Business Development Services. These services are customized to meet client objectives, with particular emphasis in the growing marketplaces of Florida, Mexico, Central and South America. Recent consulting assignments include customized marketing strategies, country risk assessments for investment decisions and corporate spokesperson activities and speeches on behalf of the client at public or private meetings.

For a full description of WEG capabilities and services, please visit our website at:

www.weg.com

Representative Client List 1993-2008

MULTINATIONAL CORPORATIONS

• Lockheed Martin • FedEx Latin America • IBM • Motorola • SBC Communications • Ameritech International • Lucent Technologies • MediaOne/AT&T • Joseph E. Seagram & Sons, Inc. (Vivendi) • Microsoft Latin America • Carrier • Medtronic • Phelps Dodge • Esso Inter-America • Visa International • MasterCard International • Telefonica Data Systems • Bureau Veritas (BIVAC) • Merck Latin America • DMJM & Harris • Wilbur Smith Associates • PBSJ

FLORIDA-BASED CORPORATIONS

• Sprint of Florida • Florida Marlins • Flo-Sun Sugar Corp. • Farm Stores • The BMI Companies • Spillis Candela & Partners • The Biltmore Hotel/Seaway • Trammel Crow Company • Advantage Capital • WCI Development Companies • Iberia Tiles • Florida Hospital • Mercy Hospital • The St. Joe Companies • Florida Power & Light (FPL) • International Speedway Corporation

LATIN AMERICA-BASED INSTITUTIONS • Federation of Inter-American Financial Institutions

(FIBAFIN) • The Brunetta Group of Argentina • Association of Peruvian Banks • Peruvian Management Institute (IPAE) • Mercantil Servicios Financieros, Venezuela • Allied-Domecq, Mexico • Fonalledas Enterprises

FINANCIAL INSTITUTIONS

• International Bank of Miami • Pan American Life • ABN-AMRO Bank • Barclays Bank • Lazard Freres & Co. • Banque Nationale de Paris • HSBC/Marine Midland • Fiduciary Trust International • Sun Trust Corporation • First Union National Bank (Wachovia) • Union Planters Bank of Florida (Regions) • Bank Atlantic Corp. • Hemisphere National Bank • BankUnited, FSB • Mercantil Commercebank N.A. • PointeBank, N.A. • The Equitable/AXA Advisors

PUBLIC INSTITUTIONS, NON-PROFIT ORGANIZATIONS & UNIVERSITIES

• Baptist Health Systems • Jackson Health Systems • Miami-Dade Expressway Authority • Miami-Dade College • Miami Museum of Science • Zoological Society of Florida • Florida International University • University of Miami • Inter-American Development Bank (IDB) • United Nations Economic Development Program (UNDP) • Universidad Politécnica de Puerto Rico • Sistema Universitario Ana G. Méndez (SUAGM) • Keiser University • Full Sail Real World Education • Florida Retail Federation • Florida Ports Council • Florida Sports Foundation • Florida Citrus Mutual • Florida Nursing Homes Alliance • Florida Bankers Association • Florida Outdoor Advertising Association • City of Plantation • City of West Palm Beach • Econ. Dev. Commission of Lee County • Econ. Dev. Commission of Miami-Dade (Beacon Council) • Econ. Dev. Commission of Mid-Florida • Jacksonville Chamber of Commerce • SW Florida Regional Chamber of Commerce • Enterprise Florida, Inc. • The Beacon Council • Visit Florida • Louisiana Committee for Economic Development • University of South Florida/ENLACE • Space Florida • State of Florida

estimation and forecasting retail sales in florida

Documents

intangible

washington

capita incomes

air conditioning

taxable sales

taxable sales

taxable sales

taxable sales