Utilization of Research in Policymaking for Graduated Driver Licensing

GOVERNMENT, POLITICS, AND LAW

Effective Use of Frameworks andResearch to Advance Policy

An Analysis of Public Health Policy and Legal IssuesRelevant to Mobile Food VendingJune M. Tester, MD, MPH, Stephanie A. Stevens, JD, Irene H. Yen, PhD, MPH, and Barbara A. Laraia, PhD, MPH, RD

Mobile food vending is

a component of the food envi-

ronment that has received little

attention in the public health

literature beyond concerns

about food sanitation and hy-

giene issues. However, sev-

eral features of mobile food

vending make it an intriguing

venue for food access.

We present key components

of mobile vending regulation

and provide examples from 12

US cities to illustrate the vari-

ation that can exist surround-

ing these regulations.

Using these regulatory fea-

tures as a framework, we high-

light existing examples of

‘‘healthy vending policies’’ to

describe how mobile food ven-

ding can be used to increase

access to nutritious food for

vulnerable populations. (Am

J Public Health. 2010;100:2038–

2046. doi:10.2105/AJPH.2009.

185892)

THERE IS A GROWING FOCUS

on the role of the food environ-ment for the obesity epidemic.1 Inparticular, there is a need forgreater access to nutritious foodand more limits on energy-densefood with low nutritional value.Greater relative availability of

nutritious food in local food stores isassociated with greater intake ofthose foods.2 Although there aresome existing strategies to increasepurchase of fruits and vegetableswithin grocery stores,3 access tostores with nutritious food remainsan issue. Supermarkets are morelikely to carry fresh produce,4 butthey are less likely to be found inlow-income neighborhoods andcommunities of color.5,6 There area variety of factors that have his-torically been barriers to super-market location in lower-incomeurban areas,7,8 and the rural poorappear to have even less access tosupermarkets than do their metro-politan counterparts.9 Neighbor-hoods without supermarkets tendto have small corner stores or con-venience markets that have limitedinventories of nutritious food.10

Although public health scholarshave given some attention to cor-ner store interventions, mobilefood vending has received littleattention in the public health lit-erature beyond concerns aboutfood sanitation and hygiene is-sues.11,12 But several features ofmobile food vending make it anintriguing venue for food access.Unlike a corner store, mobile foodvendors sell a small range of

merchandise. Specialized vendors(e.g., vendors selling only fruit) canmore easily ensure fresh merchan-dise because of rapid turnover.Because these vendors are mobile,they have the capacity to reachplaces that otherwise lack access tofood establishments or food stores.Mobile food vendors have beenfound to converge around schoolsto sell foods to students afterschool.13 Mobile vendors appear tobe a familiar phenomenon in urbanas well as rural communities withlarge numbers of Latino immi-grants,13–15 and understanding howto encourage the sale of nutritiousfood rather than energy-dense foodwould be valuable to these commu-nities and others that have dispro-portionately high rates of obesity.16

The need for increased accessto nutritious food and the uniquefeatures of mobile food vendinglead to some compelling questions.Could mobile vendors contributeto the accessibility of nutritiousfood, particularly for underservedand vulnerable communities?Could a mobile cart or truck func-tion like a supermarket produceaisle on wheels? We focused onhow local government law andpolicy could support healthy mo-bile vending mainly because the

law has the advantages of broaderapplication and permanence. Here,we present key components ofmobile vending regulation by usingexamples from the municipal codesof the 10 most populous US citiesto illustrate the variation that canexist surrounding these regulations.Then, using this framework ofregulatory features, we describehow mobile food vending can beused to increase access to nutri-tious food for vulnerable urbanpopulations, highlighting 2 citiesfrom this list and discussing 2additional noteworthy policy ex-amples. We chose to limit ourscope to mobile food vendors inurban settings because, eventhough the potential for mobilevending to increase nutritious foodaccess in rural areas is also worthexploring, the unique characteris-tics of rural settings such as lowpopulation density and differencesin local government authority war-rant a separate examination thattakes these features into account.

MOBILE FOOD VENDING INURBAN HISTORY

Mobile food vending is a world-wide phenomenon. Common inLatin America and Asia, it is often

2038 | Government, Politics, and Law | Peer Reviewed | Tester et al. American Journal of Public Health | November 2010, Vol 100, No. 11

an opportunity for individuals tomake a living with a small enter-prise.17 Mobile vendors have alsoexisted in the United States formanyyears, and records from New YorkCity as early as1691show that streetvendors (‘‘hucksters’’) were forbid-den from selling until competingpublic markets had already beenopen for 2 hours. New York Cityvendors persisted despite a com-plete ban in1707, and their growthwas closely connected to immigra-tion.18 In the1800s, whereas indoorretail stores catered to middle- andupper-class customers, street ven-dors catered to poor, mostly foreign-born residents, and, for many im-migrants with little English-speakingskills, the neighborhood pushcartbusiness was an accessible way toearn a living.14 Vendors started toestablish informal market areas, andstreet vending thrived in New YorkCity in the 1880s through the1920s, but was almost completelyabolished in the1930s when en-closed market buildings were builtto ‘‘tidyup thestreets’’ in preparationfor the World’s Fair.19 It is interest-ing to note that in1925, the majorityof fruit and vegetable peddlers wereJewish immigrants (63%), and therest were primarily Italian (32%).20

Mobile food vending continues to-day, often in communities withmany foreign-born residents, andmunicipal codes still focus on manyof the same issues, such as compe-tition with local businesses andprohibiting vendors from operatingin ‘‘upscale’’ neighborhoods.

MOBILE VENDINGREGULATION

Mobile vending regulationstypically include a number of

standard requirements regardingfood safety, permits and fees,vendor location, and traffic safety.With the exception of state retailfood codes, mobile vending istypically regulated at the local (cityor county) level. There is typicallycitywide regulation found in mu-nicipal codes, and a city’s overallapproach to regulation of mobilefood vending can range from re-strictive to permissive. Municipalcodes can also grant city agenciesthe authority to regulate mobilevending with a limited context,as in Kansas City, Missouri, andSan Francisco, California, wherepark and recreation departmentsregulate mobile vending inparks.

We examined the municipalcodes of a subset of US cities tocompare mobile vending regula-tions. For ease of comparison wechose the 10 most populous citiesranked by 2007 estimates.21 Mu-nicipal codes were all availableonline, either hosted by the city’sown Web sites or via an onlineservice that hosts city ordi-nances.22,23 Between October andDecember of 2008, we searchedfor all sections pertaining to mobilefood vending to identify languagerelevant to the 4 major a prioridomains listed in Table 1. Thesedomains pertained to health andsafety, permits and fees, location-based regulation, and whetherthere were any nutrition incentives.From the 10-city analysis we iden-tified 2 cities, Chicago, Illinois, andNew York, New York, that hadnutrition incentives for mobilevending carts. We assessed healthyfood policies for these cities plus 2additional cities (Kansas City andSan Francisco) that we identified

through our involvement in theNational Policy and Legal AnalysisNetwork as cities with a healthymobile vending policy (Table 2).Highlighted in the following sec-tions are examples of the variationin existing policy with respect to the3 domains of health and safety,permits and fees, and location-based regulation. (Nutrition incen-tives, when present, are discussedin the subsequent section wherecomponents of healthy mobilevending policies are considered.)

Health and Safety Regulation

Municipal codes regarding mo-bile vending must comply withapplicable state laws. Most statesregulate the health and safety ofmobile vending under their retailfood codes, and state retail foodcodes often charge local agencieswith carrying out the code’s pro-visions.94–97 State retail food codesare focused on protecting the publicfrom food-borne illness, with pro-visions designed to prevent con-tamination and promote hy-giene.98,99 To promote uniformfood safety regulations, the USFood and Drug Administration de-veloped a model Food Code forstates to adopt,100 under whichmobile vending facilities are con-sidered a type of food establishmentand, therefore, subject to the code’shealth and safety provisions.101

To further promote safe foodhandling practices, vendors areoften required to operate froma commissary. A commissary isa centralized facility where ven-dors clean and store their vehiclesas well as sanitize their equip-ment.102 The commissary may alsoserve as a common kitchen fromwhich vendors can prepare their

food, as laws generally prohibitvendors from preparing food athome.103 Local authorities (usuallymandated by state law) may alsorequire inspection of commissariesto ensure compliance with food-safety laws.104

Permits and Fees

Local governments requirevendors to obtain a license orpermit. To obtain a permit, fre-quently the vending vehicle mustpass inspection by the local healthdepartment or other designatedauthority. Municipalities chargea fee for vendor permits andamounts can vary greatly. In ad-dition, local laws may cap thenumber of mobile vending permitsallowed at any one time. For ex-ample, until recent legislationadded new permits for fruit andvegetable vendors, New York Citylaw had historically limited thetotal number of general permitsfor mobile food vendors at3100.105 Permits continue to be ingreat demand in New York City andthere is a sizeable waiting list forprospective vendors.106,107

Location

Local governments commonlyrestrict where mobile vendorsmay operate. Some cities havecomplex laws regulating vendingstreet by street. For example,Philadelphia, Pennsylvania, codeexplicitly refers to the specificstreets within the central part ofthe city where vendors areallowed to conduct business.108

Others might have a designatedarea for vending or allow vendingcitywide but have certain restric-tions. For example, Phoenix, Ari-zona; San Antonio, Texas; and San


November 2010, Vol 100, No. 11 | American Journal of Public Health Tester et al. | Peer Reviewed | Government, Politics, and Law | 2039

TAB

LE1

—C

ompa

riso

nof

Mob

ileFo

odVe

ndin

gLa

ws

inth

e1

0M

ost

Pop

ulou

sU

SC

itie

s:2

00

8

Heal

than

dSa

fety

Perm

itsor

Othe

rRe

gula

tions

Loca

tion

Nutri

tion

Regu

latio

n

City

Are

Vend

ors

Requ

ired

to

Oper

ate

From

aCo

mm

issa

ry?

Are

Vend

ors

Subj

ect

to

Insp

ectio

n?

Fees

for

Mob

ileVe

ndor

Perm

itsor

Lice

nse

Othe

r

Spec

ial

Regu

latio

ns

Rest

rictio

n

onDu

ratio

n

ofVe

ndor

Stop

s

Hour

sW

hen

Vend

ors

Are

Allo

wed

to

Oper

ate

Rest

rictio

ns

onVe

ndor

Prox

imity

to

Scho

ols

Nutri

tion

Ince

ntive

s

Chic

ago,

ILCo

mm

issa

ryor

othe

rlic

ense

d

fixed

food

serv

ice

esta

blis

hmen

t.24

Yes.

Vend

ors

mus

t

pass

insp

ectio

n

befo

relic

ense

will

beis

sued

.24

$165

ever

y2

yfo

r

‘‘ped

dler

s’’o

ffru

its

and

vege

tabl

es25

;

othe

rwis

e$2

75,

paya

ble

ever

y2

y.26

No.

Nore

gula

tion.

7AM

to5

PMfo

r

pedd

lers

offru

its

and

vege

tabl

es.27

Othe

rwis

e,

ther

eis

no

rest

rictio

non

hour

s.

Nore

gula

tion.

Vend

ors

selli

ngon

lyfru

its

and

vege

tabl

espa

y

are

duce

dpe

rmit

fee.

28

Dalla

s,TX

Yes.

29Ye

s.30

$100

for

mos

t

vend

ors,

but

$465

for

am

obile

food

prep

arat

ion

vehi

cle

such

asa

‘‘hot

truck

.’’31

Vend

ors

mus

t

prov

ide

am

onth

ly

itine

rary

indi

catin

g

wher

eth

eyin

tend

to

oper

ate32

and

mus

t

beab

leto

prov

ide

proo

fof

liabi

lity

insu

ranc

e.33

Vend

ors

may

not

stop

for

mor

eth

an

60co

nsec

utive

min

utes

ora

tota

lof

3h

in1

loca

tion

ina

24-h

perio

d.31

Nore

gula

tion.

Nore

gula

tion.

No.

Hous

ton,

TXYe

s,an

d

com

mis

sarie

s

are

requ

ired

to

keep

serv

icin

g

reco

rds

for

each

mob

ile

vend

or.34

Yes.

Vend

ors

mus

t

pass

insp

ectio

n

befo

rere

ceivi

ng

ape

rmit,

and

then

are

subj

ect

toin

spec

tion

with

out

notic

e.35

,36

$200

for

ape

rmit,

$310

for

a‘‘m

edal

lion’

’to

be

plac

edon

the

vend

ing

vehi

cle,

and

a$2

00

elec

troni

cm

onito

ring

syst

ems

fee

for

‘‘unr

estri

cted

mob

ile

food

units

.’’37

Vend

ors

inth

e

down

town

dist

rict

need

perm

issi

on

from

abut

ting

stor

e

owne

rs.38

Ape

rson

certi

fied

insa

fefo

od

hand

ling

mus

tbe

ondu

tyat

allt

imes

.39

Nore

gula

tion.

Vend

ors

may

desi

gnat

ea

site

for

24-h

use.

40

Nore

gula

tion.

No.

Los

Ange

les,

CA

N/A

N/A

N/A

Vend

ors

mus

t

esta

blis

ha

‘‘spe

cial

side

walk

vend

ing

dist

rict’’

;at

pres

ent,

nodi

stric

tex

ists

.41

N/A

N/A

N/A

N/A

New

York

,NY

Yes.

Allv

endo

rs

mus

top

erat

efro

m

aco

mm

issa

ry,

depo

t,

orot

her

licen

sed

faci

lity.42

Yes.

Vend

ors

are

not

allo

wed

to

oper

ate

until

they

have

pass

ed

insp

ectio

n.43

Perm

itsar

eva

lidfo

r

2y,44

$50

iffre

shfru

its

orve

geta

bles

or

prep

acka

ged

food

45;

$100

iffo

odis

proc

esse

don

site

.46

Gree

nCa

rtve

ndor

s

only

inun

ders

erve

d

area

s,47

and

mus

t

have

educ

atio

nal

broc

hure

s.48

Nore

gula

tion.

Varie

sby

loca

tion.

49No

regu

latio

n.Gr

een

Cart

vend

ors

selli

ngwh

ole

fruits

and

vege

tabl

es.50

Gree

nCa

rtve

ndor

s

have

prio

rity

onpe

rmit

waiti

nglis

t.51

Phila

delp

hia,

PAYe

s.52

Yes.

Vend

ors

mus

t

subm

itto

anof

ficia

l

insp

ectio

n53an

d

perfo

rma

self-

insp

ectio

nev

ery

3m

o.54

$125

annu

ally

for

vend

ors

onfo

ot;

othe

rwis

e$3

00

annu

ally

for

all

othe

rve

hicl

es.55

No.

Nore

gula

tion.

7AM

to12

AM.56

Nore

gula

tion.

No.

Cont

inue

d



TAB

LE1

—C

onti

nued

Phoe

nix,

AZYe

s.Ve

ndor

sm

ust

repo

rtda

ilyto

a

com

mis

sary

.57

Yes.

Vend

ors

mus

t

bein

spec

ted

at

leas

tev

ery

6m

o

unde

rth

eAr

izona

Food

Code

.58

$250

first

-tim

e

licen

seap

plic

atio

n

fee,

then

$30/

y.59

Ther

eis

also

a1-

time

fee

for

crim

inal

inve

stig

atio

n

finge

rprin

ts.60

Vend

ors

may

not

oper

ate

onan

y

stre

etab

uttin

ga

publ

icpa

rkwi

thin

150

ftof

ala

wful

ly

esta

blis

hed

park

conc

essi

on.61

Vend

ors

may

not

stop

for

mor

eth

an1

h

with

inan

8-h

perio

don

any

publ

icst

reet

oral

ley.62

6AM

to2

AMon

priva

tepr

oper

ty63

;

the

late

rof

7PM

orsu

nset

and

befo

resu

nris

eon

publ

icpr

oper

ty.64

Vend

ors

onpr

ivate

prop

erty

may

not

oper

ate

with

in

300

ftof

any

scho

olbe

twee

n

6AM

and

5PM

,65

orwi

thin

600

ftof

any

scho

ol,

or

betw

een

7am

and

4:30

pm

when

loca

ted

onpu

blic

prop

erty

.66

No.

San

Anto

nio,

TXYe

s.Ve

ndor

sm

ust

oper

ate

from

a

com

mis

sary

,un

less

they

sell

food

that

exem

pts

them

from

this

prov

isio

n.67

Yes.

Vend

ors

are

subj

ect

toro

utin

e

unan

noun

ced

insp

ectio

ns.68

$48

to$3

50an

nual

ly

depe

ndin

gon

the

type

ofve

hicl

eus

edan

d

the

type

offo

odso

ld.69

Vend

ors

may

not

sell

with

in30

0ft

ofan

y

food

esta

blis

hmen

t

unle

ssth

eyob

tain

perm

issi

onfro

m

the

owne

r.70

Nore

gula

tion.

7AM

to30

min

afte

rsu

nset

in

resi

dent

iala

reas

.

7AM

to10

PM

inJu

ne,

July,

and

Augu

st.71

Vend

ors

may

not

sell

with

in30

0ft

ofan

y

scho

ol1

hbe

fore

,

1h

afte

r,an

ddu

ring

scho

olho

urs.

72

Vend

ors

selli

ng

whol

efru

itsor

vege

tabl

es,

fresh

fish,

orsh

rimp

dono

t

have

toop

erat

efro

ma

com

mis

sary

.73

San

Dieg

o,CA

Yes.

73Ye

s.74

$164

to$4

27an

nual

ly

depe

ndin

gon

the

type

ofve

hicl

eus

edan

d

the

type

offo

odso

ld.75

Units

prop

elle

dby

‘‘mus

cula

rpo

wer

eith

erhu

man

or

anim

al’’

cann

otbe

used

tose

ll

peris

habl

efo

od.76

Rest

rictio

nson

dura

tion

oftim

e

vend

ors

are

allo

wed

tost

opva

ries

bylo

catio

n.77

9AM

to8

PM.78

Vend

ors

may

not

oper

ate

with

in50

0ft

ofan

ypu

blic

scho

ol

betw

een

7AM

and

4PM

onre

gula

r

scho

olda

ys.79

Vend

ors

may

sell

farm

prod

uce

from

the

farm

prop

erty

with

out

payin

ga

perm

itfe

e.80

San

Jose

,CA

Yes,

orot

her

appr

oved

faci

lity.81

Yes.

82$4

18fo

ran

‘‘app

rove

d

loca

tion’

’ven

dor

perm

it;

$149

for

allo

ther

vend

ors,

plus

$45

for

anID

card

.83

Vend

ors

oper

atin

g

from

ade

sign

ated

‘‘app

rove

dlo

catio

n’’

mus

tha

velia

bilit

y

insu

ranc

e.84

Exce

ptfo

r‘‘a

ppro

ved

loca

tion’

’(st

atio

nary

)

vend

ors,

vend

ors

may

not

stop

in1

plac

e

for

long

erth

an15

min

ina

2-h

perio

d.85

10AM

to7

PMor

suns

et;

vend

ors

at

cons

truct

ion

orin

dust

rial

site

sar

eex

empt

from

this

regu

latio

n.86

Appr

oved

loca

tion

vend

ors

may

oper

ate

from

6AM

until

7PM

orsu

nset

.87

Vend

ors

may

not

oper

ate

with

in

500

ftof

any

scho

ol

prop

erty

.88

No.



Diego, California, restrict vendorsfrom locating near schools basedapproximately on school hours,whereas San Jose, California, pro-hibits vending near schools irre-spective of the time of day.

Local regulations also attemptto prevent vendors from compet-ing with restaurants or otherbusinesses. For instance, in someplaces, vendors must obtain writ-ten permission from any abuttingbusiness owners to locate nearthem.109 Local laws may alsoprohibit vendors from locating inclose proximity to certain sta-tionary businesses altogether,such as Chicago’s prohibition ofvendors within 1000 feet of theMaxwell Street Market.110 Theprocess in Los Angeles, California, is

particularly prohibitive for legalmobile food vending. First, LosAngeles law requires vendors toobtain the consent of at least 20%of the business owners and resi-dents in the area before the city willbegin the bureaucratic process ofestablishing a ‘‘special sidewalkvending district.’’111 The law thenrequires vendors to get writtenpermission from the propertyowner or tenant closest to wherethe vendor intends to locate,112 anda petition of 20% of the nearbyresidents and business owners canultimately close the vending dis-trict.113

Another common regulationis to require vendors to moveafter a designated interval oftime. In San Jose, some mobile

vendors are prohibited fromremaining in the same locationfor more than 15 minutes in a2-hour period.114 This type ofregulation may discourage mobilevending as constantly movingmakes it more difficult for vendorsto draw on regular customers oroperate efficiently.

Vendors must also comply withlocal and state vehicle or trafficsafety regulations. These regula-tions are generally aimed at pre-venting interference with flow oftraffic and ensuring pedestriansafety. For example, San Diego’scode prohibits vendors fromlocating or operating in anymanner that would ‘‘interferewith the free use of the publicright-of-way.’’115

COMPONENTS OF AHEALTHY VENDINGPOLICY

Using the regulatory compo-nents of health and nutrition reg-ulation, permits and fees, and lo-cation regulation, we describe howlocal government can utilize mo-bile food vending to increase ac-cess to nutritious food.

Health and Nutrition

Regulation

Health departments alreadyplay an important role in theregulation of mobile food vendingbecause of their duty to ensurefood safety. As an additional steptoward increasing access to nutri-tious food, health departments

TABLE 2—Examples of Existing Healthy Vending Policies Enacted Within US City Ordinances by City Agencies: 2008

Type of Policy City Health or Nutrition Regulation for ‘‘Healthy Vendor’’ Permits or Fees for ‘‘Healthy Vendor’’ Location for ‘‘Healthy Vendor’’

City ordinance New York, NY ‘‘Green Carts’’ program applies

only to vendors selling whole,

unprocessed fruits and vegetables.47

Increased city’s overall number

of permits to include 1000

designated Green Carts. Reduced

fee for Green Carts vendors.89

Special permit prioritizes selling

in underserved boroughs.47

City ordinance Chicago, IL Vendors selling fruits and vegetables

eligible for a permit at a reduced cost.90

Reduced fee of $165 every 2 y,

otherwise $275 every 2 y.91

Not specified.

City agency policy San Francisco,

CA, Parks and

Recreation Department

Favorable products: grown or produced

locally, are organic, minimally processed,

have no genetic modification, no

unnecessary antibiotics, no added

growth hormones, and meet animal

welfare or fair trade policies.92

$1000 per mo.92 City parks.92

City agency policy Kansas City, MO,

Parks and Recreation

Department

Food guidelines (per serving):

d £ 5 g of total fat

d £ 30 g carbohydrate

Beverage guidelines:

d water

d milk (1% or skim, any flavor)

d 50% or more fruit or vegetable juice

with no sweeteners

d £ 50 calories per 12 oz93

‘‘Healthier’’ vendors:

d ‡ 50% items meeting guidelines

d 50% reduced permit ($250/year)

‘‘Healthiest’’

d vendors ‡ 75% items meeting guidelines

d Full permit ($500), though have

roaming privileges. 93

‘‘Healthier’’ vendors are limited

to 1 city park. ‘‘Healthiest’’ vendors

have roaming permit for 3 city parks.93



could evaluate mobile vendors forcompliance with nutritional stan-dards. Health departments couldconfer special ‘‘healthy food ven-dor’’ status to vendors who meetnutritional standards, thus creat-ing a category of vendors who areeligible for other regulatory in-centives. There are a variety ofapproaches that health depart-ments could take to define keyterms. One step would be to re-serve designation of ‘‘healthy foodvendor’’ status to vendors carryinga threshold percentage of fooditems that meet the Food and DrugAdministration’s Nutrition Label-ing criteria for designation as a‘‘healthy’’ food. Per serving, quali-fying food items would need to below in fat (3 g or less) and saturatedfat (1 g or less), contain limitedamounts of sodium and cholesterol,and provide 10% of the daily valuefor vitamin A, vitamin C, iron,calcium, protein, or fiber.116,117

Alternatively, health departmentscould limit ‘‘healthy food vendor’’status to vendors who sell exclu-sively fresh produce. Because fresh,uncut produce is exempt fromregulation under many state foodretail codes, this is a relatively easylegal intervention for some locali-ties. For example, New York Citypassed Local Law 9 in March of2008, amending the existing mu-nicipal code to create 1000 addi-tional mobile vending permits for‘‘Green Carts.’’118 A Green Cart isone selling exclusively whole, uncut,and unprocessed produce. Therehas been a high demand forobtaining permits as a Green Cart,119

and Green Cart vendors alsohave priority on the city’s overallwaiting list for vendor permits(Table 2).104

The parks and recreation de-partments in Kansas City and SanFrancisco are both encouragingthe sale of ‘‘healthy foods’’ thoughusing different criteria. KansasCity’s Department of Parks andRecreation has a policy that allowsvendors to sell in the city’s parks,provided that their food complieswith explicitly defined nutritionguidelines.93 Vendors with at least50% of their foods meeting theseguidelines are deemed ‘‘healthier’’vendors, and vendors with at least75% of foods meeting these guide-lines are considered ‘‘healthiest.’’Vendors meeting these criteriaqualify for reduced fees and areallowed to sell in more areas.93

San Francisco’s Parks and Rec-reation Department recently re-leased a request for proposalssoliciting specialty food carts withan interest in selling within thecity’s public parks.92 This requestfor proposals for specialty foodpushcarts focuses on ‘‘health,’’ butdoes not specifically require food tobe nutrient-rich or low in caloriesor fat. The request for proposalsstates that the department will‘‘view favorably menus that incor-porate healthy, sustainably grownfood and beverages.’’ Priority foodsare those that are grown or pro-duced locally, are organic, are min-imally processed, have no geneticmodification, have no unnecessaryantibiotics, have no added growthhormones, and meet animal welfareor fair trade policies.92

Permits and Fees

Cities often set a limit on thetotal number of permits for ven-dors that are allowed at any giventime. This is presumably to preventsaturation from mobile vendors.

One approach toward a healthyvending policy would be to dis-proportionately increase the num-ber of permits allowed for vendorsthat sell nutritious foods. This ap-proach was taken under the NewYork City Green Carts program.

Another potential healthy ven-dor policy is for local governmentto subsidize, waive, or reduce per-mit fees that a prospective vendorwould pay if the food that they sellmeets nutritional requirements. InChicago, vendors that sell onlyfruits and vegetables pay a reducedpermit fee of $160 instead of $475every 2 years.28 Kansas City ven-dors selling in parks who qualify asbeing ‘‘healthier’’ vendors (with atleast 50% of food meeting nutri-tional guidelines) are given a 50%discount on their vending permit(a savings of $250).93

Location Regulation

Another approach is to modifyrestrictions on where vendors areallowed to operate to give ‘‘healthyfoods vendors’’ a geographic ad-vantage over other vendors sellingless nutritious items. Kansas Cityvendors selling in parks whoqualify as being ‘‘healthiest’’ ven-dors are given a special ‘‘roaming’’permit that allows them to sell in 3parks instead of just 1 park.93 It isalso possible to translate this sameprinciple of geographic advantageto increase sale of nutritious foodnear schools.

To address racial, ethnic, oreconomic disparities in access tonutritious food, a local governmentcan also create incentives for‘‘healthy foods vendors’’ to locate inneighborhoods most in need ofincreased access to fresh produceand other nutritious food. The

Green Carts Program in New YorkCity seeks to address the disparityin access to healthful food bydesignating a greater number ofGreen Cart permits in neighbor-hoods with historically low accessto fresh fruits and vegetables(Figure 1).120

CHALLENGES

Healthy mobile vending poli-cies face several potential chal-lenges. First, such policies requiresufficient infrastructure for en-forcement. Increasing the numberof available permits for an existingtype of vendor necessitates in-creased capacity to administerthese permits and resources toenforce the new policy. Addition-ally, the presence of ‘‘healthyfoods vendors’’ creates the needfor regulation and enforcement ofnutritional quality beyond currentregulations, which are focusedsolely on food safety and hygiene.But even though additional infra-structure requires an investmentof resources, a healthy vendingpolicy has the potential to createnew job opportunities that wouldin turn generate tax revenue.

A second challenge is that thepresence of extra permits or in-centives for certain types of mo-bile vendors may create resent-ment from other vendors who donot have the same privileges andalso can create fears of competi-tion from nearby store ownerswho may have nutritious items ontheir shelves.121,122 Additionally,some vendors are undocumentedimmigrants who are earning a livingby conducting a business with rel-atively low overhead costs. Forthese vendors, increased attention



on their mobile vending businessmay be unwelcome.

A third challenge is the inherentdifficulty in establishing a mean-ingful definition of ‘‘healthy food’’and determining whether thisdefinition will lead to the con-sumption of foods with a highernutritional value. The guidelinesin the Kansas City Parks and Rec-reation Department regulationsinclude very strict definitions re-garding calories and fat. New YorkCity’s Green Carts program clearlylimits itself to nutritious food by

focusing only on produce. SanFrancisco’s Department of Parksand Recreation defines ‘‘healthyfood’’ with a focus on sustainabil-ity. Although this supports a moresustainable food system, this ap-proach does not ensure that thefoods sold would be any lower in fator calories than standard fast food.

A final challenge lies in whethera healthy vending policy actuallyincreases access for populations inneed of improved access to nutri-tious foods. There is a strong needfor increased access to nutritious

food among low-income commu-nities and communities of color.An ideal healthy vending policywould attract vendors to provideservices within these communities.However, if permits come withfees that are prohibitively steep,or if the food deemed ‘‘healthy’’ istoo expensive (or unfamiliar) tovendors or customers, a healthyvending policy may be unsuccess-ful in optimally targeting thecommunities most in need of in-creased access to healthy food.

WHERE TO GO FROM HERE

As healthy vending initiativessuch as the New York City GreenCarts Program develop, research isneeded to evaluate the effects ofthese natural policy experiments.Specifically, we need to understandat a population level whether thesepolicies actually result in increasedaccess to healthier foods, andwhether they lead to improved di-etary intake. Feasibility and sus-tainability of such programs alsoneed to be documented and un-derstood. Vulnerable populationsthat experience a higher prevalenceof obesity, such as low-income andethnic minority communities, area particular research priority area.In light of the current obesity epi-demic among youths and the factthat students appear to make pur-chases at vendors after school,13

addressing the relationship of mo-bile food vending specifically toyouths should also be a priority.

Additionally, there is a need tostudy not only consumer accept-ability of mobile-vended nutritiousfood, but also how competitivethese food items can be whencompared with less-nutritious

options. Previous research withvending machines showed thatreductions in price of low-fat itemsin vending machines led to theirincreased sale compared withhigh-fat options.123 Similar experi-mental work looking at the sale ofnutritious items in close proximityto less-nutritious options would bevaluable.

Legal Community

This article serves as thegroundwork for exploring the role,benefits, and practical limitationsof using mobile food vending reg-ulation to improve access to nu-tritious food. More work is neededto examine the balance betweenfully realizing the positive poten-tial of mobile food vendors andnot creating undue burdensfor municipalities, regulatoryagencies, or vendors themselves.

There is also a need for techni-cal expertise and guidance fromthe legal community to create thetools needed to translate desiredchanges into local policy. In recentmovements such as the increasingadoption of soda-free school dis-tricts, public health lawyers havebeen instrumental by providingmodel ordinances with exemplarlanguage that can be used by localgovernments to implement thedesired health-promoting policy.

Community Action,

Leadership, and Political Will

Finally, it is not enough to pro-pose novel ways to regulate mo-bile vending and hope that localgovernments take up the cause.Obtaining the support and politicalwill to enact new policies is critical.Advocates may need buy-in froma range of constituents, including

Source. New York City Department of Health and Mental Hygiene Web site120; used with

permission.

FIGURE 1—Map of designated areas for New York City’s specially

permitted Green Carts that sell fresh produce in underserved

areas: 2008.



the business community, law en-forcement, or health departmentofficials, to get a healthy mobilevending policy successfully sup-ported by local governance bod-ies. For example, advocates withthe Healthy Eating Active Com-munities collaborative in SantaAna and in South Los Angeles,California, have worked with ven-dors and city officials alike to un-derstand and convey the needs ofvendors to have incentives forselling nutritious foods. Garneringthe support of a diverse group ofinterested parties will create thepolitical climate necessary to enactinnovative healthy mobile vend-ing policies as part of an over-all strategy to improve accessto nutritious food in vulnerablecommunities. j

About the AuthorsJune M. Tester is with the Department ofPreventive Cardiology, Children’s Hospitaland Research Center Oakland, Oakland,CA. Stephanie A. Stevens is with PublicHealth Law and Policy, Oakland, CA. IreneH. Yen and Barbara A. Laraia are with theDepartment of Medicine, University ofCalifornia, San Francisco.

Correspondence should be sent to JuneTester, MD, MPH, Children’s Hospital &Research Center Oakland, Department ofCardiology, Healthy Hearts Clinic, 74752nd St, Oakland, CA 94609 (e-mail:[email protected]). Reprints can be orderedat http://www.ajph.org by clicking the‘‘Reprints/Eprints’’ link.

This article was accepted February 27,2010.

ContributorsJ. M. Tester originated the study and ledthe writing. S.A. Stevens led the legalanalysis and contributed significantly tothe writing. I.H. Yen and B. L. Laraiaassisted with the study and analyses.

AcknowledgmentsThe Robert Wood Johnson Foundationprovided funding for this work through

a grant from Healthy Eating Research(grant 63049) and through the NationalPolicy and Legal Analysis Network toPrevent Childhood Obesity (grant62083).

We thank Gregg Kettles for commentson an early version of this article.

Human Participant ProtectionNo human participants were involved inthis analysis.

References1. Story M, Kaphingst KM, Robinson-O’Brien R, Glanz K. Creating healthy foodand eating environments: policy and en-vironmental approaches. Annu Rev PublicHealth. 2008;29(1):253–272.

2. Cheadle A, Psaty BM, Curry S, et al.Community-level comparisons betweenthe grocery store environment and in-dividual dietary practices. Prev Med.1991;20(2):250–261.

3. Glanz K, Yaroch AL. Strategies forincreasing fruit and vegetable intake ingrocery stores and communities: policy,pricing, and environmental change. PrevMed. 2004;39(supp 2):S75–S80.

4. Sallis JF, Nader PR, Rupp JW, AtkinsCJ, Wilson WC. San Diego surveyed forheart-healthy foods and exercise facilities.Public Health Rep. 1986;101(2):216–219.

5. Powell LM, Slater S, Mirtcheva D, BaoY, Chaloupka FJ. Food store availabilityand neighborhood characteristics in theUnited States. Prev Med. 2007;44(3):189–195.

6. Morland K, Wing S, Diez-Roux A,Poole C. Neighborhood characteristicsassociated with the location of food storesand food service places. Am J Prev Med.2002;22(1):23–29.

7. Nayga RM Jr, Weinberg Z. Super-market access in the inner cities. J Re-tailing Consum Serv. 1999;6(3):141–145.

8. Pothukuchi K. Attracting supermar-kets to inner-city neighborhoods: eco-nomic development outside the box. EconDev Q. 2005;19(3):232–244.

9. Kaufman PR. Rural poor have lessaccess to supermarkets, large grocerystores. Rural Dev Perspect. 1999;13(3):19–26.

10. Gittelsohn J, Franceschini MC,Rasooly IR, et al. Understanding thefood environment in a low-incomeurban setting: implications for food store

interventions. J Hunger Environ Nutr.2008;2(2,3):33–50.

11. Burt BM, Volel C, Finkel M. Safety ofvendor-prepared foods: evaluation of 10processing mobile food vendors in Man-hattan. Public Health Rep. 2003;118(5):470–476.

12. Lues JF, Rasephei MR, Venter P,Theron MM. Assessing food safety andassociated food handling practices instreet food vending. Int J Environ HealthRes. 2006;16(5):319–328.

13. Tester JM, Yen IH, Laraia B. Mobilefood vending and the after-school foodenvironment. Am J Prev Med. 2010;38(1):70–73.

14. Taylor D, Fishell V, Derstine J. Streetfoods in America - a true melting pot. In:Simopoulos A, Bhat R, eds. Street Foods:World Review of Nutrition and Dietetics.Basel, Switzerland: Karger; 2000:25–44.

15. Cupers K. Tactics of mobility: thespatial politics of street vending in LosAngeles. Paper presented at: Urbanism &Urbanization Conference: A New Mod-ernity—Approaches Theories, and De-signs; August 31, 2006; University Iuavof Venice, Venice, Italy. Available at:http://www.iuav.it/Didattica1/SCUOLA-DI-/DOTTORATI-/urbanistic/eventi/U-U_2006_papers_collection.pdf. AccessedJuly 28, 2010.

16. Hedley AA, Ogden CL, Johnson CL,Carroll MD, Curtin LR, Flegal KM. Prev-alence of overweight and obesity amongUS children, adolescents, and adults,1999-2002. JAMA. 2004;291(23):2847–2850.

17. Simopoulos A, Bhat R, eds. StreetFoods: World Review of Nutrition and Di-etetics. Basel, Switzerland: Karger; 2000.

18. Wright R. Hawkers and walkers inearly America. In: Ward D, Zunz O, eds.Landscape of Modernity. New York, NY:Russell Sage Foundation; 1992:233–234.

19. Bluestone D. The pushcart evil. In:Ward D, Zunz O, eds. Landscape ofModernity. New York, NY: Russell SageFoundation; 1992.

20. French E. Push cart markets in NewYork City (US Dept of Agriculture, Agri-cultural Economics Bureau and the Portof New York Authority, March 1925). In:Ward D, Zunz O, eds. Landscape ofModernity. New York, NY: Russell SageFoundation; 1992:34–35.

21. Annual estimates of the populationfor incorporated places over 100,000,ranked by July 1, 2007 population: April

1, 2000 to July 1, 2007 (SUB-EST2007-01). Washington, DC: Population Divi-sion, United States Census Bureau; July10, 2008.

22. American Legal Publishing Corpo-ration [search engine]. Available at:http://www.amlegal.com. Accessed Sep-tember 29, 2010.

23. Municipal Code Corporation [searchengine]. Available at: http://www.municode.com. Accessed September 29, 2010.

24. IL Admin Code tit 77, x750.1550(2009).

25. Chicago, IL Code x4-8-030(b) (2008).

26. Chicago, IL Code x4-5-010(31)(2008).

27. Chicago, IL Code x4-244-120(2008).

28. Chicago, IL Code xx 4-244-020, 4-5-010(66) (2008).

29. Dallas, TX Code x17-8.2(g)(1)(2008).

30. Dallas, TX Code x17-8.2(c)(1)(B)(2008).

31. Dallas, TX Code x17-8.2(h)(2)(F)(iv)(2008).

32. Dallas, TX Code x17-8.2(h)(2)(B)(2008).

33. City of Dallas. Requirements formobile food vendors. Available at: http://www.dallascityhall.com/pdf/ehs/MobileFoodVendorRequirements.pdf.Accessed July 28, 2010.

34. Houston, TX Code, art II, div 1, x20-22(e)(1),(4) (2008).

35. Houston, TX Code, art XI, div 2, x40-269(b) (2008).

36. City of Houston, Mobile FoodService Units, xVIII. Available at: http://www.houstontx.gov/health/Food/MOBILEREQ.html. Accessed July 28,2010.

37. Houston, TX Code, art II, div 2, x20-37 (2008).

38. Houston, TX Code, art XI, div 2, x40-263(3) (2008).

39. City of Houston, Mobile Food Ser-vice Units, xXIV. Available at: http://www.houstontx.gov/health/Food/MOBILEREQ.html. Accessed July 28,2010.

40. Houston, TX Code, art XI, div 2, x40-263(2) (2008).

41. City of Los Angeles, CommunityDevelopment Department. Sidewalkvending program. Available at: http://



www.lacity.org/cdd/bus_side.html.Accessed July 28, 2010.

42. New York, NY, tit 24, Health Codex89.05(a)(2) (2008).

43. New York, NY, tit 24, Health Codex89.5(a) (2008).

44. New York, NY, Code x17-307(e)(2008).

45. New York, NY, Code x17-08(c)(1)(2008).

46. New York, NY, Code x17-308(c)(2)(2008).

47. New York, NY, Code x17-307(b)(4)(2008).

48. New York City Dept of Health andMental Hygiene. Eat Street Smart. 2008.Available at: http://www.nyc.gov/html/doh/downloads/pdf/cdp/greencarts-brochure-online.pdf. Accessed July 27,2010.

49. New York, NY, Code x17-315(l)(2008).

50. New York, NY, Code x17-307(b)(4)(2008).

51. New York, NY, Code x17-307(b)(4)(e)(2008).

52. Philadelphia Dept of Public Health.Food safety for mobile food vendors:preparing and servicing safe food frommobile food vending units in Philadelphia.Available at: http://www.fsis.usda.gov/OPPDE/fslgrs/Retail/FY01/Philadelphia/VendorBrochure.pdf. Accessed July 28,2010.

53. Philadelphia, PA Health Code x6-301(3)(a) (2009).

54. Philadelphia, PA Health Code x6-301(8)(b) (2009).

55. Philadelphia, PA Code x9-203(3)(b)(2009).

56. Philadelphia, PA Code x9-205(8)(l)(2009).

57. Maricopa County Environmental.Health Code ch VIII, x3, reg (5)(e) (2007).

58. AZ Food Code x8-401.10 (2000).Available at: http://www.azdhs.gov/phs/oeh/rs/pdf/fc2000.pdf. Accessed July28, 2010.

59. Phoenix, AZ Code art XIV, x10-162(A)to (B) (2009).

60. Phoenix, AZ Code art XIV, x10-162(F) (2009).

61. Phoenix, AZ Code art II, x31-24(5)(2009).

62. Phoenix, AZ Code art II x31-24(1)(2009).

63. Phoenix, AZ Code art XIV, x10-166(B)(2) (2009).

64. Phoenix, AZ Code art II, x31-24.1(c)(2009).

65. Phoenix, AZ Code art XIV, x10-166(B)(3) (2009).

66. Phoenix, AZ Code art II, x31-24(2)(2009).

67. San Antonio, TX Code art IV, x13-64(2) (2009).

68. San Antonio, TX Code art IV, x13-62(j) (2009).

69. San Antonio, TX Code art IV, x13-62(d) (2009).

70. San Antonio, TX Code art IV, x3-63(a)(10) (2009).



73. San Diego, CA Code xx42.0130 &42.0161(m) (2009).

74. San Diego, CA Code x42.0103(2009).

75. San Diego County Code x8 65.104& 65.106(a)(7)-(9) (2009).

76. San Diego, CA Code xx42.0101.2(2009).

77. County of San Diego, Dept of Envi-ronmental Health. Construction and op-erational guide for mobile food facilitiesand mobile support units. Available at:http://www.sdcounty.ca.gov/deh/food/pdf/publications_plancheckmff.pdf.Accessed July 28, 2010.


79. San Diego, CA Code x54.0122(g)(2009).


81. CA Health & Safety Code x114295(West 2009).

82. CA Health & Safety Code xx113715,113725 (West 2009).

83. San Jose, CA Resolution 74981(2009).

84. San Jose, CA Code x6.54.270(2009).

85. San Jose, CA Code x6.54.240(1)(2009).

86. San Jose, CA Code x6.54.205(2009).

87. San Jose, CA Code x6.54.260(R)(2009).

88. San Jose, CA Code x6.54.240(2)(2009).

89. New York, NY Code x17-307(b)(4)(e) (2008).

90. Chicago, IL Code xx4-5-010(66)(2008).

91. Chicago, IL Code xx4-5-010(31) &(66) (2008).

92. City and County of San Franciscoand San Francisco Recreation and ParkCommission. Request for proposalsfor the operation of specialty food push-carts at various park locations citywide.Available at: http://sf-recpark.org/ftp/uploadedfiles/wcm_recpark/RFP/PushcartRFPFinal.pdf. Accessed July 27,2010.

93. Kansas City Parks and RecreationVending Policy. Available at: http://www.kcmo.org/idc/groups/parksandrec/documents/parksrecreation/012710.pdf. Accessed July 27, 2010.

94. CA Health & Safety Code x113713(West 2009).

95. TX Health & Safety Code Annxx437.002-437.0055 (Vernon 2009).

96. 25 TX Admin Code xx22.9.162(83)& 229.171(a)(1) (2010).

97. AZ Food Code x1-201.10(B)(69)(2000).

98. AZ Food Code x3-202.11 (2000).

99. CA Uniform Retail Food FacilityLaw x114265(h) (2010).

100. FDA Model Food Code, Preface x3(2005).

101. FDA Model Food Code x1-201.10(B) (2005).

102. NY Admin Code x17-306(a)(2008).


104. Houston, TX Code x20-22(f)(1)(2009).

105. New York, NY Code x17-307(b)(4)(e) (2008).

106. Jacobo F. Street vendors face longwait for permits. The Bronx Beat. April 19,2008. Available at: https://cranberry.cc.columbia.edu/cs/ContentServer?childpagename=Bronxbeat08%2FJRN_Content_C%2FRW1StoryDetailLayout2&c=JRN_Content_C&p=1175373931411&pagename=JRN%2FRW1Wrapper&cid=1175374630808&site=Bronxbeat08.Accessed July 27, 2010.

107. New York City Dept of Health andMental Hygiene. Mobile food vendor

permit waiting list instructions. Availableat: http://home2.nyc.gov/html/doh/html/permit/permit1.shtml. AccessedAugust 24, 2009.

108. Philadelphia, PA Code x9-204(8)(2009).

109. Houston, TX Code x40-263(3)(2008).

110. Chicago, IL Code x4-244-147(2008).

111. Los Angeles, CA Code x42(m)(2)(B)(2008).

112. Los Angeles, CA Code x42(m)(7)(2008).

113. Los Angeles, CA Code x42(m)(12)(E) (2008).

114. San Jose, CA Code x6.54.240(1)(2009).

115. San Diego Code x42.0166(c)(2009).

116. Nutritional Labeling: GeneralGuidelines. Crookston, MN: AgriculturalUtilization Research Institute; 1996.Available at: http://www.auri.org/research-article.php?raid=43. AccessedJuly 27, 2010.

117. 21 CFR 101.65(d)(2) (2008).

118. New York City, NY. Local Law 9,2008. Amendment to Municipal Codex17-306. Available at: http://webdocs.nyccouncil.info/textfiles/Int%200665-2007.htm. Accessed April 2, 2009.

119. New York City Dept of Healthand Mental Hygiene. NYC green cartsprogram. Available at: http://www.nyc.gov/html/doh/html/cdp/cdp_pan_green_carts.shtml. Accessed April 21,2009.

120. New York City Dept of Health andMental Hygiene. NYC green carts pro-gram areas. Available at: http://www.nyc.gov/html/doh/downloads/pdf/cdp/green_carts_areas.pdf. Accessed April 1,2009.

121. Collins G. Customers prove there’sa market for fresh produce. New YorkTimes. June 11, 2009;A:24.

122. Levi S. Green cart proposal takes onHarlem health. Columbia Spectator. No-vember 30, 2008. Available at: http://www.columbiaspectator.com/2008/02/01/green-cart-proposal-takes-harlem-health.Accessed July 27, 2010.

123. French SA. Pricing effects on foodchoices. J Nutr. 2003;133(3):841S–843S.



Public Health Professionals as Policy Entrepreneurs:Arkansas’s Childhood Obesity Policy ExperienceRebekah L. Craig, MPH, Holly C. Felix, PhD, MPA, Jada F. Walker, MA, and Martha M. Phillips, PhD, MPH, MBA

In response to a nationwide

rise in obesity, several states

have passed legislation to

improve school health envi-

ronments. Among these was

Arkansas’s Act 1220 of 2003,

the most comprehensive

school-based childhood obe-

sity legislation at that time.

We used the Multiple Streams

Framework to analyze factors

that brought childhood obesity

to the forefront of the Arkan-

sas legislative agenda and re-

sulted in the passage of Act

1220. When 3 streams (prob-

lem, policy, and political) are

combined, a policy window is

opened and policy entrepre-

neurs may advance their goals.

We documented factors that

produced a policy window and

allowed entrepreneurs to en-

act comprehensive legislation.

This historical analysis and

the Multiple Streams Frame-

work may serve as a roadmap

for leaders seeking to influ-

ence health policy. (Am J

Public Health. 2010;100:2047–

2052. doi:10.2105/AJPH.2009.

183939)

IN AN UNPRECEDENTED RISE,

the prevalence of overweightamong US children has morethan tripled over the past 3 de-cades. Recent National Health andNutrition Examination Surveydata (2003 through 2006) esti-mated that 32% of children andadolescents had a body mass

index (BMI; defined as weight inkilograms divided by height inmeters squared) for age at orabove the 85th percentile.1 WhenOgden et al. used the 97th percen-tile as an identifier of those with thegreatest body mass for age, theyreported that more than 11% of USchildren and adolescents fit intothat category.1 Overweight in child-hood is likely to persist into adult-hood2,3 and obesity predisposes fora number of diseases of both child-hood and adulthood.4 Adolescentswith very high BMI have also beenshown to have adult mortality ratesup to 40% higher than those ob-served in adolescents with mediumBMI.4

Obesity interventions and pre-vention have, consequently, be-come a major priority for policy-makers, health care professionals,economists, and the general pub-lic.5 Prior to 2003, several statesand the federal government hadenacted limited legislation aimed atreducing and preventing childhoodobesity.6 Incremental school-basedprevention efforts were largely fo-cused on emphasizing and improv-ing nutrition and physical educationcurricula, reinforcing classroomlearning throughout the school en-vironment, rewarding voluntaryadoption of healthy nutrition andphysical activity standards, andproviding model vending policiesand toolkits.6,7 Arkansas

policymakers recognized that halt-ing the epidemic necessitated pro-gressive steps to outpace increasingdisease rates. With the passage ofAct 1220 in 2003, Arkansasenacted comprehensive legislationto combat childhood obesity.

Act 1220 included 6 compo-nents aimed at combating child-hood obesity. First, a 15-memberChild Health Advisory Committeewas created and tasked with mak-ing recommendations to the StateBoard of Education and StateBoard of Health regarding physi-cal activity and nutrition standardsin public schools. Further, Act1220 required school districts toestablish Nutrition and PhysicalActivity Advisory Committees toguide the development of locallyspecific policies and programs.With Act 1220, Arkansas becamethe first state to enact statewideschool-based BMI screening withreports to parents for all publicschool children in grades Kthrough 12. Act 1220 both re-stricted student access to vendingmachines in public elementaryschools and required that schoolsdisclose vending contracts andpublicly report vending revenues.Lastly, the Arkansas Departmentof Health was required to employcommunity health promotion spe-cialists to provide technical assis-tance to schools in formulatingand implementing the rules and

regulations.8 Thus, Act 1220 man-dated some limited immediate ac-tion while establishing the mecha-nisms for short- and longer-termchange at both state and local levels.

In the policymaking process, in-cremental health policy change isthe norm, as opposed to innovative,comprehensive reforms such as Act1220. Legislators often face a mul-titude of issues, have little time toconsider all the data they need toaddress them, and may have tochoose from among a number ofpolicy alternatives to address anygiven issue. Zahariadis observedthat policymakers often ‘‘are lesscapable of choosing issues theywould like to solve and more con-cerned with addressing the multi-tude of problems thrust uponthem.’’9(p75) Arkansas is no excep-tion. In the 2003 Arkansas legis-lative session in particular, a largenumber of bills were introducedconcerning education. How, then,did the single issue of childhoodobesity rise to the forefront of anoverburdened legislative agenda?Who garnered political attentionfor this issue and formulated pol-icy solutions? What were the keyevents that led the Arkansas Leg-islature to abandon incrementallegislation and adopt a bold, com-prehensive policy initiative?

We sought to answer thesequestions by documenting thefactors and events that influenced

November 2010, Vol 100, No. 11 | American Journal of Public Health Craig et al. | Peer Reviewed | Government, Politics, and Law | 2047


the policy process allowing thepassage of Arkansas’s school-based childhood obesity legisla-tion policy. This historical reviewand analysis of the policy processprovides a roadmap for publichealth advocates interested inpursuing policies directed towardcurbing the childhood obesity ep-idemic and other critical publichealth issues. By studying thispolicy process roadmap, publichealth advocates interested inshaping health policy can under-stand more fully their role in theprocess of setting agendas andformulating policy and can moreeffectively act when windows ofopportunity arise.

METHODS

In our policy analysis, we usedsecondary data collected as part ofa comprehensive evaluation ofArkansas Act 1220.10 Specifically,key informant interviews wereconducted by a research team atthe University of Arkansas forMedical Sciences (UAMS) Fay W.Boozman College of Public Healthwith persons knowledgeable of orinvolved in the passage of Act1220. Interviews were conductedin a semistructured format with 3questions aimed at understandingthe key events, policy entrepre-neurs, and processes that led to thedevelopment of Act 1220: ‘‘Howdid Act 1220 get started?’’; ‘‘Whohad the initial idea, and how did itget from that idea to a piece oflegislation ready for introductioninto the legislature?’’; and ‘‘How doyou see the early processes of policydevelopment for Act 1220?’’ Theopen-ended nature of the questionsallowed respondents to relate the

process of policy formation ina narrative format. When answerswere abbreviated or nonspecific,probing questions were asked toclarify or to obtain greater detail.

We used John Kingdon’s Multi-ple Streams Framework to guidethe review of interview transcripts.11

Kingdon said of the policy process:

[T]he development of policyproposals is a little bit like bi-ological natural selection. . . .Ideas float around in a policyprimeval soup. Much like mole-cules . . . ideas start, combine,recombine, and through this longprocess of evolution, some ideasfall away, while others will sur-vive and prosper.12(p333)

In the Multiple Streams Frame-work, the policy stream representsthe ideas towhich Kingdon referred(i.e., the policy alternatives andpossible solutions to a problem).The political stream represents themood, ideology, or attitudes of pol-icymakers and the public. Theproblem stream represents themany issues that may require gov-ernmental action. These 3 streamsflow independently until a policywindow (or window of opportunity)

is presented. Such windows openwhen changes occur in the problemor political streams, perhaps be-cause of new problem indicators,focusing events, or changes in po-litical parties or ideology. Feasible,acceptable, affordable proposalsfrom the policy stream then emergethrough the policy window withthe help of a policy entrepreneur.Such a person will invest his or herown resources to advocate a partic-ular policy leading to its adoption(Figure1).11,13

A research team at UAMS FayW. Boozman College of PublicHealth received training in theKingdon framework (e.g., 3streams, policy entrepreneur, andpolicy window) and then read thekey informant interview tran-scripts for overall content. Theythen reviewed the transcriptsagain to identify significant factors(e.g., person, idea, event, or pro-cess) in the policy formation pro-cess and to extract potentiallyrelevant quotes. Each transcriptwas assigned to a single reviewer;reviewers consulted with oneanother during the process to

facilitate consistency amongthemselves.

Factors were mapped to com-ponents of the analysis framework(e.g., 3 streams, policy entrepreneur,and policy window), and all pas-sages coded to a specific compo-nent were merged into a singleframework component docu-ment. Secondary source docu-ments were used to confirm find-ings and to provide greater detailand context. A total of 23 infor-mants completed interviews, in-cluding policymakers such asArkansas legislators (n=8), gov-ernment-appointed advisors onhealth and education (n=11), andstate agency leaders (n=4).Twelve secondary source docu-ments salient to the Arkansasinitiative or addressing childhoodobesity through state educationpolicy were reviewed. Five ofthese documents were paperspublished in peer-reviewed jour-nals,14–18 3 were acts or resolutionsin the Arkansas Code,19–21 2 werearticles from the popular press,22,23

and 2 were state task force re-ports.24,25

FIGURE 1—Multiple Streams Framework.

2048 | Government, Politics, and Law | Peer Reviewed | Craig et al. American Journal of Public Health | November 2010, Vol 100, No. 11


RESULTS

As complex as the MultipleStreams Framework concept maybe, it can elegantly help to describethe fast-paced and muddled expe-rience of a policy design and in-stallation such as the Arkansas Act1220 to combat childhood obesity.

The Policy Stream

As an informant commented,‘‘the Act represented the culmina-tion of a longer developmentalprocess around the policy optionsavailable to the legislature.’’ In thepolicy stream we found that

a policymaker was motivated toaction by concern about theamount of caffeine and sugar infoods and beverages available tochildren in school vending ma-chines. Determined to uncoverpossible solutions, the respondentled legislative hearings to raiseawareness of the issue and gener-ate policy alternatives during the1999 and 2001 legislative ses-sions (Figure 2). Respondentswere also familiar with the use ofschools as a venue for child healthscreenings. Knowledge of Arkan-sas’s long-standing history of pro-viding school health

services—including screening forscoliosis, vision, and hearing—andreporting adverse outcomes toparents was evident among poli-cymakers.19,20 Additionally, someschools in Arkansas routinely mea-sured student height and weight aspart of health, physical education,and other curricula. This practiceis not uncommon among schoolsnationwide. For example, Floridapublic schools began collectingheight and weight for students in3 grades in 1973.14 Story et al.reported that, as of 2000, 26% ofstates had requirements thatschools measure students’ height

and weight and 61% of those statesrequired parental notification ofresults.18

Policymakers nationwide recog-nized that only1additional stepwould be necessary for schools toconvert those measurements intoreportable BMI surveillance data. Asearly as1995, California imple-mented collection of BMI measure-ments for public school students in 3grades.14 During the 2000 to 2001academic year, Cambridge PublicSchools and the Institute for Com-munity Health, both in Massachu-setts, conducted a pilot study ofschool-based BMI screening with

Note. ACH = Arkansas Children’s Hospital; ADH = Arkansas Department of Health; AR PNPA = Arkansas Preventive Nutrition and Physical Activity Summit; ASTHO = Association of State and Territorial

Health Officials; BMI = body mass index; NCSL = National Conference of State Legislatures; NFWL = National Foundation for Women Legislators; NGA = National Governors Association; UAMS

COPH = University of Arkansas for Medical Sciences Fay W. Boozman College of Public Health.

Source. Felix HC.13

FIGURE 2—Application of Multiple Streams Framework to the enactment process for Arkansas Act 1220.



parental notification of results.15 Inthe same year, the Tennessee legis-lature authorized optional BMI sur-veillance for all public school stu-dents.14 The Michigan Departmentof Education recommended BMIscreening in 2001and about half ofschool districts elected to screensome or all of their students.22,24

The Political Stream

Prior to the passage of Act1220, several significant actionsoccurred in the political streamto influence the attitudes ofArkansas’s policymakers. In1999, the Arkansas Legislaturecommissioned the Arkansas De-partment of Health to establishan Obesity Task Force to studythe effects of obesity on childrenand adults and to make recom-mendations for future state ac-tion to reduce obesity.21 Thetask force findings were reportedpublicly and to the legislature in2000. Regarding childhood obe-sity, the task force recommendedlegislation to enact a comprehen-sive statewide program with 14specific aims to raise public aware-ness and enhance school policiesand practices for nutrition andphysical activity.25

Soon after, Arkansas legislatorsattended the 2001 National Foun-dation for Women LegislatorsConference where public healthadvocates made quite an impact.Their tactics of raising awarenessof state-specific childhood obesityindicators influenced a respondentto support efforts to combat theepidemic. This respondent noted,

All across the whole wall wasplastered ‘Little Rock, Arkansas—number 1 in the nation for child-hood obesity and type 2 diabetes.’

That really woke me up and wasone reason that I spoke out.

Subsequently, legislators werepresented with a health resolutioncalling on them to take personalaction and serve as role models inthe state’s efforts to combat child-hood obesity.

Further support for state policyefforts to combat childhood obe-sity was garnered in January 2002when Arkansas legislators andother policymakers, includingrepresentatives from the Gover-nor’s Office and the Arkansas De-partment of Health, attendeda meeting sponsored by the Na-tional Conference of State Legis-latures, National Governors Asso-ciation, and Association of Stateand Territorial Health Officials. Atthe meeting, attendees from 6contiguous states considered dif-ferent approaches to addressinghealth issues in their states, in-cluding childhood obesity.23 Re-spondents noted interventions forchildhood obesity as a primarytopic of discussion within theArkansas delegation.

Public health professionals fromthe Arkansas Department ofHealth’s Cardiovascular HealthProgram and the UAMS hostedthe first Arkansas Preventive Nu-trition and Physical Activity Sum-mit in March 2002. Leaders whowere thought ‘‘most able to initiateand implement change’’ were in-vited to attend the 1-day confer-ence.16 Attendees were divided intowork groups, 1 of which was taskedwith devising practical, achievablepolicy alternatives for education,including school environment. Arespondent from the health com-munity summarized the Summit’s

impact: ‘‘The [Arkansas PreventiveNutrition and Physical Activity Sum-mit] set the framework for thinkingabout what the problems are, thescope of the problem, and possibleinterventions.’’ The Summit’s work-ing recommendations includedschool-based BMI surveillance withparental notification for all publicschool students and creation of anoffice devoted to nutrition andphysical activity.16

The Problem Stream

Throughout that time, publichealth leaders from the UAMS FayW. Boozman College of PublicHealth and the Arkansas Depart-ment of Health presented annualupdates to legislators about theburden of obesity in Arkansas.Multiple policymakers interviewedrecounted health information theylearned during those updates. Evi-dence of obesity’s consequences forthe state made a strong impressionon legislators. One respondent,whose remark is representative ofseveral others, recalled learningthat because of earlier onset ofobesity and diabetes, ‘‘40-year-oldpeople are getting their feet andlegs cut off.’’ The informationhelped focus the attention of poli-cymakers on the obesity issue.

Annual updates to the ArkansasLegislature often included indica-tors of the severity of childhoodobesity. By 2002, for example, anestimated 31% of American chil-dren aged 6 to 19 years wereoverweight or obese, and physi-cians at the Arkansas Children’sHospital began discussing thesharp rise in the number of casesof child and adolescent onsetof type 2 diabetes seen in theirclinics.17 A fitness clinic was

planned at Arkansas Children’sHospital to provide behavioral andsurgical weight-loss interventionsfor children with a BMI measure-ment greater than the 95th per-centile.

Prior to the Arkansas 2003 leg-islative session, then-Speaker of theHouse Herschel Cleveland, a Demo-crat, and then-Arkansas GovernorMike Huckabee, a Republican, eachexperienced serious obesity-relatedpersonal health problems. Thoseexperiences, made public becauseof their offices, served as focusingevents that brought attention to theseriousness of the issue and made itclear that the battle against obesitywas bipartisan.

The Policy Window and Policy

Entrepreneurs

Because of these changes inthe political and problemstreams, a temporary policy win-dow opened, providing the op-portunity for comprehensivepolicy changes to combat child-hood obesity. After summarizingsome events in the 3 streams,a respondent aptly described thisphenomenon:

It just happened to be that thelegislators were interested inhearing about ways to improvechild health and were willing tostick their neck out with a bill thatwas really different from any-thing that had been done in therest of the United States.

Speaker Cleveland was broadlycredited by respondents as the pri-mary policy entrepreneur. Onepolicymaker remarked, ‘‘I think[Speaker Cleveland] had a personalexperience, professional interest,and a legislative responsibility thatcame together.’’ Speaker Cleveland

2050 | Government, Politics, and Law | Peer Reviewed | Craig et al. American Journal of Public Health | November 2010, Vol 100, No. 11


requested that the Arkansas De-partment of Health draft potentiallegislation for school-based policychanges to reduce childhood obesityand then invested himself person-ally in advocating the bill’s passage.

Public health professionals actedas secondary policy entrepreneursby coupling the problem with via-ble alternatives from the policystream. Several of those individualshad been involved in the eventsnoted in the political and problemstreams and were ready with prac-tical, achievable policy optionsgenerated at the Arkansas Preven-tive Nutrition and Physical ActivitySummit and other venues. Thetimely coupling of streams by pol-icy entrepreneurs led to the passageof Arkansas Act 1220 of 2003.

DISCUSSION

In the political stream, advocacyby public health professionals atnational legislative conferencesand state-level meetings influ-enced policymakers’ beliefs aboutchildhood obesity. In the problemstream, focusing events, namelythe personal health problemsof 2 policymakers, paired withthe changes in indicators pre-sented at local and national meet-ings, turned attention and focusto the issue. Arkansas Speaker ofthe House Herschel Clevelandand Arkansas’s public health pro-fessionals were most frequentlynoted by key informants to bethe primary policy entrepreneurs.Public health advocates whoparticipated in the ArkansasObesity Task Force and theArkansas Preventive Nutritionand Physical Activity Summitgenerated policy alternatives and

formulated the legislation, whichwas then sponsored by colleaguesof Speaker Cleveland, at his urg-ing.

Kingdon’s Multiple StreamsFramework continues to be auseful model for understandingmany cases of health policy re-form, particularly comprehensivereforms such as Arkansas Act1220. It is notable that during thepolicy process in Arkansas, somepublic health leaders advocatedan incremental approach toaddressing school-based obesitypolicy. Several respondentsremarked about the many ver-sions drafted before the bill wasfiled. Key elements that werefeared to diminish political feasi-bility, such as vending restrictionsand the BMI initiative, were de-bated, eliminated, and then addedback into the proposal as keypublic health professionals tire-lessly advocated comprehensivelegislation.

Kingdon asserted that althoughgeneration of policy alternativesmay be incremental, as was thecase for Act 1220, agenda changeis nonincremental and occurswhen a combination of the 3streams opens a policy window.11

During that short window of time‘‘there is often sufficient ambiguityin the nature of the problem orwhat can be done about it so thata leader can offer his or her pro-posal as a plausible solution.’’26(p216)

Speaker Cleveland was a policyentrepreneur; thus, his experience,interests, and responsibility to thestate of Arkansas poised him toadvocate policy change when hewas presented with a window ofopportunity. Public health pro-fessionals, armed with policy

alternatives, found that legislatorswere willing to take bold stepstoward eliminating childhoodobesity in Arkansas. The processwas described, even by a pro-ponent of incrementalism, as ‘‘awonderful progression of compro-mise, of discussion, of verythoughtful people being passionateabout it and it’s turned out to be anoutstanding piece of legislation.’’

The Arkansas law known as Act1220 of 2003 provides an illus-trative example of comprehensivepublic health policy on a state level.When Act 1220 is viewed throughthe lens of the Multiple StreamsFramework, the influence of publichealth professionals is clearly seenin raising awareness and proac-tively generating policy alternatives.The Multiple Streams Frameworkincorporates the important role ofchance in the policymaking process.Policy windows are short andoften unpredictable. Whether atthe federal, state, local, or agencylevel, public health professionalsmust understand their policy en-vironment and not lose a momentin recognizing the convergence ofthe 3 streams and ‘‘champion’’policy entrepreneurs. With thecorrect balance of strategicplanning and timely responses topolicy windows, public healthprofessionals can use Kingdon’sMultiple Streams Framework asa roadmap for improving thehealth and well-being of thepopulation. j

About the AuthorsAt the time of the study, all authors werewith the Fay W. Boozman College of PublicHealth, University of Arkansas for MedicalSciences, Little Rock. Martha M. Phillipswas also with the Department of Psychiatry,

College of Medicine, University of Arkansasfor Medical Sciences.

Correspondence should be sent to JadaWalker, MA, University of Arkansas forMedical Sciences, Fay W. Boozman Collegeof Public Health, 4301 W Markham St,Slot 863, Little Rock, AR, 72205 (e-mail:[email protected]). Reprints can be or-dered at http://www.ajph.org by clicking the‘‘Reprints/Eprints’’ link.

This article was accepted on April 5,2010.

ContributorsR. L. Craig synthesized concepts, inte-grated research findings, and led thewriting. H. C. Felix assisted with thepolicy theory specifics and analyses. J. F.Walker assisted with the research andenvisioned the policy implications. M. M.Phillips originated the research and su-pervised all aspects of its implementa-tion. All authors helped to conceptualizeideas, interpret findings, and reviewdrafts of the article.

AcknowledgmentsThis work was supported by the RobertWood Johnson Foundation (grants051737, 60284, and 30930).

The authors thank James M. Raczynski,PhD, for his support, as well as BritniMitchell and Matilda Louvring for assis-tance in preparation of this article.

Human Participant ProtectionThis study was approved by the Univer-sity of Arkansas for Medical Sciencesinstitutional review board.

References1. Ogden CL, Carroll MD, Flegal KM.High body mass index for age among USchildren and adolescents, 2003-2006.JAMA. 2008;299(20):2401–2405.

2. Whitaker RC, Wright JA, Pepe MS,Seidel KD, Dietz WH. Predicting obesityin young adulthood from childhood andparental obesity. N Engl J Med. 1997;337(13):869–873.

3. Engeland A, Bjorge T, Tverdal A,Sogaard AJ. Obesity in adolescence andadulthood and the risk of adult mortality.Epidemiology. 2004;15(1):79–85.

4. Regan F, Betts P. A brief review of thehealth consequences of childhood obe-sity. In: Cameron N, Norgan NG, EllisonGTH, eds. Childhood Obesity: Contempo-rary Issues. Boca Raton, FL: CRC Press;2006:26–38.



5. Paxson C, Donahue E, Orleans T,Grisson JA. The Future of Children.2006;16:9–10. Available at: http://www.aecf.org/upload/publicationfiles/obesity.pdf. Accessed July 27, 2010.

6. Rosenthal J, Chang D. State ap-proaches to childhood obesity: a snapshotof promising practices and lessonslearned. Portland, ME: National Academyfor State Health Policy; 2004. Availableat: http://nashp.org/sites/default/files/childhood_obesity.pdf. Accessed July 27,2010.

7. National Conference of State Legisla-tures. Childhood obesity—An overview ofpolicy options in 2003–2004. Availableat: http://www.ncsl.org/default.aspx?tabid-14112. Accessed August 29, 2009.

8. Arkansas Act 1220 of 2003 toCombat Childhood Obesity. ArkansasCode Annotated, 20-7-133-135 (2003).

9. Zahariadis N. Ambiguity, time, and mul-tiple streams. In: Sabatier P, ed. Theories ofthe Policy Process. 2nd ed. Cambridge, MA:Westview Press; 2007:73–93.

10. Raczynski J, Phillips M, Bursac Z, et al.Establishing a baseline to evaluate Act1220 of 2003: an act of the ArkansasGeneral Assembly to combat childhoodobesity. Little Rock, AR: University ofArkansas for Medical Sciences College ofPublic Health; 2005:1–32.

11. Kingdon JW. Agendas, Alternatives,and Public Policies. 2nd ed. New York, NY:Addison-Wesley Educational PublishersInc; 1995.

12. Kingdon JW. A model of agenda-setting with applications. Mich St DCL L.2001:331–339.

13. Felix HC. The rise of the community-based participatory research initiative atthe National Institute for EnvironmentalHealth Sciences: an historical analysisusing the policy streams model. ProgCommunity Health Partnership. 2007;1(1):31–39.

14. Nihiser AJ, Lee SM, Wechsler H, et al.Body mass index measurement in schools.J Sch Health. 2007;77(10):651–671.

15. Chomitz VR, Collins J, Kim J, KramerE, McGowan R. Promoting healthy weight

among elementary school children viaa health report card approach. ArchPediatr Adolesc Med. 2003;157(8):765–772.

16. Wheeler JG, Kahn R, Garner C,Vannatta P. Obesity in Arkansas: fromcontemplation to action. The 2002Arkansas Preventive Nutrition andPhysical Activity Summit. J Ark Med Soc.2004;100(8):268–272.

17. Hedley AA, Ogden CL, Johnson CL,Carroll MD, Curtin LR, Flegal KM. Prev-alence of overweight and obesity amongUS children, adolescents and adults,1999–2002. JAMA. 2004;291(23):2847–2850.

18. Story M, Kaphingst KM, French S.The role of schools in obesity prevention.Future Child. 2006;16(1):109–142.

19. Arkansas Code Annotated, 6-18-701(1987).

20. Arkansas Code Annotated, 20-15-802 (1987).

21. 82nd General Assembly, RegularSession, SCR 8 (1999).

22. Upton J. Health reform effort enlistsMichigan children. The Detroit News.March 23, 2004.

23. State Legislatures Standing Commit-tee on Health. Identifying health goals.Report from: National Conference of StateLegislatures; July 2003. Available at:http://www.ncsl.org/portals/1/documents/health/healthcostsrpt.pdf.Accessed July 27, 2010.

24. Michigan Department of Education.The role of Michigan schools in promot-ing healthy weight: a consensus paper.September 2001. Available at: http://www.michigan.gov/documents/healthyweight_13649_7.pdf. Accessed June 28,2008.

25. Obesity Task Force. The impact ofobesity: economics, health, prevention &treatment. 2000. Available at: http://www.healthyarkansas.com/newsletters/obesity_report.pdf. Accessed June 28,2008.

26. Oliver TR. The politics of publichealth policy. Annu Rev Public Health.2006;27:195–233.

Utilization of Research in Policymaking for Graduated Driver LicensingReece Hinchcliff, PhD, Rebecca Q. Ivers, PhD, MPH, Roslyn Poulos, PhD, and Teresa Senserrick, PhD

Young drivers are overrep-

resented in road trauma and

vehicle-related deaths, and

there is substantial evidence

for the effectiveness of gradu-

ated driver licensing (GDL)

policies that minimize young

drivers’ exposure to high-risk

driving situations. However, it

is unclear what role research

plays in the process of making

GDL policies.

To understand how research

is utilized in this context, we

interviewed influential GDL

policy actors in Australia and

the United States. We found

that GDL policy actors gener-

ally believed that research

evidence informed GDL policy

development, but they also be-

lieved that research was used

to justify politically determined

policy positions that were not

based on evidence.

Further efforts, including

more effective research dis-

semination strategies, are re-

quired to increase research

utilization in policy. (Am J Pub-

lic Health. 2010;100:2052–

2058. doi:10.2105/AJPH.2009.

184713)

YOUNG DRIVERS (AGED 17-25

years) are overrepresented in roadtrauma, and vehicle-related

crashes are a leading cause of

death among young people.1,2

Governments in many high-income

countries, including Australia and

the United States, have addressed

this problem by developing gradu-

ated driver licensing (GDL) sys-

tems.1 GDL systems minimize

young drivers’ exposure to high-risk

driving situations and may use any

of a variety of policies, such as

minimum age of licensing and speed

limitations. Research has shown that

such systems can be very effective

in reducing crashes and injuries,

although their effectiveness depends

on the inclusion of several key fac-tors.3

Restrictions on night drivingand on the ages of passengers areamong the most effective ways toreduce crash involvement.4 How-ever, policymakers in many statesand jurisdictions have opposedthese restrictions for a number ofpolitical (e.g., electoral support) andideological reasons, and because ofconcerns regarding the legitimacyof using evaluations from otherjurisdictions to determine appropri-ate policies.5 Such widespread gov-ernmental opposition to these re-strictions indicates that, despite the

2052 | Government, Politics, and Law | Peer Reviewed | Hinchcliff et al. American Journal of Public Health | November 2010, Vol 100, No. 11


prominence of evidence-based ar-guments in GDL policy discourse,6

the creation of policy is mediated bya variety of other factors that havenothing to do with evidence.

An approach to policymakingthat utilizes technical rationalityand is based on evidence wouldstrive to make effective use ofscientific research, the better tomaximize the societal benefitsresulting from policy implementa-tion.7 However, policies in variouspublic sectors throughout the worldare infrequently based on researchevidence.7–10 This discrepancy hasbeen identified as a serious issuedemanding urgent attention.11 Suchdisparity between the rhetoric andthe reality of evidence-based policyhas generated a body of literatureaiming to increase the transparencyof policymaking processes by iden-tifying factors preventing12 andfacilitating13 research utilization.These findings have been pooledinto a number of significant re-views,7,14,15 and on the basis ofthese reviews several frameworkshave been developed to explainresearch utilization in policymak-ing.16–19 The diversity of theseframeworks demonstrates the diffi-culty of understanding this complexphenomenon and articulating ex-planations that may be applicableto different policy contexts.

To our knowledge, no studyto date has examined researchutilization within the context ofroad-safety policy or novice-driverpolicy, despite such a study’s po-tential to identify critical points ofresistance to evidence-informedpolicies and to reveal strategies toencourage their adoption. Weaimed to fill this knowledge gap byseeking out individuals involved

in GDL policy and asking themtheir opinions regarding researchutilization in GDL policymaking.

METHODS

From 2004 through 2008, theAustralian states of New SouthWales, Victoria, Western Aus-tralia, and Queensland engaged inprotracted policymaking processesinvolving night driving and pas-senger restrictions as parts of GDLpolicies. Yet despite considerableadvocacy by various stakeholders,including researchers,6 none of thepolicies introduced in these statesreflected the best practices identi-fied in the GDL literature.1,20 Thus,these policymaking processes offerideal real-life examples of the bar-riers to and facilitators of researchutilization in GDL policy.

Our study was primarily situ-ated in the Australian context, butGDL policy debates involvingnight driving and passenger re-strictions have received a similarly

high level of attention in theUnited States. Therefore, the in-clusion of a US state in the studyoffered the opportunity to assessthe relevance of the results outsidethe Australian context and identifykey themes relevant in both na-tional contexts. We decided toinclude Pennsylvania becausepassenger restrictions receivedsignificant attention in that stateduring the time when we wereconsidering the inclusion of a USstate. Our Pennsylvanian partici-pants identified the important roleof federal advocacy and researchgroups in state-based decision-making, so we conducted severalinterviews in Washington, DC, togain further insights.

To learn more about these poli-cymaking processes, we interviewedpeople who had participated in theprocesses. We developed our poolof potential interviewees by firstlisting the key groups and indi-viduals involved in the novicedriver policy arena. We created

this list on the basis of an analysisof transcripts of parliamentarydebates and committee hearings,newspaper articles, and otheronline information identifying theroad safety policymaking struc-tures in each state. The analysispermitted us to identify as poten-tial interviewees 8 to 10 key in-dividuals in each state from amongthe following groups: politicians,senior public servants (includingthose from jurisdictional licensingauthorities and other relevantgovernment agencies), insuranceand motoring organizations, re-searchers, media, police, roadsafety, victim rights advocates, andyouth rights advocates. Each ofthese groups and individuals haddirect impact on formal GDL pol-icymaking processes or on rele-vant community debate that couldinfluence policy decisions.

We invited our potential inter-viewees to participate in this study,and we provided them with studyinformation, including assurances

TABLE 1—Barriers to and Facilitators of Research Utilization in Young Driver Policy, by Influential Policy

Actors: Selected States in Australia and the United States, 2007–2009

Barriers Facilitators

Influence of political factors (e.g., electoral support for policy

alternatives) and ideological factors (e.g., protection of civil

liberties) on policymaker reception of research

Evidence-based policymaking culture;

high value placed on research

Researcher awareness of existing political constraints

Opportunities for ongoing research input into

formal policymaking structures

Bureaucratic filtering of research information Policy networks, conferences, and research studies that

produce personal relationships among long-term policy actors

Poor-quality evaluation research High-quality evaluation research

Politically determined, small-scale research funding Researcher-determined, large-scale, policy-relevant research funding

Increasing prioritization of peer-reviewed publications as the

sole research output

Researcher and research institute emphasis on policy impact

Media reporting not based on evidence Effective media advocacy for evidence-based alternatives

Inability of researchers to effectively disseminate research Effective research dissemination techniques

November 2010, Vol 100, No. 11 | American Journal of Public Health Hinchcliff et al. | Peer Reviewed | Government, Politics, and Law | 2053


of confidentiality. Eighty-four per-cent of our invitees agreed to par-ticipate. We conducted a 1-hoursemistructured interview with eachof our 48 participants. The inter-views took place from August2007 through December 2009.

The interview guide was adap-ted from the RAPID Context, Ev-idence and Links framework foranalysis.21 The main interviewtopics were: policymaker demandfor research, opportunities for re-search input into policymakingstructures, types and quality ofresearch evidence, relative effec-tiveness of different research dis-semination strategies, barriers toand facilitators of research utiliza-tion, and strategies to increase re-search utilization in policymaking.Interviews were transcribed andthen categorized using NVivo ver-sion 7 textual reference software(QSR International, Doncaster, Vic-toria, Australia), allowing us toidentify the key themes emergingfrom interviewee responses. Rele-vant quotes were used to exemplifycritical emergent study themesfound to be applicable in bothAustralian and US policymakingcontexts.

RESULTS

Despite the differences amonginterviewees, their responsesexhibited minimal variation, andthe most critical emergent inter-view themes remained constant.

Structures, Actors, and

Networks Influencing Policy

Although interviewees namedseveral federal agencies that influ-enced GDL policy in Australia andthe United States, all interviewees

said that policy decision-makingwas largely controlled by politi-cians and senior public servantsfrom the lead road safety agency ofeach state. The authority of theminister responsible for road safetyin each Australian state was par-ticularly emphasized, and Ameri-can interviewees identified the im-portance of state legislators actingas the main sponsors of bills regu-lating novice drivers.

Interviewees identified severaladditional important decision-making bodies, such as road safetycouncils, that were composed ofrepresentatives from various gov-ernment agencies (e.g., health, ed-ucation) and that sometimes in-cluded influential interest groups(e.g., motoring organizations) andresearch experts. Intervieweessuggested that these interagencybodies provided opportunities forresearch information to be circu-lated among disparate policy ac-tors. Road safety conferences andinfluential research and policynetworks unrelated to formal pol-icymaking structures were alsoidentified as ways to link membersof the wider policy community,including researchers and policy-makers from different statesand countries. All intervieweesfrom influential policy groups sug-gested that these networks helpedensure that the relatively smallcommunity involved in craftingpolicy for young drivers wouldremain open to new ideas.

Politicians and chief executiveofficers of key state governmentagencies were described as ‘‘regu-larly fluctuating,’’ whereas middle-level managers, professionaladvocates, and research expertswere described as being ‘‘here

for the long haul,’’ with theirexpertise significantly utilized indrafting policy. This long-terminvolvement was identified asa facilitator of the developmentof personal relationships amongkey policy actors, including publicservants and researchers. Inter-viewees said these informalchannels provided a way for re-search and other types of infor-mation to freely circulate.

Demand for Research

All interviewees suggested thatGDL policy is generally evidence-based, and they supported thisclaim by citing the relationshipbetween increasingly positiveevaluations in the scientific litera-ture and its diffusion throughoutAustralia, New Zealand, NorthAmerica, and much of Europeover the past decade.1 They saidthey believed that this evidence-based policymaking culture hadfostered significant demand for re-search among individuals andgroups seeking to affect GDL policy,such as policymakers, motoring or-ganizations, advocacy groups, andsome journalists.

Nonetheless, participants fromall included groups of influentialpolicy actors claimed that researchevidence did not in itself deter-mine policy because other politicalfactors (e.g., electoral support ofpolicy alternatives) and ideologicalfactors (e.g., civil liberties) influ-enced decisions. In support of thisclaim, they cited the fact that eithernight-time restrictions or passengerrestrictions—but not both—hadbeen introduced in each state re-ferred to in the study: New SouthWales, Victoria, Queensland, West-ern Australia, and Pennsylvania.

Uses, Quality, and Funding of

Data and Research on Novice

Drivers

Several types of data were de-scribed as influencing GDL policy,including information used forproblem identification and agendasetting,22 such as crash, enforce-ment, and hospitalization data.Although public servants and in-terviewees from other influentialgroups generally argued that policywas not ‘‘driven by crook statistics,’’all interviewees said that suchdata were frequently manipulatedby policy actors emphasizingspecific features of the findings(e.g., brief spikes in road traumadespite long-term downwardtrends) to produce ‘‘startlingstatistics’’23(p161) designed to initiateor block policy reform.

Research that assessed commu-nity attitudes (e.g., public consulta-tion) was identified as influentialbecause of its utility in assessingpolicy alternatives from a politicalperspective. As a professional roadsafety advocate explained, ‘‘There’snothing a minister likes more thanwhen you say, ‘Here’s the policy. Bythe way, everyone loves it.’’’

Most importantly, all inter-viewees represented evaluationresearch as a key battleground inpolicy debates. Interviewees par-ticularly noted the difficulty ofevaluating the effectiveness ofspecific GDL components whenintroduced as elements of largerpackages: ‘‘You can’t provide exactestimations of, ‘If you introducethis policy, it [the road toll] willdrop by this much,’’’ an Americanadvocate complained. Such issueswere mentioned by all inter-viewees as evidence of the gener-ally poor quality of GDL



evaluation research, with an Aus-tralian researcher explaining, ‘‘Thequality of the work is not particu-larly good. . . . We need better-quality studies because many areso flawed you can’t make reason-able conclusions about whethersomething works.’’ Previous stud-ies have identified GDL researchquality issues as an obstacle tometa-analyses assessing the pre-cise level of effectiveness of spe-cific GDL models.1,24

All nongovernmental inter-viewees said that GDL researchmay be produced for prede-termined political purposes.Government-funded research wasoften described as a political strat-egy aiming to reduce communitydemand for policy reform byrepresenting government as pro-actively engaging in an evidence-based policy reform process. Asa professional advocate suggested,‘‘Research is often commissioned. . . to either procrastinate or delaydecisions. You’ve seen, ‘Yes, Minis-ter . . . yeah, we’re doing some-thing. We’re looking at it. Let’s geta report.’’’ This issue representedthe clearest division between theresponses of interviewees fromdifferent groups of policy actors,with government representativesvigorously denying this contention.

Nongovernmental intervieweesexplained that this politicization ofresearch impairs research qualitybecause it results in the issuance ofshort-term ‘‘Mickey Mouse softmoney’’ government contracts toinvestigate the politically deter-mined priority issues of policy-makers. Interviewees said thecurrent dominance of this fundingmodel in Australia and the UnitedStates was a major barrier to

research utilization because itprevented the development oflarger research studies with broadpolicy implications, and it directedresearcher attention toward gain-ing new funding contracts ratherthan increasing the quality of re-search outputs (Table 1) . A re-searcher in Victoria put it this way:

The government decides what itwants to purchase. . . . The trou-ble is, it’s not so much that they’recommissioning research butcommissioning consultancies. . . .Once the government has deter-mined the issue, we go out and doinvestigations, but the funda-mental research—about what theoptions are that could be consid-ered in the first place—perhapsisn’t being done. . . . There isn’t asmuch input from the researchersas there might be.

Research Dissemination

Interviewees said the methodby which research was dissemi-nated largely determined its use inpolicymaking. Several researchersand public servants argued that‘‘the currency of road-safety in-formation has changed,’’ in thatresearch is increasingly publishedin peer-reviewed journals gener-ally accessible only to otherresearchers instead of being dis-seminated through mediums moreaccessible to policymakers, suchas conferences and governmentreports. Interviewees said thischange had occurred because ofthe growth of large road-safetyresearch institutes, whose focus ismore academically oriented thanpolicy-oriented.

Researchers said the move to-ward peer-reviewed publicationswas slowly improving the qualityof GDL research, but they andinterviewees from other influentialgroups of policy actors argued that

this shift was nonetheless prob-lematic, because policymakersmay not have the time or expertiseto monitor these sources. All in-terviewees said that the extent ofindividual researchers’ dissemina-tion efforts was largely determinedby the degree to which policyimpact formed part of a researchinstitute’s organizational vision.

The presentation of researchemerged as critical. A public ser-vant in Pennsylvania describedthe main goal as ‘‘being receiver-oriented in any communication.So what do they want to know?’’The provision of clear and suc-cinct policy-relevant advice wasparticularly emphasized by pro-fessional advocates, public ser-vants, and motoring organizationrepresentatives, with a profes-sional road safety advocate argu-ing that when communicatingresearch to policymakers, it ismost important ‘‘to get from 120pages of data down to 1 page of 5simple bullet points that explain, ‘Ifwe do this, here’s the problem,here’s the remedy, here’s the likelybenefit.’’’ However, intervieweeswho were not researchers gener-ally described researchers asunwilling to produce such sum-maries. As an Australian profes-sional road safety advocateexplained, ‘‘They get defensive . . .

[because] it’s hard when you’velived and breathed the report toget it down to three main points.’’

Government representativesfurther suggested that researchwas ‘‘the base level in a pyramid ofdecision-making.’’ As a legislatorcommented, ‘‘While overseasevidence is important, you needmore than research on yourside if you’re going to bring the

community and politicians withyou.’’ Therefore, intervieweesclaimed that in addition to clearlylaying out the research, effectivedissemination required attentionto broader factors of political in-terest, apart from the road safetyliterature. A public servantexplained that such factors mayinclude ‘‘whether it’s workable,whether you can bring somethingin that sounds good, that’s practi-cal enough to be administered andsupported by the community.’’

Barriers to Research

Utilization

Interviewees identified severalcritical barriers to research utiliza-tion (Table 1). In their assessmentsof formal policymaking structures,motoring organization representa-tives, various professional advo-cates, and researchers (includingsome with previous bureaucraticexperience) represented publicservants as information filterswho only provided superiors withresearch supporting politicallyfeasible policy alternatives, such aspolicies that enjoyed significantcommunity support. As an Aus-tralian researcher with bureau-cratic experience explained,

There is a tendency for the bu-reaucratic policymakers to applya perspective based on their in-terpretation of what their minis-ter or even their boss within thedepartment will think. . . . Weunfortunately have politicizedthe public service.

All interviewees said that dis-seminating research to the com-munity via the media was just asimportant as dissemination topolicymakers. As a public servantexplained:



The political process is influ-enced by what politicians thinkthe electorate wants, and that’sinfluenced strongly by what themedia says. . . . So the media doestend to marshal arguments whichhave an influence on the way thatthings are discussed by the publicand the way that they’re dis-cussed by politicians.

All interviewees said the mediasometimes assisted advocates ofevidence-based policy reform bystimulating and reinvigoratingcommunity interest and debateregarding young drivers. How-ever, they also said that mediareporting may largely ignore re-search that may be relevant topolicy debates if that research isperceived as having limited appealto audiences. Interviewees whosupported evidence-based GDLreform suggested that thesemedia-related issues were exem-plified by the policy processessurrounding night-time and pas-senger restrictions in each state:media reporting has highlightedthe unacceptable level of roadtrauma involving novice driversand the need for government ac-tion, yet media outlets havegranted less attention to researchevidence supporting competingpolicy alternatives.6A researchercited this generally ill-informedstyle of media reporting as a reasonwhy researchers needed to

put forward a broader base ofinformation . . . [because] in theabsence of such activities, youdon’t have a researcher or re-search evidence being brought tothe table [in policy debates].

Motoring organization representa-tives, journalists, professional ad-vocates, and several researcherssuggested that using emotive im-ages (e.g., crashed cars, mass

funerals) and linking these to‘‘startling scientific statistics’’23 andnonscientific but policy-relevanttypes of information (e.g., victims’tragic anecdotes) facilitated thecommunication of research to a layaudience via the media. Thesemedia advocacy strategies wereseen as encouraging more evi-dence-informed GDL policy de-bates and increasing communitysupport for evidence-informedpolicy alternatives.25

DISCUSSION

Research-utilization literaturemost commonly involves inter-views with policymakers and re-searchers from single states orjurisdictions.26 To prevent the re-sults of this study from being overlyreliant on the interpretations of anysingle group of policy actors, wecreated a much more varied sample,drawing participants from differentstates and countries. This allowed usto conduct a more holistic analysisby triangulating the interviewdata.27 Our strategy also enabled usto discover differences of opinionbetween government representa-tives and other influential groups ofpolicy actors regarding the existenceof bureaucratic filtering of policy-relevant information and ideal re-search funding models in both theAustralian and US contexts. None-theless, other major relevantthemes remained similar across allincluded groups of policy actors,indicating both their broad accep-tance and their likely relevance inall GDL policy settings in bothAustralia and the United States.

Studies situated in other policysectors have found minimal

research utilization in policy,7,28–

30 but our results show that in-dividuals involved in GDL policydebates and processes believed thatpolicy decisions were generally‘‘evidence-informed.’’31(p20) How-ever, this result may partially reflectthe qualitative nature of our as-sessment of research utilization; in-terviewees may have believed itmore appropriate to indicate thatthey used, rather than ignored,research in their policymakingactivities.

Although interviewees said re-search was predominantly usedinstrumentally to determine thecauses and potential solutions forthe problems of young drivers,participants from all includedgroups also argued that policyactors may use research tacticallyto justify politically determinedpolicy positions.17 As with otherpolicy areas, research evidence andexpertise constituted valuable cur-rency in modern GDL policy dis-course, with the scientific commu-nity, motoring organizations, andrelevant government departmentsacting as the major suppliers of thenecessary ‘‘intellectual ammuni-tion.’’32(p35)

A critical issue within debatesover the policy implications ofresearch evidence is whetherappropriate research utilizationinvolves a high level of fidelity(accurate replication of interven-tions from one context to an-other) as opposed to some levelof adaptation being required foreffective outcomes in differentcontexts.9 As it is impracticable forhigh-quality research evaluations ofthe effectiveness of complex inter-ventions to exist for every setting,reasonable translation and

interpretation of research findingsfrom one setting to another is re-quired to produce effective policies.This issue has particular relevancein assessing whether the versions ofnight-time or passenger restrictionsintroduced in the states understudy represent effective researchutilization, given their divergencefrom best-practice models becauseof practicality and political con-cerns.20

Policy Affected by Many

Factors, Including Evidence

Such policy deviations from re-search evidence indicate that de-spite evidence-based rhetoric,GDL policy is affected by a rangeof other factors. As a senior publicservant commented, ‘‘It boilsdown to what the research evi-dence is telling you, what thepracticality of your system allowsyou to do, and what we interpretfrom the community as being ini-tiatives they think are workable.’’Therefore, research representsonly a single piece of the GDL‘‘policymaking puzzle,’’33 with itsapplicability to local settingsa critical influence on policymakerrationale.

The GDL policy context in-volves significant community andmedia attention because of theemotionality attached to roadtrauma involving young drivers.The resulting social atmosphereinevitably influences politicians,who hold actual policymakingpower. Thus, political rationalesare the most critical mediator ofresearch utilization in GDL policy.Yet, as Frommer and Rychetnikhave suggested,34 although policydecisions may be dominated bypolitical factors, research evidence



may nonetheless inform policy de-bates. Therefore, regardless of theinescapable influence of politics onGDL decision-making identified inthis study, efforts to increase re-search utilization in policy by facil-itating more evidence-informedpolicy debates remains a validstrategy for those aiming to reduceroad trauma involving youngdrivers.

Despite the presence of anevidence-based policymaking cul-ture and close links between re-searchers and other policy actors,individuals involved in GDL pol-icy debates and processes identi-fied several barriers to increasedresearch utilization. Many of thesebarriers—including poor researchquality, an overreliance on peer-reviewed publications as the mainform of research dissemination,and ineffective research fundingmodels—have been identified inother policy settings, with a partic-ular emphasis on the need forimproved methods of disseminat-ing research to policymakers andthe community.9,35 Higher-qualityevaluation research and, in particu-lar, more effective communicationof research to policymakers havebeen found to increase researchutilization in other policy contexts.9

Interviewees suggested that thesestrategies are also likely to be ca-pable of improving existing GDLpolicymaking systems in Australiaand the United States. Thus, ourfindings confirm the contention ofLomas16a that greater ‘‘linkage andexchange’’16b between researchersand policymakers may representan effective method of improvingpolicymaking systems.

Our findings indicate that re-searcher engagement in effective

media advocacy may facilitateadoption of evidence-based strat-egies, given the influence of com-munity perceptions on policy-maker rationales and the highmedia profile of issues relatedto young drivers.20 Therefore, inaddition to the production of high-quality research and the com-munication of research results topolicymakers, effective utilization ofresearch in policy may also requireresearchers to use the media tocommunicate their findings to au-diences outside the scientific andpolitical communities.

Media reporting on GDL issuestends to highlight nonscientificinformation,25 so it may be benefi-cial for GDL policy actors to frameresearch to render it meaningfuland legitimate for nonresearcheraudiences.36,37 Studies situatedin other policy contexts havefound that linking ‘‘startlingstatistics’’23(p161) and tragic victimanecdotes with research on effectiveprevention strategies may help initi-ate reform processes by highlightingopportunities for the introductionof evidence-informed policies thatmay be controversial but that arelikely to be effective.37 Because ofthe prominence of various practical,political, and research issues withinGDL policy discourse,6 linkingevaluation and public consultationresearch to moral arguments (e.g.,community demand for govern-ment to introduce evidence-basedpolicies to ensure the safety of ourchildren) when debating policy al-ternatives may also facilitate effec-tive research dissemination.37

Conclusions

Although studies in other pol-icy sectors have found minimal

utilization of research in pol-icy,7,28–30 individuals involved inpolicy debates and processes gen-erally perceived GDL policy to beevidence-informed, although theyfelt that various factors, includingpolicymaker perceptions of com-munity support for policy alterna-tives, mediated research utilizationin policy. The use of standardizedmeasures to assess the impacts ofthose factors may help verify theself-reports of interviewees and de-termine the relative influence ofsuch key factors in different policy-making environments.

Effective research dissemina-tion strategies may represent thebest way for proponents of evi-dence-based policies to encouragebetter-informed policy debates inthe community and increasedpublic and policymaker supportfor evidence-informed GDL poli-cies. Some researchers may bereluctant to engage in media ad-vocacy, but it is nonetheless criti-cal that researchers strive to en-sure that their findings aredisseminated beyond the scientificcommunity. Given the acuteglobal impact of novice drivertrauma and the potential for in-creased research utilization tofoster more effective policy out-comes, such dissemination strate-gies may offer considerable publichealth benefit. j

About the AuthorsReece Hinchcliff, Rebecca Q. Ivers, andTeresa Senserrick are with the George In-stitute for Global Health, Sydney, Australia.Rebecca Q. Ivers and Teresa Senserrick arealso with the Sydney School of PublicHealth, University of Sydney, Australia.Roslyn Poulos is with the School of PublicHealth and Community Medicine, Univer-sity of New South Wales, Sydney.

Correspondence should be sent to ReeceHinchcliff, George Institute for GlobalHealth, PO Box M201, Missenden Road,Sydney, NSW 2050, Australia (e-mail:[email protected]). Reprints can beordered at http://www.ajph.org by clickingthe ‘‘Reprints/Eprints’’ button.

This article was accepted on March 14,2010.

ContributorsR. Hinchcliff and R.Q. Ivers conceptual-ized the study and supervised all aspectsof its implementation. All authors partic-ipated in interpreting the findings andwriting the article.

AcknowledgmentsR. Hinchcliff was funded by a PhD schol-arship from the NRMA-ACT Road SafetyTrust for this research.

Human Participant ProtectionThis study protocol was approved by theUniversity of Sydney human researchethics committee. Participants providedwritten consent.

References1. Senserrick T, Whelan M. GraduatedDriver Licensing: Effectiveness of Systemsand Individual Components. Clayton, Aus-tralia: Monash University Accident Re-search Centre; 2003.

2. European Conference of Ministers ofTransport, Organisation for EconomicCooperation and Development. YoungDrivers: The Road to Safety. Paris, France:Transport Research Centre; 2006.

3. Williams AF, Mayhew DR. GraduatedLicensing: A Blueprint for North America.Arlington, VA: Insurance Institute forHighway Safety; 2004.

4. Lin ML, Fearn KT. The provisionallicense: nighttime and passenger restric-tions—a literature review. J Safety Res.2003;34(1):51–61.

5. Smith G. License to Drive: YoungDrivers and Nighttime Curfews in Aus-tralia. Canberra, Australia: Asia PacificSchool of Economics and Government,Australian National University; 2004.

6. Blows S, Ivers RQ, Chapman S.‘‘Banned from the streets I have paid touse’’: an analysis of Australian printmedia coverage of proposals for passen-ger and night driving restrictions foryoung drivers. Inj Prev. 2005;11(5):304–308.



7. Hanney S, Gonzalez-Block M, BuxtonM, Kogan M. The utilisation of healthresearch in policy-making: concepts, ex-amples and methods of assessment.Health Res Policy Syst. 2003;1(2).

8. Lavis JN, Ross SE, Jurley JE, et al.Examining the role of health servicesresearch in public policymaking. MilbankQ. 2002;80(1):125–154.

9. Nutley SM, Walter I, Davies HTO.Using Evidence: How Research Can InformPublic Services. Bristol, England: PolicyPress; 2007.

10. Davies HTO, Nutley SM, Smith PC,eds. What Works? Evidence-Based Policyand Practice in Public Services. Bristol,England: Policy Press; 2000.

11. World Health Organization. Reportfrom the Ministerial Summit on HealthResearch: Identify Challenges, Inform Ac-tions, Correct Inequities. Geneva, Switzer-land: World Health Organization; 2004.

12. Jacobson N, Butterill D, Goering P.Development of a framework for knowl-edge translation: understanding usercontext. J Health Serv Res Policy. 2003;8(2):94–99.

13. Shonkoff JP. Science, policy, andpractice: three cultures in search ofa shared mission. Child Dev. 2000;71(1):181–187.

14. Innvaer S, Vist G, Trommald M,Oxman A. Health policy-makers’ percep-tions of their use of evidence: a systematicreview. J Health Serv ResPolicy. 2002;7(4):239–244.

15. Court J, Young J. Bridging Researchand Policy: Insights From 50 Case Studies.London, England: Overseas DevelopmentInstitute; 2003.

16a. Lomas J. Improving Research Dis-semination and Uptake in the Health Sector:Beyond the Sound of One Hand Clapping.Hamilton, Ontario: Centre for HealthEconomics and Policy Analysis; 1997.

16b. Lomas JL. Using ‘‘linkage and ex-change’’ to move research into policy ata Canadian foundation Health Affairs.2000;19(3):236–241.

17. Weiss CH. The many meanings ofresearch utilization. Public Adm Rev.1979;39(5):426–431.

18. Caplan N. The two-communities the-ory and knowledge utilization. Am BehavSci. 1979;22(3):459–470.

19. Crewe E, Young J. Bridging Researchand Policy: Context, Evidence and Links.London, England: Overseas DevelopmentInstitute; 2002.

20. Poulos RG, Zwi AB. Evidence-basedpolicy making? Med J Aust. 2005;182(8):429.

21. Start D, Hovland I. Tools for PolicyImpact: A Handbook for Researchers. Lon-don, England: Overseas Development In-stitute; 2004.

22. Ham C. Health Policy in Britain. Lon-don, England: Macmillan; 1983.

23. Tong A, Chapman S, Sainsbury P,Craig J. An analysis of media coverage onthe prevention and early detection ofCKD in Australia. Am J Kidney Dis. 2008;52(1):159–170.

24. Hartling L, Wiebe N, Russell K,Petruk J, Spinola C, Klassen TP. Gradu-ated driver licensing for reducing motorvehicle crashes among young drivers.Cochrane Database Syst Rev. 2004;2:CD003300.

25. Hinchcliff R, Ivers RQ, Poulos R,Senserrick T, Chapman S. What is the roleof researcher media advocacy within theAustralian road safety policy process?Paper presented at: Australasian RoadSafety Research, Policing and EducationConference; November 10, 2008; Ade-laide, Australia.

26. Waddell C, Lavis JN, Abelson J, et al.Research use in children’s mental healthpolicy in Canada: maintaining vigilanceamid ambiguity. Soc Sci Med. 2005;61(8):1649–1657.

27. Yin R. Case Study Research: Designand Methods. 2nd ed. Thousand Oaks,CA: Sage Publications; 1994.

28. Elliott H, Popay J. How are policymakers using evidence? Models of re-search utilization and local NHS policymaking. J Epidemiol Community Health.2000;54(6):461–468.

29. Percy-Smith J, Burden T, Darlow A,Dowson L, Hawtin M, Ladi S. PromotingChange Through Research: The Impact ofResearch in Local Government. York, En-gland: Joseph Rowntree Foundation;2002.

30. Walshe K, Rundall T. Evidence-based management: from theory to prac-tice in health care. Milbank Q. 2001;79(3):429–457.

31. Nutley SM, Webb J. Evidence and thepolicy process. In: Davies HTO, Nutley SM,Smith PC, eds. What Works? Evidence-Based Policy and Practice in Public Ser-vices. Bristol, England: Policy Press;2000:13–42.

32. Fischer F, Forester J. The Argumen-tative Turn in Policy Analysis and

Planning. Durham, NC: Duke UniversityPress; 1993.

33. Lewis JM. Evidence-based policy:a technocratic wish in a political world. In:Lin V, Gibson B, eds. Evidence-BasedHealth Policy: Problems and Possibilities.Melbourne, Australia: Oxford UniversityPress; 2003:250–262.

34. Frommer M, Rychetnik L. From evi-dence-based medicine to evidence-basedpublic health. In: Lin V, Gibson B, eds.Evidence-Based Health Policy: Problemsand Possibilities. Melbourne, Australia:Oxford University Press; 2003:56–69.

35. Nutbeam D. How does evidence in-fluence public health policy? Tacklinghealth inequalities in England. HealthPromot J Austr. 2003;14(3):154–158.

36. Chapman S, Lupton D, Squires N. TheFight for Public Health: Principles andPractice of Media Advocacy. London, En-gland: BMJ Publishing Group; 1994.

37. Nutbeam D, Boxall A- M. What in-fluences the transfer of research intohealth policy and practice? Observationsfrom England and Australia. PublicHealth. 2008;122(8):747–753.



Utilization of Research in Policymaking for Graduated Driver Licensing

Documents

Transcript of Utilization of Research in Policymaking for Graduated Driver Licensing