Nick Warner Forecasting School Districts’ Revenues and ...cslf.gsu.edu/files/2015/12/GASBO-School-District-Forecasting... · Forecasting School Districts’ Revenues and Expenditures

Forecasting School Districts’ Revenues

and Expenditures

Nick Warner

November 5, 2015

Forecasting Overview

Forecasting Districts’ Revenues and Expenditures2

What is forecasting?

• Predicting the future as accurately as possible, given all of the information available, including historical data and knowledge of any future events that might impact the forecasts

• Many methods ‐ Forecasters use the one that best fits the known data.


Why is forecasting challenging?


Why is forecasting challenging?

• A forecast must be based on what we know at the time, and things could change.

• Some data is predictably variable or volatile, so any forecast includes a range of possible amounts.


Why would a school district use forecasting?


Why would a school district use forecasting?• Inform long‐ and short‐term budget, capital, and staffing plans– Are expected revenues sufficient to cover expected expenditures under current conditions?

– Are expected revenues sufficient to implement new improvement initiatives?

– Is the district positioned to handle changes in student population?

– And more...


How can we forecast data important to school districts?


How would a school district forecast?

• Do it yourself, based on known data sources and info about your district– Methods range from very basic to extremely advanced, but all will provide very useful insights and help inform decisions and planning.

• Get help from experts

Warning: A forecast should be used to inform decisions, not dictate budget planning or decisions. Budget planning and forecasting should remain separate.


Property Tax Revenue

• Importance: Primary local funding source for operations (with the exception of certain systems that have access to sales taxes for operations)

• Data Sources: Digest values, observed home sales, info from property tax assessor’s office

• Volatility: Low, relatively easy to forecast– Exception: The Great Recession created declining statewide property tax digests.

• Method to Try: Linear trend forecast (ordinary least squares)




$500

$700

$900

$1,100

$1,300

$1,500

$1,700

$1,900

$2,100

2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014

Net

M&

O D

iges

t in

$ M

Illio

ns

Tax Year

District A Linear trend line



$500

$700

$900

$1,100

$1,300

$1,500

$1,700

$1,900

$2,100

2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015

Net

M&

O D

iges

t in

$ M

Illio

ns

Tax Year

District A With Moving Average Forecastg g

Two period moving average

Which Line fits the data better? otherwise known as “goodness of fit”

Sales Tax (ESPLOST)

• Importance: A major fund source for capital projects– Districts enter into five year plans based in part on expected collections

• Data Sources: Sales tax collections (GDOR) and information on community business activity

• Volatility: High, forecast results in a wide range– Seasonality and other trends generally are incorporated into forecast.

• Method to Try: An observed average over a long period of time.


Sales Tax (ESPLOST)


40

50

60

70

80

90

100

110

ESP

LOST

Rev

enue

$ in

Tho

usan

ds

Month -Year

Sample District, 1999-2015

g1999-2015 monthly

average

G d f f ? N b d d

average

Goodness of fit? Not bad and its just the long term monthly

average

Sales Tax (ESPLOST)


40

50

60

70

80

90

100

110

1-Ja

n-99

1-Ju

n-99

1-N

ov-9

9

1-A

pr-0

0

1-Se

p-00

1-Fe

b-01

1-Ju

l-01

1-D

ec-0

1

1-M

ay-0

2

1-O

ct-0

2

1-M

ar-0

3

1-A

ug-0

3

1-Ja

n-04

1-Ju

n-04

1-N

ov-0

4

1-A

pr-0

5

1-Se

p-05

1-Fe

b-06

1-Ju

l-06

1-D

ec-0

6

1-M

ay-0

7

1-O

ct-0

7

1-M

ar-0

8

1-A

ug-0

8

1-Ja

n-09

1-Ju

n-09

1-N

ov-0

9

1-A

pr-1

0

1-Se

p-10

1-Fe

b-11

1-Ju

l-11

1-D

ec-1

1

1-M

ay-1

2

1-O

ct-1

2

1-M

ar-1

3

1-A

ug-1

3

1-Ja

n-14

1-Ju

n-14

1-N

ov-1

4

1-A

pr-1

5

1-Se

p-15

ESPL

OST

Rev

enue

in $

Tho

usan

ds

Month Year

Sample District 1999-2015

gOne Year Moving

Average

Sales Tax (ESPLOST) – Helpful Data

• Counties governments might be producing sales tax forecasts, even within their budget documents, which can help districts predict their portion of the sales tax revenues.

• Speak to local business owners and parents. • Georgia budget documents forecasts state sales tax revenues into the future. Check and see how well state sales tax collections track with your counties.


Number of Students

• Importance: Drives state funding (QBE), expenditures, staffing decisions...

• Data Sources: Previous student counts, data on <5 y.o. children in your area, and historic drop out and migration info

• Volatility: Low– Based on steady drop outs, in and out migration

• Method to Try: Linear trend (OLS) to start – move to age specific calculations improve forecast


Number of Students


8,600

8,800

9,000

9,200

9,400

9,600

9,800

10,000

fall 2006 fall 2007 fall 2008 fall 2009 fall 2010 fall 2011 fall 2012 fall 2013 fall 2014 fall 2015

Fall

FTE

Cou

nt

School Year

District A

Looks like a good fit and would forecast steadily declining student

counts

Maybe we could dig a little deeper?

Number of Students


600

650

700

750

800

fall 2006 fall 2007 fall 2008 fall 2009 fall 2010 fall 2011 fall 2012 fall 2013 fall 2014 fall 2015

DISTRICT A

First Grade Second Grade Third Grade Forth Grade

gIncreasing early

gradesFollowed by Increasing

older Grades

Should we expect steadily declining student counts?

Number of Students – Helpful Data

• Georgia Office of Planning and Budget provides population by age group forecasts for all of the counties in Georgia

• Other publicly available datasets do the same by race.


Forecasting State Funding – QBE

• Predicating future student population, knowledge of proposed funding formula changes, and awareness of state budget shortfalls are the ways to forecast future QBE/State funding.


State Funding – QBE


$6.75

$7.25

$7.75

$8.25

$8.75

$9.25

$9.75

$10.25

$10.75

$11.25

2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014

QB

E F

orm

ula

Ear

ning

s $

in M

illio

ns

Schoolyear

What about expenditures?

• Forecasting expenditures – Based on revenue and student population forecasts

• Other changes would be based on district decisions on curriculum and improvement efforts.


Questions?

Contact InformationNick [email protected]


Documents

Nick Warner Forecasting School Districts’ Revenues and ...cslf.gsu.edu/files/2015/12/GASBO-School-District-Forecasting... · Forecasting School Districts’ Revenues and Expenditures