Articles and Updates

Sales Tax Increase vs. Expenditure Cuts: An Economic Impact Study

March 19, 2010

By Alberta H. Charney, Ph.D.

Introduction

In order to balance Arizona’s budget for fiscal year (FY) 2011, a temporary one cent increase in the sales tax has been proposed and will be voted on in a May 18, 2010 referendum. The Governor has proposed a list of “conditional enactments” (budget cuts) that will take effect if the referendum does not pass.

The purpose of this study is to assess and then compare the economic impacts of these two alternative scenarios. In the first case the temporary one percent sales tax increase is implemented, and in the alternative scenario the sales tax referendum is not approved and the proposed conditional enactments take effect. In Section 2, the research method and the model used for the impact analysis are described. The results of the economic impact analysis for the tax increase are presented in Section 3, and the impact analysis results for the proposed expenditure cuts in Section 4. A brief summary and conclusions are in Section 5.

Research Method for Impact Analysis

Information on both the sales tax increase, by tax category, and on government spending cuts, by agency, were inputted into an input-output (I-O) model to estimate the economic impacts of the one percent increase in the sales tax and the proposed conditional enactments. An I-O model is a collection of mathematical relationships that represent the interrelationships within an economy.

In the present analysis, the IMPLAN (Impact Analysis for PLANning) I-O model for Arizona is used. This model consists of a social accounting matrix, in which transactions between producers, intermediate and final consumers are represented. This matrix is used to assess the economic impact of purchases made within a region.

For example, an increase in demand for the output of a particular industry requires that industry to purchase additional goods and services, some of which are produced in-state. When the local demand for those inputs increases, producers of those inputs must also increase their demand for goods and services, some of which are produced in-state.

At each round of purchases there is an increase in state economic activity. In addition to purchasing additional goods and services, industries also hire additional labor and pay additional wages at each round. These additional wages produce an increase in local economic activity via an increase in consumer expenditures. This cycle continues until, in the final round, all money remaining leaks out of a region.

The cumulative effect of these purchases is referred to as the multiplier effect and the additional purchases and economic activity created as a result of the initial increased demand is referred to as the economic impact. The reverse of all these effects occurs if there is a decrease in the demand for the output of an industry.

Three different types of impacts are measured with an I-O model. Direct impacts are the changes that occur in the immediately impacted industry or industries. Indirect impacts are economic changes that occur because of all the inter-industry purchases that are made at each round. Induced impacts occur when impacted workers spend their money locally.

In the present analysis, impacts for the following economic measures will be reported: jobs, labor income, and output. Labor income includes both employee compensation and an estimate of proprietor’s income. Output represents gross sales, or the accumulation of sales at each round of purchases.

Economic Impact of Raising Revenue with the Sales Tax Increase

When assessing the negative economic impact associated with raising revenue from an increase in the sales tax, it is important to note that not all of the sales tax revenue is paid by local households and businesses; a significant portion is paid by tourists. It is estimated that approximately 10.4% of all taxable sales taxes paid in Arizona are paid by tourists.

The amount of state sales tax revenues paid by tourists is estimated using the 2008 visitor expenditures, by type of commodity purchased, produced by Dean Runyan and Associates, prepared for the Arizona Office of Tourism. For each category of expenditures (lodging, food and beverage service, food stores, etc.), the taxable portion is estimated. For example, in the food stores category, it is assumed that only 10% is taxable because most of those purchases would be for non-taxable food. However, 10% is assumed taxable to allow for other purchases, which are taxable, such as personal care products, laundry products, etc., which are taxable.
Tourism does not contribute uniformly to all categories of taxable sales. Tourism accounts for almost all of hotel/motel collections, a substantial portion of restaurant and bar tax collections and lesser amounts of personal property rentals (car rentals), retail sales and amusements. Very little of tourism expenditures occurs for other sales tax categories, such as contracting, utilities and communications. Tourism taxable sales, by category were subtracted prior to assessing the impact of the sales tax.

A one cent sales tax increase is comparable to increasing the price of taxable goods and services by one percent. In Arizona, the sales tax applies mostly to the sales of goods, although some services, such as telephone services and car rental services, are taxable.

In order to assess the impact of the non-tourism portion of the sales tax increase, some assumptions must be made about how consumers and businesses will respond to the tax. First, it is assumed that any sales tax paid by businesses will be passed either forward onto consumers or backwards onto labor in the form of lower wages. Second, households view the sales tax increase as an increase in the price of taxable goods and services.

Local consumers will be purchasing less as a result of the tax but it is impossible to know exactly how consumers will respond. As a result, two different impact scenarios are created and averaged. The first scenario has consumers reducing their consumption of taxable goods and services by the full amount of the tax for each taxable sales category. In this scenario, the impact of reducing taxable sales, by category, by the amount of the tax, is estimated. This implicitly assumes that the price elasticity of all taxable purchases is equal to 1 so that consumers continue to spend the same amount of money on taxable items, but receive 1 percent fewer goods and services.

The second scenario assumes that households view the one percent sales tax as a reduction in income. That is, their budgets require cutting back on all expenditure components in proportion to their personal consumption expenditures mix. Both of these scenarios are assessed excluding the portion of the tax paid by tourists. Both of these scenarios overstate the negative impact of the tax if the tax is paid for out of savings. In the present analysis, it is assumed that there is no change in savings.

The results of the impact calculations of the two scenarios were within five percent of each other. Because it is impossible to know the precise consumer response to the sales tax increase, the two impacts are averaged and presented in Table 1 below. The direct impact on retailers is a reduction of 4,283 jobs and $180.5 million in labor income. Including the multiplier effects, the one percent sales tax is estimated to reduce jobs (man-years) by approximately 7,383 jobs.

Table 1: Total Economic Impact of Increasing the Sales Tax by One Cent ($918M)

	Direct	Indirect	Induced	Total
Employment	(4,283)	(1,165)	(1,935)	(7,383)
Labor Income	(180,522,912)	(59,885,211)	(85,775,696)	(326,183,818)
Output	(547,280,652)	(170,784,749)	(207,445,581)	(925,510,986)

Economic Impact of Cutting Government Expenditures

The budget cuts that will be made next fiscal year if the one percent sales tax increase is not approved are detailed in the FY2011 Conditional Enactments. The cuts total $867.5 million and are listed by government agency. For those agencies that are conditionally going to receive the largest cuts (over $10 million), detailed budgets were pulled from the 2010 State of Arizona Appropriations Report. These detailed budgets were put into Excel and each spending category for each agency was matched with North American Industrial Classification System (NAICS) codes and inputted into IMPLAN. For the agencies receiving smaller expenditure cuts, the cuts were summed (totaling $26 million) and a portion was entered into IMPLAN as a reduction in non-educational government payroll.

The direct economic impact of cutting government expenditures by $867.5M is to reduce jobs 10,429 and labor income by $489.5 million. When indirect and induced impacts are computed, the expenditure cuts will reduce jobs by 14,092 and payroll by $652.9million.

Table 2: Economic Impact of Cutting Government Expenditures by $867.6M

	Direct	Indirect	Induced	Total
Employment	(10,429)	(554)	(3,133)	(14,092)
Labor Income	(489,519,026)	(27,766,178)	(135,652,371)	(652,937,575)
Output	(676,767,164)	(79,505,150)	(328,337,852)	(1,084,610,165)

The figures in Table 2 severely underestimate the job loss of the expenditure cuts because, in addition to the $867.6 million cut in state government expenditures, Arizona will lose an estimated $442.5 million in federal government matching funds due to cuts in health and welfare agencies. When the federal matching funds are included, the direct job loss and loss to labor income are estimated to be 14,629 jobs and $700.3 million, respectively. Total economic losses associated with the combined state and federal cut in expenditures is estimated to be 20,510 jobs, $963.7 million in labor income, and $1.7 billion in output (gross sales)1, Table 3.

Table 3: Economic Impact of Cutting Government Expenditures by $1.3B
(Includes $442.5M in Federal Government Matching Funds)

	Direct	Indirect	Induced	Total
Employment	(14,629)	(1,257)	(4,647)	(20,510)
Labor Income	(700,265,061)	(62,195,267)	(201,273,554)	(963,733,881)
Output	(1,045,469,702)	(177,463,146)	(504,161,404)	(1,727,094,252)

Summary and Conclusions

Increasing the sales tax by one percent is estimated to reduce employment by a total of 7,383 jobs. Cutting Arizona state government expenditures is expected to reduce jobs by 14,092 (without considering federal matching dollars). Adding the expenditure cuts associated with federal matching dollars will reduce employment by an estimated 20,510.

Obviously, the loss in federal matching dollars substantially contributes to the finding that job losses are larger with expenditure cuts than with a sales tax increase. But even without the federal matching money, expenditure cuts reduce jobs substantially more than comparably sized tax increases. There are three major reasons why a cut in government expenditures result in a larger economic impact than an increase in the sales tax rate.

First, the government is a service provider and, generally, services employ more persons per $1 million of expenditures than do non-services providers. Consequently, a larger portion of government expenditures is paid as labor income than most sectors of the economy. Sometimes the state government provides services directly, such as parts of higher education, sometimes the state government subsidizes other government entities to provide the services, such as K-12 education, and often the state government contracts with private entities to provide services, such as health care, day care, and social service providers.

Second, from the taxation side, a portion of the sales tax is paid by out-of-state visitors, so only slightly under 90 percent of the tax is paid by Arizona residents.

Third, the sales tax is imposed primarily on the sale of tangible goods, rather than services. Very few of the good purchased by Arizona consumers are produced in Arizona. For most sales, only the retail margin (the difference between final sale price and the wholesale cost of the item) is retained in the state. Retail margins can be as low as 27 percent of the total sales price for purchases made at general merchandise stores.

It is also important to note that both the estimated job loss associated with the one percent increase in the sales tax (7,383) and the job gains associated with not cutting government spending (20,510, with the federal match) are both extremely small when compared to Arizona’s overall economy. Arizona currently has approximately 2.4 million jobs, which is down approximately 290 thousand jobs from peak employment levels in 2007.

For a detailed comparison of the simulation model employed in this article, please read:

Comparison of UA, REMI, and STAMP Simulations of Tax/Spending Increases. Alberta H. Charney, Ph.D., March, 30, 2010.

What Will It Cost If Arizona Voters Reject the One-Cent Sales Tax Hike On May 18, 2010? Alberta H. Charney, Ph.D., March, 30, 2010.

Find more research on Arizona Fiscal Issues .

Note:

1 Gross sales impact estimates represent the accumulation of sales as impacts cycle through the economy. It should not be compared with Gross State Product, which is a net production measure.

For additional information, please contact us.

| More