OECD – Working Party on National Accounts
Paris, France, 14-16 October 2008
Session 10
Non-market Services
Comparison of input and output-based volume
measures of education and health
Collesi, Versace, Zannoni
Istat, Directorate of National Accounts
Daniela Collesi
ISTAT, Via Agostino Depretis 74/B, 00184
Rome, ITALY
daniela.collesi@istat.it
INTRO
OUTPUT
METHODS
INPUT
METHODS
This presentation aims at proposing:
The comparison between the use of output and
input methods for the two most relevant individual non
market services: education ad health
The comparison of results
INPUT AND
OUTPUT
METHODS
IMPACT ON
GDP
The comparison of the impact on the key
variables of the National Accounts framework
Daniela Collesi - Comparison of input and output-based volume measures of education and health
Deflation methods
INTRO
A short description of the methods currently in use
OUTPUT
METHODS

INPUT
METHODS
INPUT AND
OUTPUT
METHODS
INPUT methods: the value of the “output at constant prices is
based on the deflation of each cost component that, in the
case of non-market services, contributes in forming its value”
P1
PPY
= D1
PPY +
P2
PPY +
K1
PPY +
D29
PPY

OUTPUT methods: the value of the output is based on the
“identification of a volume component, by means of indices of
the quantities produced of the service examined”

Price indices
IMPACT ON
GDP
Daniela Collesi - Comparison of input and output-based volume measures of education and health
Output methods – Education
INTRO
OUTPUT
METHODS
INPUT
METHODS
The public production of educational services is divided into the
followings activity areas:




School system
Vocational training
University education
Subsidiary services to education
INPUT AND
OUTPUT
METHODS
IMPACT ON
GDP
Daniela Collesi - Comparison of input and output-based volume measures of education and health
Education: School system - 1/2
INTRO
OUTPUT
METHODS
INPUT
METHODS
INPUT AND
OUTPUT
METHODS
IMPACT ON
GDP
Public school system is supplied in state schools managed by
Ministry of Education, University and Research, and in other
public schools managed by local authorities: Municipalities,
Provinces and Regions.
Scholastic education is supplied mainly in state schools, and is
divided into four levels:




pre-primary education;
primary education;
lower secondary;
upper secondary education.
Daniela Collesi - Comparison of input and output-based volume measures of education and health
Education: School system - 2/2
INTRO
Data Sources
OUTPUT
METHODS
INPUT
METHODS
INPUT AND
OUTPUT
METHODS
IMPACT ON
GDP
Quantities
Number of pupils in state schools and
non-state schools broken down into the
four levels of education, and in upper
secondary education by type of institute:
- classical lyceum
- scientific lyceum
- teacher-training institutes and schools
- vocational institutes
- technical institutes
- art institutes and lyceums.
Detailed breakdown of the basic data is
essential in order to ensure homogeneity
among the elementary indices and the
costs assigned to them.
Costs
The costs are based on the Cofog
classification of Government expenditure.
The expenditure on Division 9, Education,
provides the cost elements, by the 2°
level analysis, Groups.
For the part concerning services provided
to individuals (groups from 9.1 to 9.6), the
COFOG classification is fully in line with
the ISCED 97.
The quality adjustments are based on
the number of pupils per class
Daniela Collesi - Comparison of input and output-based volume measures of education and health
Education – University - 1/2
INTRO
OUTPUT
METHODS
INPUT
METHODS
INPUT AND
OUTPUT
METHODS
IMPACT ON
GDP
The production of services supplied by universities is split in two
CPA classes (and in COFOG groups too):
 Research and Development, for the part related to research
 Education, for the part concerning didactic services.
The first is deflated by the input method, while the second one is
deflated with the output method.
The quantity indicator used is the number of enrolled students per
faculty and/or group of homogenous faculties (18 faculties)
The quantities are weighted by the cost per student by faculty,
defined by using a methodology based on the standard cost per
student
Daniela Collesi - Comparison of input and output-based volume measures of education and health
Education – University – 2/2
INTRO
OUTPUT
METHODS
INPUT
METHODS
INPUT AND
OUTPUT
METHODS
IMPACT ON
GDP
The quality adjustments are outcome-based.
The indicators used are:
The ratio between the enrolled “regular students” in the course
and the total number of enrolled students
The reduction of the distance between the actual number of
years for graduation and the theoretical length
The two indicators have the same direction, they both tend
to 1. The simplest aggregation of the two indicators has been
preferred, supposing that both indicators have the same weight.
Daniela Collesi - Comparison of input and output-based volume measures of education and health
Health
INTRO
The volume index is based on Diagnosis Related Groups (DRG).
OUTPUT
METHODS
The quantities are represented by the patients discharged from
the public hospitals classified by DRG
INPUT
METHODS
INPUT AND
OUTPUT
METHODS
The costs, that constitute the system of weight, are based on the
sets of fees that are used by the National Health Service to
refund the services supplied
IMPACT ON
GDP
The quality correction is based on the availability of large diagnostic
apparatus with high-tech content at a regional level.
Daniela Collesi - Comparison of input and output-based volume measures of education and health
Input Methods
INTRO
OUTPUT
METHODS
INPUT
METHODS
INPUT AND
OUTPUT
METHODS
IMPACT ON
GDP
The input methods are currently used for the evaluation of
collective services at constant prices; they also serve for
the residual part of the government individual services.
The calculation is based on the deflation of compensation
of employees according to the AW method, as suggested
by the Eurostat Task Force on NACE L services.
The deflation of the other components of output cost, that
is to say other taxes on production,
intermediate
consumptions and consumptions of fixed capital, does not
present any relevant difficulties.
Daniela Collesi - Comparison of input and output-based volume measures of education and health
INPUT methods – The Average Wage method
INTRO
OUTPUT
METHODS
INPUT
METHODS
INPUT AND
OUTPUT
METHODS
For the deflation of the compensations of employees (D1) of
General Government sector (which make up more than 60% of
the overall production value at current price) the Average Wages
Method was used, based on Annual Count Survey (CA) data,
produced yearly by the State accounting department (RGS) .
The AW method allows to evaluate in a complete manner this
cost component that represents the most important part of the
General Government output.
100%
90%
80%
Percentage incidence of the
items making up the Total
NMK output of the General
Government. Elaborations
on data at current prices,
2000 - 2007
70%
60%
IMPACT ON
GDP
50%
40%
30%
20%
10%
0%
2000
2001
2002
2003
D1
D29
2004
K1
2005
2006
2007
P2
Daniela Collesi - Comparison of input and output-based volume measures of education and health
Comparison for Education
INTRO
Education Output at constant prices
OUTPUT
METHODS
2000
2001
2002
2003
2004
2005
2006
2007
_____________________ previous year price, millions euro ____________________
INPUT
METHODS
INPUT AND
OUTPUT
METHODS
IMPACT ON
GDP
Input method
Output method
Difference
48,795
48,622
-173
51,292
51,148
-144
52,042
54,073
2,031
56,238
55,875
-363
59,541
57,304
-2,237
56,619
56,788
169
59,305
59,781
476
60,859
61,670
811
Output
Compensation of employees
_____________________ current prices, millions euro ________________________
50,099
52,897
54,858
57,963
56,551
60,173
61,119
64,665
41,506
43,660
45,971
48,322
46,476
49,654
50,996
54,540
_____________________price index, previous year = 100 _____________________
Input method
Output method
102.7
103.0
103.1
103.4
105.4
101.5
103.1
103.7
95.0
98.7
106.3
106.0
103.1
102.2
106.3
104.9
The compensation of employees represents more than 83 per cent of the output
at current prices. This explains the relevant effect that the changes in this
aggregate, due to the contract renewals, have on the output price index (the
implicit deflator).
Daniela Collesi - Comparison of input and output-based volume measures of education and health
Comparison for Education
INTRO
OUTPUT
METHODS
In order to apply the input method by product, it has been done an attempt to
breakdown the compensation of employees to take into account the level of education.
% distribution of workforce by occupation profile and product (level of education)
Level of education
Detailed of occupation profile
INPUT
METHODS
Managers
Headmaster
General manager
INPUT AND
OUTPUT
METHODS
IMPACT ON
GDP
Pre-primary
education
Pre-primary
Lower
and primary secondary
education
education
Primary
education
Lower and
Not
upper
allocated by
secondary
level of
education
education
Upper
secondary
education
Total
-
-
-
-
-
-
0.69
0.68
0.00
0.69
0.68
0.00
Levels
Accountant
Assistant
Coaching staff
General manager
Staff
Teacher
Teacher for "special classes"
Technician
8.22
0.16
7.33
0.73
-
24.39
21.47
2.91
-
1.17
1.17
-
17.73
15.37
2.36
-
24.82
0.14
0.02
22.92
1.74
-
1.05
1.05
-
21.94
5.32
13.73
0.26
0.86
0.08
1.69
99.31
5.62
13.75
0.26
0.86
0.08
69.31
7.74
1.69
Total
8.22
24.39
1.17
17.73
24.82
1.05
22.62
100.00
IGOP: Conto annuale 2006
As almost 23% of employees is not allocated by level of education, only three
products can be analysed: Scholastic and other education, University and
Subsidiary services to education
Daniela Collesi - Comparison of input and output-based volume measures of education and health
Comparison for Education
INTRO
2001
2002
2003
2004
2005
2006
_____________________ previous year price, millions euro ______________________
OUTPUT
METHODS
Input method, of which :
Compensation of employees by level of education:
Scholastic and other education
University
Subsidiary services to education
51,215
42,116
39,570
2,453
93
51,989
43,166
40,561
2,531
74
56,181
46,811
43,898
2,794
119
59,535
49,484
46,585
2,804
95
56,599
46,416
43,465
2,857
94
59,299
49,348
46,189
3,060
99
INPUT
METHODS
Output method
Difference
51,148
-67
54,073
2,084
55,875
-306
57,304
-2,231
56,788
189
59,781
482
INPUT AND
OUTPUT
METHODS
IMPACT ON
GDP
Output
Compensation of employees
_____________________ current prices, millions euro ________________________
61,119
60,173
56,551
57,963
54,858
52,897
50,996
49,654
46,476
48,322
45,971
43,660
______________________ price index, previous year = 100 _____________________
Input method
Output method
103.3
103.4
105.5
101.5
103.2
103.7
95.0
98.7
106.3
106.0
103.1
102.2
The result is very close to that obtained without the breakdown. This could actually mean that the stratification,
already enclosed in the method, is detailed enough to allow to catch the differences since the beginning; on the
other hand it could also mean that there is no relevant difference in the ratio between prices (compensation of
employees) and quantities (number of teachers, etc.) among the different products (the level of Education).
Daniela Collesi - Comparison of input and output-based volume measures of education and health
Comparison for Health
INTRO
Health Final consumption expenditure at constant prices
2000
OUTPUT
METHODS
INPUT
METHODS
IMPACT ON
GDP
2002
2003
2004
2005
2006
2007
__________________ previous year price, millions euro ______________________
Input method
Output method
Difference
INPUT AND
OUTPUT
METHODS
2001
34,869
35,524
655
38,066
39,020
954
40,474
40,817
343
43,341
43,046
-295
45,529
45,848
319
49,491
50,138
647
51,943
53,305
1362
56,203
55,426
-777
________________ price index, previous year = 100 _____________________
Input method
Output method
106.4
104.5
103.8
101.3
103.2
102.3
100.8
101.5
106.8
106.1
104.5
103.1
105.7
103.0
98.8
100.2
When analysing the effect of the different methods on the aggregate Final
consumption expenditure the relevant role of Social transfers in kind should be
kept in mind. This item is often subject to budget constraints, as it happens to
intermediate consumption, also whether not necessarily in conjunction with them.
Also political actions/measures that aims at introducing efficiency in Health may
have effect on the expenditure for Social transfers in kind.
Daniela Collesi - Comparison of input and output-based volume measures of education and health
Impact of input and output methods on GDP
INTRO
Final consumption expenditure
OUTPUT
METHODS
GDP chain-linked volumes
(2000 = 100, millions euro)
2000
2001
2002
2003
2004
2005
2006
2007
1,191,057
1,212,713
1,218,220
1,218,013
1,236,671
1,243,525
1,266,420
1,284,868
Government final consumption expenditure in Education and Health/GDP (%)
INPUT
METHODS
Input method
7.2
7.2
7.2
7.4
7.5
7.6
7.4
7.3
Output method
7.2
7.3
7.4
7.6
7.6
7.7
7.6
7.6
2000
2001
2002
2003
2004
2005
2006
2007
1,191,057
1,212,713
1,218,220
1,218,013
1,236,671
1,243,525
1,266,420
1,284,868
Value added
INPUT AND
OUTPUT
METHODS
IMPACT ON
GDP
GDP chain-linked volumes
(2000 = 100, millions euro)
Government Value added in Education and Health/GDP (%)
Input method
5.7
5.7
5.6
5.7
5.7
5.7
5.6
5.5
Output method
5.7
5.8
5.9
5.9
5.8
5.8
5.8
5.7
Daniela Collesi - Comparison of input and output-based volume measures of education and health
Conclusions
INTRO
OUTPUT
METHODS
INPUT
METHODS
INPUT AND
OUTPUT
METHODS
IMPACT ON
GDP
The comparison between the two methods has
highlighted several problems both on the theoretical and on the
practical grounds.
The input method implies to determine a price for each
product, that is for each level of education or for each Diagnosis
Related Group in the case of Hospital services.
An appropriate application of an input method would
imply to work on a detailed database that, for the time being,
does not exist.
Furthermore the substitution of output methods with input ones
would also prevent from the calculation of whatever kind of
productivity measure of government sector.
Daniela Collesi - Comparison of input and output-based volume measures of education and health
INTRO
OUTPUT
METHODS
Thank you!
INPUT
METHODS
INPUT AND
OUTPUT
METHODS
IMPACT ON
GDP
Daniela Collesi
ISTAT, Via Agostino Depretis 74/B, 00184
Rome, ITALY
daniela.collesi@istat.it
Daniela Collesi - Comparison of input and output-based volume measures of education and health