Choice, Competition and Public Service Reform

Julian Le GrandLondon School of Economics

Australia 2014

Central Problem of Public Service Delivery

• How to drive up quality in public services (school education, higher education, health care, social care – and others)

• Do poor providers (hospitals, schools, children’s homes, social work offices) need incentives to improve? Is the provision of incentives irrelevant – or damaging?

Models of Public Service Delivery

• Trust

• Mistrust

Models of Public Service Delivery (4)

• Voice

• Choice

Models and Motivation

Each model incorporates one or more assumptions concerning the motivation of employees working in human services. They are either assumed to be largely motivated by altruism and professional values (trust and voice models), by self-interest (mistrust model) or some combination (choice). That is, they are assumed to be knights, knaves or a mixture of the two.

Of Knaves and Knights

’In contriving any system of government, and fixing the several checks and controls of the constitution, every man ought to be supposed a knave and to have no other end, in all his actions, than private interest. By this interest, we must govern him and, by means of it, notwithstanding his insatiable avarice and ambition, co-operate to the public good’ David Hume

Trust Model

• Problem is identified as the constraints: resources/silos. ‘Give us the money and we will finish the job’.

• Problem is unawareness. ‘Tell us what’s wrong and we’ll put it right’.

Motivational assumption: No incentives necessary. Resolve the central problem, and intrinsic motivation sufficient: providers are professional knights not knaves.

Trust model: advantages & disadvantages

• Advantages– Liked by producers & professionals– No gaming– No monitoring costs & low transaction costs

• Disadvantages– Rewarding failure resentment from those do

deliver.– Knights may have their own agenda– Doesn’t work…….

Mistrust Model

• Mistrust (1): Ministerial command and control. UK version: targets and performance management.

• Mistrust (2): Regulation.Motivational assumption: primarily knaves. Employees need incentives:• Promotion/dismissal• Autonomy/restriction

Mistrust model: advantages and disadvantages

Advantages:• It works…….

% patients waiting for hospital admission > 12 months

Source: Are improvements in targeted performance in the English NHS undermined by gaming: A case for new kinds of audit of performance data? Gwyn Bevan and Christopher Hood, British Medical Journal (forthcoming)

0%

5%

10%

15%

20%

25%

30%

2000 2001 2002 2003 2004 2005

% p

atie

nts

wa

itin

g >

12

mo

nth

s

England

Northern Ireland

% patients waiting for hospital admission > 12 months

Source: Are improvements in targeted performance in the English NHS undermined by gaming: A case for new kinds of audit of performance data? Gwyn Bevan and Christopher Hood, British Medical Journal (forthcoming)

0%

5%

10%

15%

20%

25%

30%

2000 2001 2002 2003 2004 2005

% p

atie

nts

wa

itin

g >

12

mo

nth

s

England Northern IrelandScotlandWales

% patients waiting for hospital admission > 12 months

Source: Are improvements in targeted performance in the English NHS undermined by gaming: A case for new kinds of audit of performance data? Gwyn Bevan and Christopher Hood, British Medical Journal (forthcoming)

0%

5%

10%

15%

20%

25%

30%

2000 2001 2002 2003 2004 2005

% p

atie

nts

wa

itin

g >

12

mo

nth

s

England

Northern Ireland

Scotland

Wales

% patients waiting for hospital admission > 12 months

Source: Are improvements in targeted performance in the English NHS undermined by gaming: A case for new kinds of audit of performance data? Gwyn Bevan and Christopher Hood, British Medical Journal (forthcoming)

0%

5%

10%

15%

20%

25%

30%

2000 2001 2002 2003 2004 2005

% p

atie

nts

wa

itin

g >

12

mo

nth

s

England

Northern Ireland

Scotland

Wales

% Patients spending less than 4 hours in A +E

0

20

40

60

80

100

Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2

2002/03 2003/04 2004/05 2005/06

%

Source: Chief Executive's Report on the NHS - Statistical Supplement (December 2005)

+ 24% increase in A+E admittances

Mistrust model: disadvantages

• Targets set centrally.• Distortion: ‘hit the target and miss the point’• Gaming.• Demotivating and demoralising – especially

for professionals. Tums knights into knaves.

Voice

• Mechanisms: Complaints, petitions, public meetings, community councils. No direct incentives

• Advantages: gives information, personal relationship.

• Disadvantages: clumsy, relies upon knightly motivations – and inequity. Favours the confident, the articulate: the better off.

Choice: types

• Choice of Provider (where?)– Hospital, GPs, Schools

• Choice of Treatment (what?)– Treatment, procedures, curriculum, teaching styles

• Choice of Time (when?)– Appointment time, opening hours

• Choice of Access Channel (how?)– Face to face, phone, web

Choice of Provider

• Providers are independent. Non-profit or for-profit. Keep any surplus they make on their budget

• Users choose provider. Money follows the choice. So hospitals, schools etc get more resources through the number of users (patients, pupils) they attract.

• Funding formula or individual budgets

Choice and competition: advantages

• Choice is intrinsically desirable• Choice is instrumental: leads to greater autonomy

and welfare

Intrinsic desirability of choice and self-determination theory

• Autonomy desirable (important basic value, promotes greater well-being, life-expectancy etc)

• Greater choice greater autonomy.

But does it lead to greater welfare? Too much choice can lead to confusion and regret

Choice and competition: advantages

• Choice is intrinsically desirable• Choice is instrumental: leads to greater autonomy

and welfare• It works. Provides strong incentives for

responsiveness and efficiency. Evidence (from US and UK) suggests that fixed price systems in health care lower costs and increase quality. Education evidence more mixed.

• Promotes equity through diminishing the power of voice.

Competition in health care: Evidence from UK.

• Following roll out of choice in UK, AMI mortality dropped faster in more competitive areas (Zack Cooper, Carol Propper).

• Introduction of ISTCs (Independent Sector specialist treatment centres) led to technical efficiency improvements in NHS competitors (Cooper)

• Expansion of market since 2008 to inclide any willing provider: no further improvement?

Waiting times: an equity problem

145

150

155

160

165

170

175

180

185

190

195

200

I II III IV V

Level of Deprivation (V = Most Deprived Quintile)

Days

Wait

ed

The poor waited up to 30-days longer than the wealthy

Hip replacement broken down by deprivation

0

50

100

150

200

250

300

1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007

Year

Days W

ait

ed I

II

III

IV

V

Choice and competition: advantages

• Choice is intrinsically desirable• Choice is instrumental: leads to greater autonomy

and welfare• It works. Provides strong incentives for

responsiveness and efficiency. Evidence (from US and UK) suggests that fixed price systems in health care lower costs and increase quality. Education evidence more mixed.

• Promotes equity through diminishing the power of voice.

• Can appeal to both the altruist (knight) and the self-interested (knave).

Choice and Competition:Non-problems/fallacies

• ‘People don’t want choice; they want a good local school’. False alternative.

• Only middle classes want choice.• Choice and competition don’t belong in

‘the public realm’.

Parental Choice in New Zealand• 96% of parents indicated they would like to select the

school their child goes to• 80% of parents agreed that education should be

funded such that parents can afford to send their children to the school of their choice.

• A higher proportion of parents with annual income of $30,000 or less strongly agreed with the statement than parents with an annual income of over $30,000.

• Source: Steven Thomas and Ruth Oates The Parent Factor Report Four: Access to Education. Auckland: the Maxim Institute, 2005

Minorities and Choice in the US

• 52 per cent of parents, and 59 per cent of public school parents, supported school choice.

• 60 per cent of minorities supported vouchers. • 87 per cent of black parents aged 26-35 and

66.4 per cent of blacks aged 18-25 supported vouchers.

Who wants choice UK: Gender

56%

69%

0%

20%

40%

60%

80%

100%

Men Women

% saying peopleshould have a greatdeal or quite a lot ofsay over which hospitalto go to if they needtreatment

62%

64%

64%

59%

67%

55% 60% 65% 70%

semi routineand routine

Lowersupervisory& technical

Self-employed

Intermediateoccupations

Managerialand

professional

soci

al c

lass

% saying peopleshould have a greatdeal or quite a lot ofsay over which hospitalto go to if they needtreatment

Who wants choice UK: Social Class

59%

57%

65%

70%

0% 20% 40% 60% 80% 100%

>£50,000

£32,000 -£50,000

£10,000 -£32,000

< £10,000

inc

om

e

% saying peopleshould have agreat deal or quitea lot of say overwhich hospital togo to if they needtreatment

Who wants choice UK: Income

56%

59%

67%

69%

0% 20% 40% 60% 80% 100%

Highereducation

A Level

GCSE/OLevel

None

Hig

hest

edu

catio

nal q

ualif

icat

ion

% saying peopleshould have a greatdeal or quite a lot ofsay over whichhospital to go to ifthey need treatment

Who wants choice UK: Educational Achievement

London Choice Pilot: % opting for an alternative hospital

Source: Evaluation of the London Patient Choice Scheme, Picker Institute (July 2005)

69%65% 65% 67% 68% 67% 66% 67%

63%

73%

0%

20%

40%

60%

80%

100%

Fe

ma

le

Ma

le

Ba

sic

Hig

he

r

Be

low

ave

rag

e

Ab

ove

ave

rag

e

No

n-w

hite

Wh

ite

No

te

mp

loye

d

Em

plo

yed

Sex Educationalstatus

Householdincome

Ethnic Group EmploymentStatus

Choice and Competition:Problems

• Natural or geographical monopoly. Possibility of franchising.

• Information. For choice to work, users must be informed, or have informed agent.

• Transactions costs high.• Opportunities for cream-skimming. Cream-

skimming: selecting easiest, least costly pupils, clients, patients. Favours less needy.

• Cover for privatisation

Choice and Competition: problems (cont)

• Failure. Need mechanism by which poor providers can ‘exit’ the market. Politically difficult.

• Fragmentation. Loss of benefits of economies of scale (If they exist)

• Need excess capacity (though not very much)

Avoiding problems of choice: cream-skimming

• Stop-loss insurance

• No discretion over admissions

• Incentives. Risk adjustment

• Pupil premium in school education

Avoiding problems of choice: Reputation Competition

Incentiveso Individuals’ esteem (spotlight effect)o Publicity of good / poor performance

Mechanismso ‘Naming & shaming’ organisationo Rankingo Publication: widely disseminated

Reputation Competition Advantages & Disadvantages

• Advantages– Can work– Low transaction costs

• Disadvantages– Problems of selection– Can result in gaming– Damage to morale from ‘naming &

shaming’. Unpopular.– No direct incentives to improve

% pupils achieving > 5 good grades GCSE A* to C

Note: data are missing for 1998, 2001 & 2005 & these have been estimated as the mean values from adjacent years

Wales stops publication league tables

1999 2001 2003 2005 200745

55

65

75

85

England Wales

Star ratings published

Target

Source: Bevan & Hamblin (2009)

% Ambulance response times to life-threatening emergencies < 8 minutes

Problems of choice and competition: privatisation?

• Is drive for competition just a cover for privatisation?

• The for-profit question. Motivation of shareholders knavish; non-profits/social enterprises mixture of knight and knave.

• The Public Service Mutual

The Public Service Mutual

• The Public Service Mutual (PSM) is an employee-led (and owned) organisation, often spun out of the public sector, delivering a public service.

• Mutuals across sectors and countries deliver:– High user satisfaction– Greater productivity– High employee morale (especially professionals)