Kilpailu ja säätely Hollannin terveydenhoitomarkkinalla, dr. Rein Halbersma.

1. Competition and Regulation
Dutch Health Care Markets
Finnish delegation
Amsterdam, November 6, 2009
dr. Rein Halbersma
Economic Expert, unit for Economic Analysis

2. Introduction
2

3. Reform and public policy objectives
• Cutler (2002): successive waves of healthcare reform aiming at
• Ensuring universal access to medical care
• Centralised regulation-based cost containment by various rationing
mechanisms
• Decentralised market- and incentive-based systems
• Promoting effective competition is not a goal in itself,
but is seen as the best way to deliver the key public policy objectives of:
• Accessibility
• Affordability
• Quality

4. expenses health care and life expectancy in the Netherlands
82 14
12
80
10
78
8
levensverwachting
76
uitgaven als %BNP
6
74
4
72
2
70 0
1953
1956
1959
1962
1965
1968
1971
1974
1977
1980
1983
1986
1989
1992
1995
1998
2001
2004
Bron: CBS en OECD Health Data

5. An overview of the framework
Competition Auctions Regulation State provision
Decentralization Hierarchy
Motivation Coordination
Transaction costs
Market failures Government failures
(market power, externalities, information (information problems, incentive problems,
problems, hold up etc) regulatory uncertainty etc)
5

6. General economic objectives
Coordination
Ensure that the right services are produced at the right
time and place.
(includes financial risk, quality and access)
Motivation
Ensure that the parties have individual incentives to make
coordinated decisions.
Transaction costs
Ensure that coordination and motivation are provided at
the lowest possible cost
(production, search, transportation, contracting,...)
6

7. Scorecard for health care delivery systems
Hierarchy Private Private Self insurance
(public integrated insurers/provider insurers/provider (health
model) (no mandatory (mandatory savings
insurance) insurance) accounts)
Coordination of risk
++ - + --
Motivating health
- ++ ++ +
care providers
public contracting selective contracting Selective contracting
yardstick competition competition policy competition policy
Freedom of choice
++ - - ++
Adverse
++ -- + +
selection/access
Moral hazard
-- - - ++
Rationing Rationing Rationing
Gate keeping Gate keeping Gate keeping
Co-payments Co-payments Co-payments
Transaction cost
Health care Providers/Third party Providers/Third party Patient
providers/government payer/patient payer/patient
7

8. Dutch Healthcare Authority (NZa) and the Regulatory Landscape
8

9. The Dutch Healthcare Authority
Legal tasks of the NZa
• NZa established by the Healthcare Market Design Act (2006)
• Roughly three complementary tasks
• regulating providers and insurers
• mitigating dominant market positions
• initiating market-based reforms where feasible
Organizational structure of the NZa
• Exezcutive Board
• supported by legal & communication staff
• Cure and Care departments
• budget and price regulation
• Supervisory department
• market power assessment
• Research & Development department
• design, advocacy and implementation of reforms
9

10. Mission of the NZa
“The NZa creates and monitors properly functioning healthcare
markets. The interests of the consumers are central in the
performance of these tasks. Efficiency, both in the short and long
term, market transparency, freedom of choice, access to
healthcare and quality are guaranteed. This gives the consumer
the best value for his or her healthcare euros.”
10

11. The Dutch regulatory landscape: competition and quality
The NZa
• ex ante regulation of dominant market positions
• advisory role in merger control
• transparency role in quality control
The NMa (Netherlands Competition Authority)
• ex post regulation of dominant market positions
• decisive role in merger control
• enforcing cartel prohibitions
The IGZ (Healthcare Inspection Agency)
• standard setting and enforcement role in quality control
11

12. The Dutch regulatory landscape: health insurers
The NZa
• ex ante: misleading policies, marketing, consumer targeting
• ex post: legality of reimbursements
The DNB (Dutch Central bank)
• compliance with solvency requirements (Basel II)
The AFM (Authority Financial Markets)
• supervises behaviour of financial institutions
The CBP (Data Protection Authority)
• ensures privacy of patient and client records
12

13. The regulatory toolbox (I)
Regulatory environment for the NZa
• absence of EU level framework (contrast: telecom, energy)
• legal tools endowed by the Healthcare Market Design Act
1. power to impose general obligations (all market parties)
2. power to impose specific obligations on individual parties
General obligations
• law contains no specific criteria for application
• policy objectives: universal access, affordability, quality
• promoting effective competition as a means to this end
• examples:
• transparency requirements (quality, marketing)
• terms of agreements (response time, exclusivitiy)
• price regulation (e.g. general price cap)
13

14. The regulatory toolbox (II)
Specific obligations for individual market parties
• key criterion: Significant Market Power (SMP) = dominance
• relevant market definition on case by case basis
• dominance analysis: market share, structure, effects
• proportional ex ante obligations
• transparency and non-discrimination
• obligation to deal and reference offer
• cost accounting principles and price regulation
• proposed priorization for application of SMP
• exclusion over exploitation
• selling power over buying power
• predatory prices and discrimination:
• only intervene if clearcut foreclosure effects
14

15. Creating effective competition
Systematic effort to deregulate primary care
• 2005: experiment with physiotherapists
• 2007: NZa consulted its deregulation framework
• 2008: proposal to deregulate psychologists, dieticians
• 2010: experiment with dental care
For secondary care, barriers of entry obstruct effective competition
• highly regulated labor markets
• restricted capacity at universities
• 2008: overhaul of capacity planning of specialists
• restricted access to capital markets
• for-profit goal is forbidden, hence equity is scarce
• 2009: proposal to experiment with for-profit goal
15

16. Insurance -Hospital Market
1. Overview
2. Insurance Market
3. The B-segment
4. The A-segment
16

17. Overview of the health care system in the Netherlands
Supplementary Health Insurance (voluntary)
Third Compartment
Mandatory Health Insurance
(compulsory for the entire population)
Second Compartment
National Health Insurance for Exceptional Medical Expenses
(compulsory for the entire population)
First Compartment

18. Health care financing in the Netherlands in the second
compartment
• Public insurance for exceptional medical expenses
• mandatory for all citizens
• home care, nursing homes, care for the handicapped
• 20 G€ annually (1.25 k€ per capita)
• Private basic insurance
• mandatory for all citizens
• general practicioner, hospital care, pharmaceutical care
• 2008: also mental care (moved from public insurance)
• 32 G€ annually (2 k€ per capita)
• Private supplementary insurance
• dental, paramedic (physiotherapy) and cosmetic care
• Total expenditures: 12.4% of GDP (including daycare, public health etc.)
• annual increase of 7.7% since 1998
18

19. The idea behind competitive health care markets
Health care providers
Insurers can selectively contract hospitals
Negotiations between
insurers and hospital
Insurers
Consumers choose between competing insurers
19

20. Economic characteristics of competition
• Consumers have free choice of health insurance company,
• no risk selection, no lock-in
• incentives for prevention?
• Competition between health insurance companies leads to
downward pressure on costs:
• Selective contracting with health care providers
• Directing consumers toward more efficient choices
• Utilization review by insurers:
• Crosschecking need for treatment received
• Best practice benchmarking
20

21. Characteristics of the Dutch health insurance market
• New 2005/2006 legal framework provides for:
• Mandatory health insurance for all Dutch citizens
• Uniform comprehensive benefits package
• Obligation for all health insurers to provide services to all
consumers without:
• risk selection
• premium differentiation
• Funding regime:
• 50% of the premium is a nominal premium (differentiated per
insurer not per consumer) and collected by insurers
• 50% of the premium is income dependent and collected by the
state (this part of the premium is redistributed to insurers based
on a risk adjustment system)
21

22. Risk adjustment system
• Remove financial incentives for risk selection
• Compensates insurers for predictable losses
• Insurers will make an effort to efficiency instead of risk selection
• Fair competition among insurers
• Ex-ante risk adjustment
• Age, sex, source of income (e.g. salary, subsidy)
• Region (classification of postcode areas based on socio-economic,
demographic and healthcare related characteristics of the postcode area)
• Recent outpatient drug consumption (chronic diseases)
• Diagnose (was the patient treated in hospital last year, and does this
predict further high cost treatments/ drugs?)
• Ex-post risk adjustment
• Correction of the ex-ante adjustment. Necessary e.g. because of the
changes in case mix from one year to the next, general cost increase,
unexpected high costs
• Net yearly risk per enrolled consumer 35 Euro
22

23. Insurer supervision: selective contracting
Health Insurance Act
• In theory, health insurers
• do not have to contract every provider, BUT
• do have to contract SUFFICIENT amount of care
• can differentiate payments per provider
• can differentiate deductibles for consumer, depending on the
chosen provider
• In practice:
• every health insurer contracts every provider
• payments are differentiated, BUT
• there is no differentiation in deductibles
• Explanation:
• quality differences between providers not transparent
• consumers value choice more than lower deductible
• we expect this to be a long-run equilibrium
23

24. Market share largest health insurers
Market Shares 2006-2008
35,0%
30,0%
25,0%
20,0% 2006
2007
15,0% 2008
10,0%
5,0%
0,0%
1 2 3 4 5 6 7 8 9 10 11
24

25. Switching behavior
50,0%
45,0%
Sw itching
40,0%
Consideration
35,0%
30,0%
25,0%
20,0%
15,0%
10,0%
5,0%
0,0%
2005 2006 2007 2008
25

26. Premium dispersion
€ 1.250,00
€ 1.200,00
€ 1.150,00
1 quartile
€ 1.100,00 minimum
median
€ 1.050,00 maximum
3 quartile
€ 1.000,00
€ 950,00
€ 900,00
2006 2007 2008
26

27. Collective contracts are popular
80
70
60
50
Collective insurance
40
Individual insurance
30
20
10
-
2005 2006 2007 2008
27

28. Price reductions collective contracts
9,
8,
7,
6,
5, 2006
2008
4, 2007
3,
2,
1,
0,
collective contract patient other collective contracts collective contract
organizations (banks, labor unions etc) employers
28

29. Some conclusions on the insurance market
• The average premium 2006 (1.028 Euro) is below expected premium
(1.106 Euro). The average premium 2007 is 1.103 Euro and 2008
1.049.
• Premiums are difficult to compare as a consequence of adjustments
in 2007 (share of the government) and 2008 (law change for
deductibles).
• 18% of the enrollees switched in 2006 (year of policy change).
Switching in 2007 and 2007 is below 5%
• Elderly people and enrollees with a bad health switch significantly
less.
• Collective contracts are important. Price reductions up to 7,5%.
• 93% of enrollees buy a supplementary insurance.
29

30. B-segment: the introduction of competition for hospitals
30

31. Conditions for competition
• Stable system of invoicing:
• Clear product description.
• Administration performs well.
• Supply side conditions:
• Lower barriers to entry.
• Risk on investment and bankruptcy rules.
• Profit possible.
• Liberalization of contracts between doctors and hospitals.
• Demand side conditions:
• Selective contracting and steering of enrollees.
• Risk adjustment is adjusted to new institutional design.
31

32. 2005: introduction of competition
Revenue of major diagnoses in the competitive
segment
Other Hip replacement
Incontinentence 26% 20%
operation
5% Diabetes Mellitus
Tonsillectomy 12%
6%
Knee 145 products (27
Groin rupture replacement
Cataract diagnoses) of elective
8% 12%
11%
outpatient care
Competitive segment is 8% of total hospital revenue
32

33. Increase of the B-segment in 2008 and 2009
2008
• Chronic heart diseases
• Pregnancy (pregnancy, delivery, after birth control and miscarriage)
• Knee operations (meniscus and cruciate ligament leasie)
• Some cancer treatments (breast and prostate)
• Umbilical hernia
• Sterilization (men and women)
2009
• Treatments in ophthalmology, surgery, orthopedics, urology, gynecology,
neurosurgery, dermatology, internal medicine, cardiology and
anesthesiology.
Competitive segment 2008 and 2009 are an
estimated 20% and 31% of total hospital revenue,
respectively.
33

34. Recent increase in hospital concentration
Voor fusietoezicht Na fusietoezicht
115
110
105
100
95
90
85
1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008
Average HHI hospitals in local market: 2,350
Average HHI insurers in corresponding market: 5,300 34

35. Average number of beds per hospital

36. Price development in the B-segment and A-segment
Percentage price increase/decrease (nominal prices)
2.5
2
1.5
A-segment
B-Slice 2005
1
0.5
0
2005-2006 2006-2007 2007-2008 2008-2009
• The price increase in the A-segment is approx. lower than the price
increase in the B-Slice 2005.
• In 2008-2009, the price increase in the B-slice 2008 was 0.7%.
• Overall, the price development in the competitive sector is more
favorable than the price development in the regulated sector.

37. Price development in the B-segment
Price B-Slice 2005 B-Slice 2008 B-Slice 2009
Development
`05-`06 `06-`07 `07-`08 `08-`09 `08 `08-`09 `09
Nominal 0.0% 2.1% 1.1% 0.8% - 0.7%
Real -1.2% 0.5% -1.4% -0.2% - -0,3%
Mark-up on cost - - - - 2.0% - -0.3%
• For each B-Slice, there is mostly a decline in the real prices.
• In 2008 (2009), we can only calculate the mark-up of B-Slice 2008
(2009) on the estimated cost.
• The estimated cost is already corrected for inflation.

38. A-segment: the regulated segment for hospitals
38

39. Current budgets
A-segment B-segment
Hospital ZBC
Price = pCTG ≤ pCTG Free
Volume Free Free Free
Revenue = TRFB Free Free

40. ‘Functional Budget’ Model (1)
• Cost
- Operational Cost
- Interest and depreciation
B-C= change in reserves
• Budget = max allowed Revenue set by NZa
B-R=change in tariff for
nursing days
• (actual) Revenue earned through billing of
- Tariff for treatments
- Tariff for nursing days

41. ‘Functional Budget’ Model (2)
Component ‘price’ Share in Budget
2004
Fixed Cost per capita 8(10)
Semi-Fixed Cost per bed and per doctor 24 (23)
Variable cost inpatient visits 43 (37)
outpatient visits
nursing days
Location Cost Depreciation/ 16 (16)
interest

42. Care

43. How is the AWBZ market organized?
The AWBZ market is divided into:
• EXTRAMURAL CARE: Health care services delivered outside a
medical institution (hospital, nursing home, psychiatric clinics, etc).
• INTRAMURAL CARE: Health care services delivered in a medical
institution.
In the extramural care there are two different actors:
• PGB clients: These people have a voucher at disposal and they
can organize their care provision themselves. They are free to
receive services from providers that are not contracted by the care
office.
• In kind clients: These people receive services from a provider who
has entered into a contract with the care office.

44. Structure of the extramural AWBZ market
• The Netherlands is divided in 32 regions; each region has a care office
which goals are to:
• Purchase care for their in kind clients. A change in legislation in
2003 made it possible to selectively contract care providers (for all
the functions of the extramural care).
• Inform clients about the contents of a care service that is provided.
• They are accountable for spending financial means for the AWBZ.
• They have a regional budget, which caps the expenditure
• Prices of providers are regulated; maximum tariffs are set by the NZa.
• Providers of care and care offices negotiate on prices. Some services
have a price which is equal to the maximum tariff, but some price
variability is also observed.

45. Overview of the extramural care market

46. Some results from research on the homecare market
1. Positive and significant effect of market share on contracted prices
2. Decreasing relative contracted prices over time.
3. Significant differences in the relative contracted prices across regions
4. No support to the argument of superior quality of large providers (based on
subjective data)

47. Analysis NZa of problems in the current regulation
• Indication: independent indication, but delegation of indication determination to
providers and no clear standard
• No incentives for buying agencies to buy the right health care. No proper health
insurance market.
• No transparency for consumers, waiting lists, low perception of quality and low
efficiency
47

48. Yardstick competition
Yardstick competition is a
dynamically updated price-cap,
that follows the development of the
average unit costs.
Firms have a strong incentive to
improve their unit costs, hence the
average will drop over time.
When the market consists of
regionally fragmented
monopolies (or oligopolies), a
yardstick based on national
average costs can create effective
competition.
Yardstick competition is applied to
Dutch energy companies.

49. What is needed for yardstick competition
Yardstick competition needs
homogenous products that can
be accurately measured.
Dutch hospitals have 30.000
products (“DBCs”) that are
classified by diagnosis and
treatment.
In 2006, the NZa proposed
yardstick competition for
Dutch hospitals, that consists of a
price-cap on the average price
per product, adjusted for their
patient mix.
Owing to bad data registration, this
plan has been cancelled.

50. Long term perspective: Market definition
1. Distinction between the market for basic care and the market for complex and
highly specialized care
2. Within the market for complex care, a number of submarkets exist.
3. Housing and care are separate markets, except when living is an integral part of
the provision of care (e.g. 24 hour supervision in mental health care).
50

51. Long term perspective
• Market for basic care
• No price regulation necessary, if the purchasing is carried out by risk
taking insurers (risk adjustment should be possible)
• Combination with normal health insurance, to solve externalities
• Government monitors market behaviour and quality.
• Market for complex/specialistic care
• Risk adjustment is not possible. Therefore a competitive health
insurance market cannot be established.
• Price regulation is necessary, due to monopolized markets.
51

52. Transition proposals by NZa
Long term model Short term model
Basic Care Extramural Care
Liberalization, possibly
Liberalization with reverse auctions
Purchased by local
communities and
health insurers (if Intramural Care
risk adjustment is
possible)
Yardstick competition
/benchmarking
Complex care
Yardstick competition/
benchmarking

53. Market Definition
53

54. Recent increase in hospital concentration
Voor fusietoezicht Na fusietoezicht
115
110
105
100
95
90
85
1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008
54

55. Background
• Concerns and discussion about mergers in healthcare
• Mergers have to be assessed by NMa and NZa
• Mergers have negative and positive welfare effects (in case of
horizontal merger):
• Reduction of competitive constraints
• Easier coordination (e.g. keeping prices higher)
• Larger size
• Different input and output mix (potentially)
• Merger assessment process
• Predicting the market developments with and without the merger
• Weighing the positive and negative effects against each other
• This project: focus on positive effects
• Idea: measure the potential efficiency gains

56. How do we define markets for antitrust purposes?
• The smallest area or group of products for which there are no close
substitutes outside the group
• The market established using this criterion determines the measure of
market concentration
• Pre-merger and post-merger competitive effects rest upon this definition
• SSNIP Test (EU Merger Guidelines)
• Area or group of products in which a hypothetical monopolist, could
impose a “small but significant and non-transitory increase in price,”
(SSNIP) holding constant the terms of sale for all products produced
elsewhere
Presumption of anticompetitive effects if there is a large
increase in concentration in this area

57. Product market versus geographic market (I)
What is a Product Market?
• Group of products with few outside substitutes
• Smallest sensible segment with Dutch data: Medical specialty
• Total of 24 (e.g. cardiology, neurology)
• Similar to ICD coding of “Major Diagnostic Category”
• Other segmentations:
• Specialties with same complexity/volume (Varkevisser)
• Care type with the same resource requirements (e.g.
primary/secondary tertiary)
• Inpatient versus outpatient
57

58. Product market versus geographic market (II)
What is a Geographic market?
• Geographic area with few outside substitutes outside the area
• Smallest sensible area: zip code
• Look for smallest group of zip codes that make up market
• Supplement markets with additional areas, defining “active
competitors” as hospitals with significant market (>1%) share in
that zip code area
58

59. Patient flow versus patient choice (I)
Patient flow methods
• Assumption: Existing travel pattern is indication of future
preference
• Elzinga-Hogarty method: find zip code area where:
• Few outflows from the area (imports of care) indicates demand self-
sufficiency
• Few inflows to the area (exports of care) indicates supply self-
sufficiency
• Intuitive and easy to compute (and often used in court)
• Problems with EH-method:
• What is “few”? (usually 10% to 25%)
• Sensitive to starting point, expansion method
• Elzinga’s testimony in US court:
• method not suitable for hospital mergers
59

60. Patient flow versus patient choice (II)
Patient choice methods
• Analyze choice conditional on characteristics (distance)
• When characteristics change, choice will change
• Can directly simulate effect of merger on prices
• Methods based on patient choice
• Critical Loss (can also be used with patient flow analysis)
• Uses willingness to travel as proxy for willingness to pay
• BUT: needs to be validated using consumer surveys
• Option Demand
• Uses hospital profits as proxy for willingness to pay
• BUT: for-profit not allowed in the Netherlands
• Logit Competition Index (LOCI)
• Computes price equilibrium for Bertrand competition
• BUT: neglects structure of the insurance market
• NZa has recently implemented all these methods
• Each method has challengeable assumptions, BUT
• Predictions are strongly correlated for all 3 patient choice methods
• This robustness should help in court
60

61. Gains from (horizontal) mergers
Based on work with P. Bogetoft (NZa, 2008)
61

62. System Model
Management
(Effort/Ability)
PROCESS
PROCESS
Resources Products
(Inputs) (Outputs)
Exogenous factors
(Non-discretionary resources or products)

63. Scale (Size)
Output
F∗(y1+y2)
D
y1+y2
A+B
y2 C
B
y1 A
O x1 x2 E∗(x1+x2) x1+x2 Input

64. Scope (Harmony)
Nurses
B
A borrows nurses
and lends doctors
L(y) = resources
necessary to produce
given output
A
O Doctors
64

65. Individual learning
y outputs
P(x)
x inputs
65

66. Interpretations and remedies
Effect Remedy Horizontal
Learning Learn, incentives, change
event
Scope / Mix Exchange/trade inputs and
outputs
Scale Merge
66

67. Average number of beds per hospital

68. Questions?
I am available for further questions at
rhalbersma@nza.nl
68

69. Thank you!
69