UI Benefit Financing Seminar Division of Fiscal and Actuarial - - PowerPoint PPT Presentation

UI Benefit Financing Seminar Division of Fiscal and Actuarial Services U.S. DOL/ETA/OUI October 23-26, 2018

• 1. Understanding The Elements That

Comprise the Payment of Benefits in Your State.

• 2. Calculating a Forecast for Total State

Benefits.

2

3

Total Benefits Paid

STA TATE LA TE LAW W VA VARI RIABLES LES

Co Coverag age / / El Eligibility / / Ben enef efit L Lev evels / / Wag age B Bas ase / e / Tax Tax Rat Rates es / / Tr Trigger ers

ECO ECONOMIC SCE CENARIO VA VARI RIABLES

Tota

• tal L

l Labo bor Forc

• rce /

/ Tota

• tal U

l Unemployment / / Avera rage Ear Earnings / / Interest Rat Rate

Answering Three Questions:

1.

How Many People are Receiving Benefits?

2.

For How Long Do They Receive Benefits?

3.

How Much Do They Receive?

4

5

2,270 (32%) 7,091 1,958 (29%)

Total Unemployed JANUARY FEBRUARY MARCH

96,743 2,313 (32%) 808 492

UI First Payments

200

UI Exhaustees

164

Not in Labor Force

161 95,439

Total Labor Force

160,037 161,494 161,548 7,189 6,671

Unemployment Flow - First Quarter CY2018 (000)

392 95,549 Unemployment Insurance Recipients

a) How Much Do

Average Weekly Benefit Average Weekly

They Receive?

Benefit

a) How Long Do

Total Weeks Average

They Claim?

Compensated Duration

a) How many

Insured Unemployment Rate First

Claimants?

(Avg. # of Claimants / Week)

Payments

6

Total Benefits 1 2

Unemployment Trust Fund Modelling

Total Labor Force / Total Unemployment /

• Avg. Wages / Interest Rate

Economic Scenario

Coverage / Eligibility / Benefit Levels / Wage Base / Tax Rates/ Triggers

State Law Variables

Weeks Claimed First Pays Average Duration Weeks Compensated

Total Benefits Paid

• Avg. Weekly Benefit
• r

Unemployment Trust Fund Modelling

Weeks Claimed

Total Labor Force / Total Unemployment /

• Avg. Wages / Interest Rate

Economic Scenario

Coverage / Eligibility / Benefit Levels / Wage Base / Tax Rates/ Triggers

State Law Variables

TUR IUR Insured Unemployment Covered Employment

* Time Period (Weeks)

1) 1) IUR

IUR = f( f(TUR UR, o

• ther va

variables) IU IU = IUR IUR x x Cove vered Empl ploy

• yment

2) 2) IU

IU = f( f(TU, U, o

• ther va

variables)

3) ) IU/TU

TU = = f(TU TUR, R, o

• ther

er var ariables) IU = IU = IU/ IU/TU x x TU

 Reg

Regres ession M Model eling

• Meth

thod f for de

• r dete

termin ining th the re rela lati tionships amon mong tw two or more

• or more vari

riables

• Meth

thod f for f

• r fore
• recasting f

future re v valu lues of

• f on
• ne

var ariab able ( (dep epen endent), g given en the v e val alues o

• f t

the e

• ther v

varia iable bles ( (indepe pende dent) t)

• *Assume

mes h histor toric ical r l relation ionship ips c contin tinue i in the futur ure

Time Series UI Data: IUR & TUR

0.00 1.00 2.00 3.00 4.00 5.00 0.00 2.00 4.00 6.00 8.00 10.00 Y -- Dependent Variable -- IUR X -- Explanatory Variable -- TUR

TUR - IUR Scatter Plot

y = a + b * x 0.00 1.00 2.00 3.00 4.00 5.00 0.00 2.00 4.00 6.00 8.00 10.00 Y -- Dependent Variable -- IUR X -- Explanatory Variable -- TUR

TUR - IUR Scatter Plot w/ Trend

0.00 1.00 2.00 3.00 4.00 5.00 6.00 7.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00 Y -- Dependent Variable -- IUR X -- Explanatory Variable -- TUR

Residuals

0.00 0.50 1.00 1.50 2.00 2.50 3.00 3.50 4.00 4.50 5.00 6.00 7.00 8.00 9.00 10.00 Y -- Dependent Variable -- IUR X -- Explanatory Variable -- TUR

TUR - IUR Scatter Plot w/ Trend

 y

y = a a + b*x

• IUR

R = = I Interce cept + + Co Coeffici icient(b) * * TU TUR

 y = a + b1*x1

*x1 + b2*x2 *x2 + … + bN bN*xN xN

1. 1.

Identify Potentia ial l Ex Expla lanatory ry Va Varia riable les

2. 2.

Co Colle llect ct Data (BLS LS, LM LMS, UI Pro rogra gram)

3. 3.

Plot and and Revi view Dat Data a and and Relat ationships

4. 4.

Choos

• ose a tim

ime perio riod

5. 5.

Choose se specifi fication(s) n(s)

• Ad

Add/Drop v var ariables in S Stepwise Ap Approach

6. 6.

Va Valid lidate

7. 7.

Te Test fore reca casts

8. 8.

Final nal mo mode del

9. 9.

Devel elop assum sumptions/ ns/sc scena enario

10.

• 10. Forecast

st

 Bas

ased on knowl

• wledge of UI pro

rogra ram

• State p

e pro rogra ram id idio iosyncrasies

• State/Nationa

nal E Econo nomy & Rece cessions ns

• Seasonalit

lity

• Structura

ral E l Economic ic or Progra rammatic ic S Shif ifts

 Avail

ilab abili lity o

• f Data f

a for R Regres ressi sion a n and nd Forec ecas astin ing

• Hi

Histor torical Period

• d Data

ta &

• Fo

Forecast P Period Dat d Data

 Must st h have o

• r produce projections/

s/assu assumptions s of e each vari riable u used in in re regre ression e equation in in o

• rd

rder to f fore recast.

 TU

TUR R – Inclu ludi ding l g lags/l /leads ds

 La

Lagged IUR IUR

 Exhaustion

tions

 Extende

ded d UI B Benefit A it Availa labili bility ty

 State l

law/admin dminis istra tration tion v varia iable bles

 Demogra

• graph

phic ics / / changin ging i g industri tries

 Long

g term u m unempl mploy

• yed

 Manufactu

turin ring E g Employ loyment

 Union

• niz

ization tion

 Job L

Losers rs ( (alternativ tive t to u unemploy mployed) d)

Year.Qtr TUR IUR TU IU … D1 D2 D3 Rececession 1998.3 3.28 1.97 57,666 31,112 … 1998.4 2.64 1.73 46,007 27,420 … 1 1999.1 3.38 2.55 58,598 40,715 … 1 1 1999.2 2.98 1.94 52,254 31,050 … 1999.3 2.68 1.96 47,438 31,553 … 1999.4 2.47 1.63 43,435 26,337 … 1 2000.1 3.13 2.31 55,011 37,580 … 1 1 1 2000.2 2.40 1.60 42,403 26,147 … 1 2000.3 2.27 1.70 40,250 27,840 … 1 2000.4 1.78 1.58 31,188 25,908 … 1 1

0.00 1.00 2.00 3.00 4.00 5.00 6.00

IUR Not Seasonally Adjusted

0.00 2.00 4.00 6.00 8.00 10.00 12.00

IUR Not Seasonally Adjusted TUR Seasonally Adjusted

0.00 2.00 4.00 6.00 8.00 10.00 12.00 0.00 1.00 2.00 3.00 4.00 5.00 6.00

IUR Not Seasonally Adjusted TUR Seasonally Adjusted

 At

At l lea east st 6 to 8 8 yea ears

 10

10 to 12 12 years rs is is ge genera rall lly adequate but use dis iscre cretio ion to:

• Include A

AT T LEA EAST o T one r e reces ecession ( (Co Consider magnitu itude de and c change ges i in U UI r relati tion

• nships

ips)

 Look for s

stat ate law c aw chang hanges

 Lo

Look for r other r stru ruct ctura ral l ch changes

 Add/

d/Dr Drop o p one variabl ble a at a time

• Stepwi

pwise appr approach

 “Stepwis ise M Model Tracker.xls xlsx” x”

 Chec

eck: k:

• Coefficie

icient nts

• Adjust

sted R R-Square uare

• Residuals

uals

 Inc

nclude vari ariables o

• f int

interest / / hig high im importance

 Significance of Individual Variables:

• t Statistic = Coefficient / Standard Error
• t Statistic > 2 or P-value of Coefficient < 0.05
• Look for correct sign (+/-) & magnitude of

coefficient

 Adjusted R Square:

• Reflects proportion of variation in dependent

variable (TUR) explained by regression line.

• Useful to compare performance across multiple

regressions

• Larger Adj. R Square = “better” fit