Diamond Testifies Before the Texas House Ways and Means Committee
Table of Contents
Author(s)
This testimony was delivered before the Texas House of Representatives Committee on Ways & Means on September 16, 2014.
Video of the testimony is available here. John Diamond's testimony starts at the 46:30 mark.
Introduction
Chairman Hilderbran, Vice Chairman Otto, and distinguished Members of the Committee, it is a pleasure to testify before you on the potential usefulness of including estimates of the macroeconomic effects of tax and expenditure policies in the budget process. I will focus my discussion on:
-
What dynamic analysis is, why it is important, as well as why it is difficult to accomplish and implement;
-
A comparison of federal and state budget scoring methods;
-
A brief overview of state and local tax and spending issues; and
- Some policy guidelines for implementing dynamic analysis.
Dynamic Analysis: What Is It and Why Is It Important
Dynamic scoring is theoretically preferred to the current budget scoring process; however, many questions remain about how best to implement a consistent and practical framework that allows information on the macroeconomic effects of various policy changes to be included in the budget process.
A popular management adage is, “If you can’t measure it, you can’t manage it.” It is important to adopt fiscal policies that maximize economic growth, subject to political constraints and distributional goals. However, to do this we must be able to measure the relative effects of alternative policies on the economy. Dynamic analysis provides valuable information about the effects of policy proposals on economic growth, and it is important that we use this information to better manage U.S. and Texas fiscal policy. Note that routinely disregarding information on the macroeconomic effects of alternative proposals may lead to a budget process that undervalues proposals that help grow the Texas economy and overvalues proposals that shrink the Texas economy.
Dynamic analysis allows the budget process to account for the effect of policy proposals on the level of aggregate output (gross domestic product), which is a function of the size of the capital stock and total hours of work in the economy. For example, if increasing the marginal tax rate on wage income decreases the number of hours individuals are willing to supply in the labor market, then such a policy would lead to a decrease in aggregate output. However, current methods of estimating revenue assume that aggregate output would not be affected by the change in labor supply, and thus the revenue estimate would over estimate the amount of revenue that would be raised from the tax increase. Under current revenue estimating procedures, analysts account for the direct effect of proposals on program costs and tax receipts and may include some analysis of the proposals’ ancillary effects on labor supply, saving and investment, tax avoidance, and the response of other entities. Dynamic analysis would account for how the direct and ancillary effects would impact aggregate output and the impact of the change in aggregate output on program costs and tax receipts. Thus, the resulting dynamic revenue estimate would include the direct program costs, the costs related to ancillary incentive effects, and the effects related to changes in aggregate output.
In addition, dynamic analysis may be used to examine the effects of policies on wages, consumption, welfare (under certain types of modeling), distributional outcomes (both within and across generations), as well as other important variables. Dynamic analysis may increase the likelihood of adopting tax policies that maximize economic growth. In addition, dynamic analysis can highlight areas of inherent uncertainty of various policy initiatives.
While dynamic analysis will provide valuable information about the relative economic effects of alternative policies, it is not a panacea for the budget issues facing the state or the nation. Policymakers will still face tough decisions regarding the use of scarce resources. In addition, it is important to note that preparing a dynamic analysis is no easy task and presenting and communicating the results to members, their staff, and the general public is difficult.
Comparing Federal and State Budget Scoring Methods
In Texas, the Legislative Budget Board (LBB) is responsible for providing information about the fiscal effects of various policy proposals. According the LBB,
“State statute requires that the LBB prepare a fiscal note to accompany a bill or joint resolution as it goes through the legislative process. A fiscal note is a written estimate of the costs, savings, revenue gain, or revenue loss that may result from implementation of requirements in a bill or joint resolution. It serves as a tool to help legislators better understand how a bill might impact the state budget as a whole, individual agencies, and in some instances, local governments.”
A fiscal note is similar in nature to a revenue or cost estimate at the federal level — revenue estimates are provided by the Joint Committee on Taxation (JCT) and cost estimates are provided by the Congressional Budget Office (CBO) as specified in the Congressional Budget Act of 1974 (CBA 1974). Thus, it is useful to begin by briefly reviewing the budget process at the federal level.
CBA 1974 required CBO to produce baseline budget projections of outlays, revenues, budget authority, and surplus and deficits. Baseline budgeting uses projections of current levels of spending and revenues as the benchmark to compare the effects of policy changes on the budget. Note that the current budget baseline is a 10-year projection that accounts for inflation and population growth. At the federal level, official revenue and cost estimates assume that policy changes do not alter gross domestic product (GDP); however, various microeconomic behavioral responses are included in official revenue estimates. For example, if the marginal tax rate on labor is increased the official revenue estimate would account for some behavioral responses, such as employers shifting compensation from taxable wages to fringe benefits, which are deductible (e.g., health insurance). The estimate would not include a reduction in hours worked that would decrease GDP in the United States. Thus, at the federal level the effects of behavioral changes on total production are ignored for official revenue estimates. It is important to note that current congressional rules require that a dynamic analysis be provided for certain policy proposals or that the JCT provide a reason for not supplying such an analysis. These analyses are purely supplemental and are intended for informational purposes only (i.e., they are not included in official budget numbers). House Rule XIII.3.(h)(2) of the Rules of the House of Representatives, adopted January 3, 2013, in the 113th Congress, includes the following requirements:
(2)(A) It shall not be in order to consider a bill or joint resolution reported by the Committee on Ways and Means that proposes to amend the Internal Revenue Code of 1986 unless —
(i) the report includes a macroeconomic impact analysis;
(ii) the report includes a statement from the Joint Committee on Internal Revenue Taxation explaining why a macroeconomic impact analysis is not calculable; or
(iii) the chair of the Committee on Ways and Means causes a macroeconomic impact analysis to be printed in the Congressional Record before consideration of the bill or joint resolution.
(B) In subdivision (A), the term “macroeconomic impact analysis” means—
(i) an estimate prepared by the Joint Committee on Internal Revenue Taxation of the changes in economic output, employment, capital stock, and tax revenues expected to result from enactment of the proposal; and
(ii) a statement from the Joint Committee on Internal Revenue Taxation identifying the critical assumptions and the source of data underlying that estimate.
Note that dynamic analysis is already used on a fairly wide scale at the federal level. For example, the Joint Committee on Taxation (JCT) has produced dynamic analyses of several significant tax proposals (JCT 2003; JCT 2005; JCT 2006; JCT 2014a; JCT 2014b). In addition, the Department of the Treasury’s Office of Tax Analysis (OTA) has published dynamic analyses of the reform proposals made by the President’s Advisory Panel on Federal Tax Reform (Carroll, Diamond, Johnson, and Mackie 2006) and the proposal to permanently extend the President’s tax relief (OTA 2006). The Congressional Budget Office also publishes macroeconomic analyses of various proposals on an annual basis, including the President’s Budget (CBO 2003a and 2003b). A number of private sector institutions also provide dynamic analyses of federal (and state) policy proposals.
In Texas, the comptroller is responsible for developing the Biennial Revenue Estimate (BRE) — which requires an economic projection for the next three years — and the LBB and governor both make budget recommendations for various governmental agencies. Once the legislature is in session, all bills are examined and the LBB provides fiscal notes — an estimate of the revenue or cost implications — with input from the comptroller. Fiscal notes are updated during the legislative process as bills are modified. However, not much information is provided in terms of how fiscal notes are calculated. For example, H.B. 3 in 2006, the major reform of the Texas Franchise tax, simply includes the following statement regarding methodology:
“The estimate was based on data from the Internal Revenue Service’s Statistics of Income publications, the U.S. Census Bureau, and information in the Comptroller’s franchise tax data files.”
This is obviously not enough information to determine what types of behavioral responses, if any, are included in the estimate. This raises the question of how the budget process could be reformed to provide lawmakers more information. And, more specifically, it raises the question of whether dynamic analysis of some policy proposals would add valuable information to the legislative process. However, H.B. 464, which passed in 2009, already requires the staff of the LBB to prepare a dynamic fiscal impact statement if a bill or resolution raises or lowers a tax or fee. After five years, the comptroller is required to report on the accuracy of the fiscal note and dynamic impact statement.
The LBB published a dynamic economic impact statement for H.B. 1 on March 30, 2011. H.B. 1 reduced expenditures by 12.3 percent relative to the 2010-11 appropriations. LBB used the REMI Policy Insight Model to examine the macroeconomic effects of H.B. 1. LBB reported that H.B. 1 would reduce the number of jobs by 272,000 relative to the baseline number of jobs that would have occurred had revenues not fallen in the recession and the slow recovery that followed. LBB also issued a dynamic impact statement for H.B. 1025 — which proposed to increase appropriations by $92.4 million in the 2012-13 biennium — that simply stated there would be no significant impact on economic activity in the state of Texas. Note that dynamic impact statement for H.B. 1 was effectively a forecast of the economic effects of the policy. However, the benefit of dynamic analysis is that it can be used to compare the effects of competing policies. For example, for H.B. 1, the competing policy could have been the effect of raising taxes to close the budget gap. Increased taxes would have also decreased economic output and job growth relative to the baseline. Dynamic analysis could be used to examine the types of spending cuts or tax increases that were likely to have the smallest negative economic effects.
A Brief Overview of State Tax and Spending Issues
The major benefit of dynamic analysis is that it would provide policymakers additional information in choosing between alternative policies. This may include information about the optimal mix of taxes, the effects of increasing or decreasing taxes or tax rates, the optimal mix of spending (i.e., transfers, education, infrastructure, health, etc.), and the effect of changes in the level of spending. Note that such analysis is unlikely to provide definitive information given the lack of consensus on a number of different topics and the lack of detailed modeling of some of these issues in the economics literature. However, even on issues which the literature does not reach a consensus, the information provided about the range of potential effects of alternative policy choices may be useful. Note that there is a large literature that examines the effects of state spending and tax policies on aggregate output, personal income growth, employment, the location decisions of individuals and firms, and other decisions.
A number of studies assert a strong negative relationship between taxes and economic growth (Mullen and Williams 1994; Holcombe and Lacombe 2004; Reed 2008; Goff, Lebedinsky, and Lile 2011; Agostini 2007). For example, Goff, Lebedinsky, and Lile (2011, p. 303) state that “our results provide strong support for the idea that lower tax burdens tend to lead to higher levels of economic growth” and they find that individual income taxes are most important as a determinant of economic growth. In addition, Harden and Hoyt (2003, p. 7) find that “states are not choosing the mix of taxes to minimize losses in employment growth with corporate income tax rates set relatively too high.”
However, other studies find results that are partially at odds with these and other studies. For example, Ojede and Yamarik (2012, p. 161) find that “property and sales tax rates have negative effects on long-run income growth, while income tax rates have no impact.” However, it is troubling from a theoretical perspective to explain how a consumption tax (e.g., the sales tax) would affect long-run growth while an income tax would not. One possible, although less than convincing, explanation is that the inclusion of business purchases under the sales tax implies that it imposes a significant burden on businesses and thus reduces economic growth. Deskins and Hill (2008) find that increasing the tax burden as a share of personal income had a significant negative effect in 1985 but that the negative effect was nonexistent by 2003. Alm and Rogers (2011, p. 516) find that the estimated coefficient of taxes is sensitive to the specification and time period chosen and state that “further work must be done before it can be determined what these results mean.” A broad reading of the literature implies that it is likely that higher taxes lead to reduced economic growth in the long run, although this will depend in part on how tax revenues are being raised (e.g., an increase in rates versus an increase in revenue from base broadening) and the uses of the revenue. In addition, the theoretical literature argues that distortionary taxes and growth are negatively related and Yamarik (2000) provides results supporting the predictions of growth theory.
Another strand of the literature examines what happens to economic growth if you raise (lower) taxes to increase (decrease) spending. Helms (1985) found that increasing state and local taxes led to significant reductions in economic growth if the revenues were used to fund transfer payments, but that if the revenues were used to fund education or highway spending that it was possible to offset the negative effects of the tax increases. Tomljanovich (2004) find results consistent with Helms (1985), except income tax rate increases only affected short-run growth and not long-run growth. However, Taylor and Brown (2006, p. 561) suggest that “increases in all but a select few categories of state and local government spending would reduce private sector growth for the average state.” And in addition Taylor and Brown (2006, p. 561) state that “as a general rule, the costs of the negative effects of the additional taxes more than offset the benefits accruing from additional government services.”
Still other studies examine the relationship between taxes and firm and individual location decisions, entrepreneurship, and the location of foreign direct investment. These issues are certain to be important determinants of the optimal mix of state tax policy and state expenditures.
Policy Guidelines for Implementing Dynamic Analysis
-
While examining the aggregate macroeconomic effects of various proposals is of interest, this approach ignores much of the additional information that could be gleaned from dynamic analyses. Thus, dynamic analysis should focus on comparing the macroeconomic effects of competing provisions as well as presenting information on the aggregate effects of all the provisions. Obviously, analyzing every provision separately would be impossible and counterproductive, as this would consume far too many staff resources. However, it is important to ensure that the choice of provisions to be analyzed is not politically driven, as this would undermine the integrity of the process. A balance must be struck on this issue.
-
Dynamic analysis may also be applied to some spending proposals. However, the demand-side effects of spending and tax proposals should not be considered, especially for permanent proposals.
-
Macroeconomic aggregates are not the only information that should be provided to policymakers. Some measure of economic well-being should also be provided in addition to the macroeconomic aggregates. This is important because positive macroeconomic effects can be associated with negative welfare effects.
-
Distributional analyses should also be conducted both within income groups and across generations for certain policies. For example, the President’s Advisory Panel on Federal Tax Reform in the United States decided against recommending a true consumption-based tax and, instead, proposed a consumption-based system supplemented with an “add-on” capital income tax at the individual level (the “Growth and Investment Tax” or GIT). Given that the report showed that the economic gains were larger under the consumption-based tax relative to the GIT, it would be interesting to compare how the plans differed from a distributional perspective, both during the transition and in the long run.
-
The extent of the uncertainty contained in a dynamic analysis should be well noted. For example, this would include discussing the sensitivity of the results to various assumptions about parameter values, the assumptions underlying the economic model, whether the policy was financed by changes in government spending (and the effects of such spending on welfare) or taxes, and assumptions about the reactions of other state governments.
-
Dynamic analysis should be timely so that it can be used effectively in the formulation of policy. At the federal level, the current House rule XIII.3.(h)(2) requires an analysis of the macroeconomic effects before the bill can be considered on the floor. This is somewhat late in the political process, as many of the major details of a bill are typically established at this point. It is important to note that there are possible logistical constraints on this issue, given the current state of macroeconomic modeling.
-
Public disclosure is imperative. As much information as possible should be released to the public. At a minimum, enough information should be released so that outside entities could replicate the work. This will ensure that the process is seen as fair and open and will serve as a check on those who provide the estimates.
References
Alm, James, and Janet Rogers, 2011. “Do State Fiscal Policies Affect Economic Growth?” Public Finance Review 39 (4), 483–526.
Carroll, Robert, John Diamond, Craig Johnson, and James Mackie III, 2006. A Summary of the Dynamic Analysis of the Tax Reform Options Prepared for the President’s Advisory Panel on Federal Tax Reform, U.S. Department of the Treasury, Office of Tax Analysis, May 25, 2006, prepared for the American Enterprise Institute Conference on Tax Reform and Dynamic Analysis May 25, 2006
Congressional Budget Office, 2003. An Analysis of the President's Budgetary Proposals for Fiscal Year 2004 (March 2003).
Congressional Budget Office, 2003. How CBO Analyzed the Macroeconomic Effects of the President's Budget (July 2003).
Deskins, John and Brian Hill, 2008. “State Taxes and Economic Growth Revisited: Have Distortions Changed?” Annals of Regional Science 44 (2) , 331–348.
Goff, Brian, Alex Lebedinsky, and Stephen Lile, 2011. “A Matched Pairs Analysis of State Growth Differences.” Contemporary Economic Policy 30 (2), 293–305.
Harden, William J., and William H. Hoyt, 2003. “Do States Choose Their Mix of Taxes to Minimize Employment Losses?” National Tax Journal 55 (1), 7–26.
Helms, Jay, 1985. “The Effect of State and Local Taxes on Economic Growth: A Time Series-Cross Section Approach.” Review of Economics and Statistics 67 (4), 574–82.
Holcombe, Randall G., and Donald J. Lacombe, 2004. “The Effect of State Income Taxation on Per Capita Income Growth.” Public Finance Review 32 (3), 292–312.
Joint Committee on Taxation, 2003. Overview of Work of the Staff of the Joint Committee on Taxation to Model the Macroeconomic Effects of Proposed Tax Legislation to Comply with House Rule XIII.3.(h)(2) (JCX-105-03), December 22, 2003.
Joint Committee on Taxation, 2005. Macroeconomic Analysis of Various Proposals to Provide $500 Billion in Tax Relief, (JCX-4-05), March 1, 2005.
Mullen, John, and Martin Williams, 1994. “Marginal Tax Rates and State Economic Growth.” Regional Science and Urban Economics 24 (6), 687–705.
Ojede, Andrew, and Steven Yamarik, 2012. “Tax Policy and State Economic Growth: The Long and Short of It.” Economics Letters 116 (2), 161–65.
Reed, Robert W., 2008. “The Robust Relationship between Taxes and U.S. State Income Growth.” National Tax Journal 61 (1), 57–80.
Taylor Lori L., and Stephen A. Brown, 2006. “The Private Sector Impact of State and Local Government: Has More Become Bad?” Contemporary Economic Policy 24 (4), 548–62.
Tomljanovich, Marc, 2004. “The Role of State Fiscal Policy in State Economic Growth.” Contemporary Economic Policy 22 (3), 318–30.
U.S. Department of the Treasury, Office of Tax Analysis, 2006. A Dynamic Analysis of Permanent Extension of the President’s Tax Relief, July 25, 2006.
Yamarik, Steven, 2000. “Can Tax Policy Help Explain State-Level Macroeconomic Growth?” Economic Letters 68 (2), 211–15.