The Fiscal Sustainability of State and Local Governments · 2016-01-19 · Interpreting Fiscal Sustainability Summary: long-term ability of state and local governments to Provide

The Fiscal Sustainability of State and The Fiscal Sustainability of State and Local Governments

Bo ZhaoNew England Public Policy CenterFederal Reserve Bank of BostonFederal Reserve Bank of Boston

FTA Revenue Estimation & Tax Research ConferenceOctober 24, 2012October 24, 2012

Disclaimer: The views expressed here are those of the authors and do not necessarily represent the views of the Federal Reserve Bank of Boston or the Federal Reserve System.

Motivation

State and local governments face long-term fiscal S a e a d oca gove e s ace o g e sca challenges: Disproportionate growth in health care costs Large unfunded pension and OPEB liabilities Impending substantial cuts in federal aid

Failure to achieve fiscal sustainability could cause: Intergenerational inequality Disruption of future public services Lower credit ratings and higher borrowing costs Instability of the broad financial system Instability of the broad financial system

Research Goals

Clarify and interpret fiscal sustainability of state and local C a y a d e p e sca sus a ab y o s a e a d oca governments

Use new data and methodology to estimate “trend gaps” Use new data and methodology to estimate trend gaps in the recent decade

Forecast trend gaps for future years Forecast trend gaps for future years

Defining State & Local Fiscal Sustainability

Chapman (2008): long-run capability to ensure the C ap a ( 008): o g u capab y o e su e e continued provision of service and capital levels that the public demand

GASB (2011): a government’s ability and willingness to generate revenues needed to meet both current service commitments and financial obligations when they come due

Ward and Dadayan (2009): a government’s ability to Wa d a d adaya ( 009): a gove e s ab y o balance revenues and expenditures in the long term

Interpreting Fiscal Sustainability

Summary: long-term ability of state and local governments to Summary: long-term ability of state and local governments to Provide public services the public demand and are willing to pay for Balance revenues and expenditures

Our interpretation: Such ability should be determined by underlying economic, social, and

demographic characteristics. Because it is a long-term concept, it should focus on the trend revenue

and expenditure, not influenced by cyclical movements or other short-and expenditure, not influenced by cyclical movements or other shortterm factors.

Existing Empirical Studies

GAO (2008, 2011, 2012) studies the whole state & local G O ( 008, 0 , 0 ) s ud es e w o e s a e & oca government sector, using aggregate data

Ulb i h (1997) t di S th C li ’ t t l f d Ulbrich (1997) studies South Carolina’s state general funds

Dye and Hudspeth (2010) study Illinois’ state “consolidated funds”

Common Measurement Problems

Directly use actual revenues and expenditures to measure ec y use ac ua eve ues a d e pe d u es o easu e fiscal balances/gaps for the past years Do not separate the trend from cyclical movements Their balance/gap measures indeed fluctuate with business cycles

A l l h l d Apply long-term growth rates to actual revenues and expenditures of a base year to make projections Implicitly assume the cyclical and other short-term influences in the base Implicitly assume the cyclical and other short term influences in the base

year are permanent Could overestimate future gaps if the base year is in recession

Data

Use state and local level data from the 1990—2009 Use s a e a d oca eve da a o e 990 009 Annual Survey of State and Local Government Finance

Combine state and local finances

Examine all revenue and expenditure categories

Pension and OPEB Data

Data source: Pew Center on the Statesa a sou ce: ew Ce e o e S a es

Use Actuarially Required Contributions (ARCs) to measure long term retirement costslong-term retirement costs Include payments for amortizing unfunded liability More comprehensive than actual government contributionsp g

ARCs underrepresent true retirement costsG t t d t h hi h di t t t tifi i ll l Governments tend to choose high discount rates to artificially lower ARCs

The Pew Center’s data underreport at local level

Example of Revenue Regressions

log(tax revenue)

log(personal income) 1.063***

Log(personal income)*(multiple state income tax rate brackets) 0.004***

State unemployment rate -0.007*

Percent of population with less than a high school degree -0.003

Percent of population with at least a college degree -0.003

State Fixed Effects Yes

Year Fixed Effects Yes

Number of observations 918

R-Squared 0.947

Note: *** p < 0.01, ** p < 0.05, * p < 0.10. Standard errors are clustered by state.

Example of Expenditure Regressions

log(educationspending)

log(social services and income maintenance spending)spending) maintenance spending)

Log(personal income) 0.556*** 0.408

Percent of population with less than a high school degree -0.003 -0.002

Percent of population with at least a college degree 0 002 0 003Percent of population with at least a college degree 0.002 -0.003

State unemployment rate -0.006 0.014**

Percent of population aged 65 and older 0.004 0.027

Percent of population aged less than 18 0.033*** -0.002*Percent of population aged less than 18 0.033 0.002

log(population density) -0.208**

Education CPI 0.006***

Medical care CPI 0.018***

State Fixed Effects Yes Yes

Year Fixed Effects Yes Yes

Number of observations 765 918

R-Squared 0.943 0.935

Note: *** p < 0.01, ** p < 0.05, * p < 0.10. Standard errors are clustered by state.

Estimating Trend Gap

Use regression coefficients and actual values of explanatory Use eg ess o coe c e s a d ac ua va ues o e p a a o y variables to estimate trend revenue and expenditure

Remove the effect of business cycles and other short-term yinfluences: Replace actual unemployment rate with the average unemployment

f h 1990 2009rate for each state across 1990-2009 Replace actual personal income with estimated income under the

long-run state average unemployment rate and potential GDP Exclude year fixed effects in estimating trends

Trend gap = trend expenditure – trend revenueg p p

Figure 1. The Combined State and Local Fiscal Gap: A t l T d

$600

$800

Gap

($)

Actual vs. Trend

$200

$400

apita

Fis

cal

-$200

$0

$200

Rea

l Per

Ca

-$200R

2000 2002 2004 2006 2008 2010

Year

Actual Gap without OPEB Trend Gap without OPEB

Actual Gap with OPEB Trend Gap with OPEB

Recession

Note : Weighted by population. All gap measures include pension.

Future Work

Identify and quantify driving forces for the increasing trend gaps Preliminary investigation shows rapid growth of SSIM (mostly Preliminary investigation shows rapid growth of SSIM (mostly

Medicaid), pension, and OPEB costs.

Forecast future trend gaps Forecast future trend gaps

Conclusion

State and local trend gaps have been steadily increasing S a e a d oca e d gaps ave bee s ead y c eas g in the recent decade.

This increasing pattern is unlikely to change substantially This increasing pattern is unlikely to change substantially in a short time period.

GASB (2011) recommends conducting long term financial GASB (2011) recommends conducting long-term financial planning to improve fiscal sustainability.

O l i t th t it i i t t t t Our analysis suggests that it is important to separate trends from cyclical, short-term responses in long-term planning.p g

Additi l M t i lAdditional Materials

Revenue Regressions

log(tax revenue) log(other own revenue) log(federal transfers)

log(personal income) 1.063*** 0.601**

Log(personal income)*(multiple state income tax rate brackets)

0.004***

S l 0 007*State unemployment rate -0.007*

1 year lag on state unemployment rate 0.018*** 0.036***

Percent of population with less than a high school degree

-0.003 -0.009***high school degree

Percent of population with at least a college degree

-0.003 0.002

log(real GDP) 1.789***

State Fixed Effects Yes Yes Yes

Year Fixed Effects Yes Yes No

Number of observations 918 918 918Number of observations 918 918 918

R-Squared 0.947 0.959 0.923

Note: *** p < 0.01, ** p < 0.05, * p < 0.10. Standard errors are clustered by state.

Other Expenditure Regressions

l ( bli l ( i l ( l ( h

p g

log(transportation)log(public

safety)log(environment

and housing)log (government administration)

log(other expenditures)

log(personal income) 0.854*** 0.442 1.061*** 0.678*** 0.844***

Percent of population with less than a high school degree

-0.009* -0.008* -0.008* -0.013*** 0.002

Percent of population with at least a ll d

-0.006 -0.009** -0.000 -0.007** -0.003college degree

State Unemployment rate 0.000 -0.002 0.004 -0.003 0.022**

Percent of population aged 65 + -0.016

Percent of population aged less than 18 -0.045***

State Fixed Effects Yes Yes Yes Yes Yes

Year Fixed Effects Yes Yes Yes Yes Yes

Number of observations 918 918 918 918 918Number of observations 918 918 918 918 918

R-squared 0.855 0.961 0.912 0.938 0.949

Note:***p 0.01,**p 0.05,*p 0.1;Standarderrorsareclusteredbystate.

$42 000$)

Appendix Figure 1. Personal Income:Actual vs. Trend

$38,000

$40,000

$42,000na

l Inc

ome

(

$34,000

$36,000

$38,000

apita

Per

son

$30,000

$32,000

Rea

l Per

Ca

R

1990 1995 2000 2005 2010

Year

Actual Income Trend IncomeRecession

Note : Weighted by population

SAGEAGECOLLEGELTHSGDPUPI

)18(021.0)65(013.0 )(003.0)(001.0)log(606.0)(007.0)log(

Appendix Figure 2. The Combined State and Local Revenue:

$8,500

$9,000

e ($

)Appendix Figure 2. The Combined State and Local Revenue:

Actual vs. Trend

$7,500

$8,000

$ ,

ita R

even

ue

$6,500

$7,000

,

eal P

er C

ap

$6,000

Re

1990 1995 2000 2005 2010

Year

Actual Revenue Trend RevenueRecession

Note : Weighted by population

Appendix Figure 3. The Combined State and Local Expenditures:

$9 000

$9,500di

ture

($)

pp g pActual vs. Trend

$8,500

$9,000

pita

Exp

end

$7,500

$8,000

Rea

l Per

Ca

$7,500R

2000 2002 2004 2006 2008 2010

Year

Actual Expenditure without OPEB Trend Expenditure without OPEB

Actual Expenditure with OPEB Trend Expenditure with OPEB

Recession

Note : Weighted by population All expenditure measures include pensionNote : Weighted by population. All expenditure measures include pension.

)

$9 000

$9,500pe

nditu

re ($

) Figure 4.Trend Revenue vs. Trend Expenditure

$8,500

$9,000

venu

e or

Exp

$8,000

r Cap

ita R

ev

$7,500

Rea

l Pe

2000 2002 2004 2006 2008 2010

Yearea

Trend Revenue Trend Expenditure

Recession

Note : Weighted by population. The trend expenditure measure includes pension but not OPEB.g y p p p p