Bortoletti, measures of corruption, commissione europea, ipa zagabria 21 23 novembre 2011
-
Upload
maurizio-bortoletti -
Category
Documents
-
view
73 -
download
0
description
Transcript of Bortoletti, measures of corruption, commissione europea, ipa zagabria 21 23 novembre 2011
TABLE OF CONTENTS Measures of corruption
1
Judicial Measures Corruption indexes based on direct experience of the phenomenon Objective measures of corruption
TABLE OF CONTENTS Measures of corruption
2
In the last few decades, there was a renovated attention in the analysis of corruption, that has been accompanied by both a substantial increase in sensitivity by public and private, national and international institutions and a greater concern in developing tools to measure corruption-related phenomena. Measuring phenomena such as corruption, whose nature, causes and consequences are the ultimate objectives of the investigation itself, is now recognized essential, not only in the social sciences but also in the public debate and, in the end, it is deemed a prerequisite to planning any policy strategy. The difficulties encountered in measuring corruption-related phenomena derive, ultimately, from the fact that corrupt practices are illegal, and for this reason, those who are responsible for such practices have a natural tendency to occult them. In this paper, we provide an overview of the various alternatives available to measure corruption. The measurement techniques are subject to empirical and conceptual limitations, forcing us to consider cum grano salis the results of the many empirical researches aiming at the identification of causes and effects of corruption.
TABLE OF CONTENTS Judicial Measures
3
The first measurement technique we are to describe is based on the use of judicial statistics. For example, once defined “corruption” within a justice system, it is possible to consider as a proxy for “corruption” the number of people sentenced for corruption-related crimes. The underlying idea is that, if it is true that corrupt practices are usually intended to be secret, the role of the justice system is precisely to bring to light and punish them. From a scientific point of view, such a reasoning can be maintained only if the justice system punishes corruption-related crimes with a constant probability, in space or in time. Such assumption is often a too strict one: in fact, the number of convictions has no unambiguous interpretation. Actually, a high number of convictions can indicate both the presence of a high number of corruption-related crimes, a rampant phenomenon whose explicit part is nonetheless very little, or the opposite, in the case that national judges were very efficient in fighting corruption-related phenomena. On the other hand, a reduced number of convictions could be a sign that the judicial power is corrupted as well or that the level of corruption in such an area is actually very low.
TABLE OF CONTENTS Judicial Measures
4
Actually, a high number of convictions can indicate both the presence of a high number of corruption-related crimes, a rampant phenomenon whose explicit part is nonetheless very little, or the opposite, in the case that national judges were very efficient in fighting corruption-related phenomena. On the other hand, a reduced number of convictions could be a sign that the judicial power is corrupted as well or that the level of corruption in such an area is actually very low. In the first scenario, a high number of convictions can be an indicator of both high corruption and of great judicial activity – how to distinguish? In the second environment, on the other hand, both the number of convictions and the probability of punishment are low – is that an indicator of a high or of a low level of corruption in society? In other words, the problem with judicial statistics is that they measure jointly two different and hardly distinguishable sides of the phenomenon: its incidence and the judicial authorities’ activism. Moreover, the probability that the judicial system identifies a case of corruption is usually unknown to the researchers, since nobody will never be able to know exactly how many crimes remain invisible and unpunished, no matter how many are uncovered and castigated. In the end, we may conclude that the judicial statistics, at best, allow very relative evaluations about the problem extent.
TABLE OF CONTENTS Judicial Measures
5
Despite these considerations, judicial statistics are still of some interest. Firstly, provided a more or less constant justice activism can be assumed, they actually allow to draw some interesting conclusions on the relative level of corruption. Namely, when the comparison is realised between different areas within the same country. Furthermore, judicial statistics can be used to control the judicial system itself. For example, observing that, in a certain area of a country, there is a significantly lower average number of convictions for corruption crimes than in other areas, could suggest a lower degree of effectiveness of the judicial power in that particular area, not to mention the possibility that the judicial power itself is corrupted. Finally, it is important to underline that, being all judicial statistics the result of conceptually different but intimately linked phenomena, their study and interpretation is always a very problematic issue. After these preliminary clarifications, let us consider some judicial statistics about different Italian regions.
TABLE OF CONTENTS Judicial Measures
6
We present some data regarding both denounces (possibly with arrest) and convictions. The two types of information differ in several dimensions: first of all, not all denounced cases result in convictions and, secondly, the actual definition and nature of the crime identified in the denounce may change gradually in the course of a hypothetical investigation (for example, from a general “abuse of office” to a more detailed “corruption”). For these reasons, we do not present information only about corruption crimes, but also about other types of crimes against the public administration, such as “abuse in office acts”, “embezzlement” and “bribery”. As data related to convictions refer to the year of the final judgement and the time of justice can be so long that some crimes can even expire, it is apparent that the date of convictions reported by the statistics could refer to crimes committed many years before
TABLE OF CONTENTS Judicial Measures
7
2004 2005 2006 2007 2008 2009 2010
Delitti
registrati
3.403 3.550 5.498 3.367 3.413 3.230 3.076
Persone denunciate
F M F M F M F M F M F M F M
2.812 9.935 2.944 10.903 5.785 15.019 3.561 10.418 3.091 11.309 3.651 9.963 3.591 10.233
12.747 13.847 20.804 13.979 14.400 13.614 13.824
Fonte : Sistema di Indagine del Ministero dell’Interno.
NOTA METODOLOGICA: sono state considerate le fattispecie p. e p. dagli artt. : 414, 316, 316 bis, 316 ter, 317, 318, 319, 319 ter, 320, 322, 322 bis, 323, 353, 354, 355, 356, 640 bis del codice penale. Pur se evidentemente significativa, la fattispecie p. e p. dall’art. 640, 2°comma, c.p., dedicata a punire la fattispecie aggravata prevista nei casi di consumazione a danno dello Stato, non è stata considerata, in quanto il Sistema di Indagine, nella registrazione delle denunce per l’art. 640 bis cp, prevede tale specificazione quale voce facoltativa lasciata alla discrezionalità dell’operatore.
TABLE OF CONTENTS Judicial Measures
8
2004 2005 2006 2007 2008 2009 2010
Persone denunciate
F M F M F M F M F M F M F M
2.812 9.935 2.944 10.903 5.785 15.019 3.561 10.418 3.091 11.309 3.651 9.963 3.591 10.233
12.747 13.847 20.804 13.979 14.400 13.614 13.824
Fonte : Sistema di Indagine del Ministero dell’Interno.
25.435 donne denunciate su 103.215 persone denunciate dal 2004 al 2010, segnalano una “presenza” prossima al 25%, con alcune annualità dove tale percentuale s@iora il 30%: • il 22 % nel 2004, con 2.812 donne su 12.747 persone denunciate; • il 22 % nel 2005, con 2.944 donne su 13.847; • il 28 % nel 2006, con 5.785 donne su 20.804; • il 25% nel 2007, con 3.561 donne su 13.979; • il 22 % nel 2008, con 3.091 donne su 14.400; • il 27% nel 2009, con 3.651 donne su 13.614; • di nuovo il 27% nel 2010, con 3.591 donne denunciate su 13.824 soggetti segnalati.
TABLE OF CONTENTS Judicial Measures
9
Reati commessi da “chiunque”,
Reati p. e p. dagli art. 322 e 353 cp
Reati consumati da P.U. o da Incaricati di P.S.
Reato p. e p. dall’art. 323 cp TOTALE
Delitti consumati 11.956 2.147 4.048 7.386 25.537
Donne denunciate 19.421 757 2.279 2.978 25.435
Maschi denunciati 40.465 6.263 12.478 18.574 77.780
T o t a l e p e r s o n e denunciate 59.886 7.020 14.757 21.552 103.215
Fonte : Sistema di Indagine del Ministero dell’Interno.
1. In tale dato è ricompresa anche l’ipotesi di cui all’art. 321 cp, “Pene per il corruttore”, che realizza la @igura della corruzione quale reato a concorso necessario estendendo al privato le stesse pene del pubblico uf@iciale o dell’incaricato di pubblico servizio.
2. Come già illustrato alla nota 69 per l’ipotesi p. e p. dall’art. 640 II comma cp, il Sistema di Indagine non consente anche per queste due fattispecie di distinguere tra le diverse ipotesi sanzionatorie rispettivamente previste, alcune delle quali costruite come reato “proprio” e altre come reato “comune”, perché prevede tale speci@icazione quale voce facoltativa lasciata alla discrezionalità dell’operatore.
3. Il dato comprende anche i soggetti cui l’art. 322 bis cp estende la punibilità rispetto alle fattispecie p. e p. dagli art. 314, 316, 317 – 320, 322 terzo e quarto comma cp.
TABLE OF CONTENTS Judicial Measures
10
2004 2005 2006 2007 2008 2009 2010
truffa per il conseguimento …
(art.640 bis cp)
4.942 5.541 11.102 6.201 4.361 4.887 5.744
F M F M F M F M F M F M F M
1.504 3.438 1.387 4.154 3.817 7.285 2.052 4.149 1.252 3.109 2.063 2.824 2.172 3.572
età
media 42 44 40 46 36 40 39 43 40 45 44 47 47 49
indebita percezione
(art. 316 ter cp)
1.322 1.818 2.034 1.033 1.244 711 1.153
F M F M F M F M F M F M F M
501 821 764 1.054 899 1.135 428 605 583 661 309 402 460 693
età
media 42 45 42 47 38 42 43 45 46 47 47 48 54 52
Fonte : Sistema di Indagine del Ministero dell’Interno.
TABLE OF CONTENTS Judicial Measures
11
2004 2005 2006 2007 2008 2009 2010 totale
corruzione (art. 318, 319, 320 cp) 158 126 112 128 148 171 96 939
concussione (art. 317 cp) 138 115 86 130 145 140 127 881
TOTALE Reati contro la PA 3.403 3.550 5.498 3.367 3.413 3.230 3.076 25.537
Fonte : Sistema di Indagine del Ministero dell’Interno.
Ribadita, per evitare equivoci, la necessità di contrastare il fenomeno senza “se” e senza “ma”, l’analisi non può evitare di evidenziarne una importanza, in termini percentuali, quasi impalpabile: • il 9 %, nel 2004, con 296 denunce sulle 3.403; • il 7 %, nel 2005, con 241 denunce su 3.550; • il 3,5 %, nel 2006, con 198 denunce su 5.448; • l’8 %, nel 2007, con 258 denunce su 3.367; • l’8 %, nel 2008, con 293 denunce su 3.413; • il 10 %, nel 2009, con 311 denunce su 3.230; • il 7 %, nel 2010, 223 denunce su 3.076, per una percentuale complessiva nel periodo “2004-‐2010” del 7 %, con 1.820 denunce per corruzione e concussione sui 25.537 reati contro la PA.
TABLE OF CONTENTS Judicial Measures
12
2004 2005 2006 2007 2008 2009 2010 totale
corruzione (art. 318, 319, 320 cp) 1.287 663 1.382 658 1.305 1.506 888 7.689
concussione (art. 317 cp) 256 258 222 262 354 315 338 2.005
TOTALE Reati contro la PA 12.747 13.847 20.804 13.979 14.400 13.614 13.824 103.215
Fonte : Sistema di Indagine del Ministero dell’Interno.
TABLE OF CONTENTS Judicial Measures
13
Fonte : Sistema di Indagine del Ministero dell’Interno.
corruzione (art. 318, 319, 320 cp)
concussione (art. 317 cp)
2004-‐2010 2009 2010 2004-‐2010 2009 2010 ABRUZZO 28 4 3 25 2 3 BASILICATA 9 0 1 10 1 2 CALABRIA 45 4 6 50 5 7 CAMPANIA 186 61 26 123 19 20 EMILIA R. 42 8 2 46 8 4 FRIULI V.G. 7 2 1 6 0 2 LAZIO 104 15 12 85 17 14 LIGURIA 29 3 6 24 2 6 LOMBARDIA 130 18 9 127 12 24 MARCHE 6 1 1 12 2 1 MOLISE 12 0 0 12 4 1 PIEMONTE 56 6 3 49 10 9 PUGLIA 70 5 14 86 13 5 SARDEGNA 8 1 2 12 2 5 SICILIA 76 18 0 117 24 14 TOSCANA 53 9 4 46 9 6
TRENTINO A.A. 6 1 0 7 2 1
UMBRIA 29 4 5 11 1 1 V. D'AOSTA 1 0 0 0 0 0 VENETO 42 11 1 33 7 2 TOTALE 939 171 96 881 140 127
TABLE OF CONTENTS Judicial Measures
14
2006 2007 2008
M F Tot M F Tot M F Tot
concussione
• inizio azione penale 267 16 283 327 20 347 365 22 387
• condannate n.d. n.d. n.d. 99 2 101 72 6 78
corruzione complessivo
• inizio azione penale 1.030 181 1.211 1.246 157 1.403 1.056 133 1.189
• condannate n.d. n.d. n.d. 212 28 240 191 28 217
Fonte: ISTAT, Rilevazione dei delitti denunciati per cui è iniziata l’azione penale e sui condannati con sentenza irrevocabile, ottobre 2010
TABLE OF CONTENTS Judicial Measures
15
2004 2005 2006 2007 2008 2009 2010 totale
Delitti
registrati
173 167 184 195 246 217 213 1.395
Persone denunciate
F M F M F M F M F M F M F M F M
12 164 22 161 21 194 12 109 25 259 25 214 28 216 145 1.317
176 183 215 121 284 239 244 1.462
Fonte : Sistema di Indagine del Ministero dell’Interno.
DENUNCE PER REATI CONTRO LA P.A., PER TIPOLOGIA DI REATO: ISTIGAZIONE ALLA CORRUZIONE
TABLE OF CONTENTS Judicial Measures
16
DENUNCE PER REATI CONTRO LA P.A., PER TIPOLOGIA DI REATO: CORRUZIONE IN ATTI GIUDIZIARI
Fatti denunciati Persone denunciate
ABRUZZO 1 4
BASILICATA 2 4
CALABRIA 3 15
CAMPANIA 6 7
EMILIA R. 3 2
FRIULI V.G. 0 0
LAZIO 9 57
LIGURIA 2 1
LOMBARDIA 7 17
MARCHE 1 11
MOLISE 0 0
PIEMONTE 0 0
PUGLIA 2 53
SARDEGNA 0 0
SICILIA 3 5
TOSCANA 4 34
TRENTINO A.A. 0 0
UMBRIA 2 4
V. D'AOSTA 0 0
VENETO 3 3 TOTALE 48 217
Fonte : Sistema di Indagine del Ministero dell’Interno.
TABLE OF CONTENTS Judicial Measures
17
DENUNCE PER REATI CONTRO LA P.A., PER TIPOLOGIA DI REATO: ABUSO D’UFFICIO
Delitti registrati Persone denunciate
M F totale
2004 1.016 2.429 284 2.713
2005 1.051 2.542 342 2.884
2006 935 2.533 396 2.929
2007 1.097 2.810 460 3.270
2008 1.168 3.021 527 3.548
2009 1.099 2.698 489 3.187
2010 1.020 2.541 480 3.021
totale 7.386 18.574 2.949 21.429
Fonte : Sistema di Indagine del Ministero dell’Interno.
2006 2007 2008 M F Tot M F Tot M F Tot
Abuso d’ufNicio (art. 323 cp) • inizio azione penale 1.992 236 2.228 2.125 222 2.347 2.417 473 2.890 • condannate n.d. n.d. n.d. 94 20 114 82 4 86
Fonte: ISTAT, Rilevazione dei delitti denunciati per cui è iniziata l’azione penale e sui condannati con sentenza irrevocabile, ottobre 2010
TABLE OF CONTENTS Judicial Measures
18
IMPORTI DELLE CITAZIONI IN GIUDIZIO DELLE PROCURE REGIONALI, PER TIPOLOGIA DI EVENTO DANNOSO. ANNO 2008. € 57.607.164,42
€ 1.386.038.981,65
€ 36.859.370,20€ 139.297.932,34
€ 578.637,80 € 394.505,17
€ 69.013.083,11
€ 1,00
€ 10.000,00
€ 100.000.000,00
€ 1.000.000.000.000,00
PERS
ON
ALE
(Con
sule
nze)
MA
LAD
MIN
ISTR
ATI
ON
ALT
RE T
IPO
LOG
IE
ERO
GA
ZIO
NE
CON
TRIB
UTI
EFI
NA
NZI
AM
ENTI
-FR
OD
I CO
MU
NIT
ARI
E
INCI
DEN
TI
DA
NN
OA
LL'IM
MA
GIN
E
CORR
UZI
ON
E,TA
NG
ENTI
,CO
NCU
SSIO
NE
EDA
LTRI
REA
TI
Fonte : allegato V alla Relazione scritta del Procuratore generale, cerimonia di inaugurazione dell’anno giudiziario, Adunanza dell’11 febbraio 2009.
TABLE OF CONTENTS Judicial Measures
19
568
290246
113159
103 96130 111125 109
84
0
100
200
300
400
500
600
2004 2005 2006
Corruzione
Concussione
Istigazione alla Corruzione
Abuso d'Ufficio
601
1037
715
348 289 270
21 2 25
2302
2560
2029
0
500
1000
1500
2000
2500
3000
2004 2005 2006
Corruzione
Concussione
Istigazione alla Corruzione
Abuso d'Ufficio
Fonte : Dati Istat, pubblicati dall’Alto Commissario per la lotta alla corruzione e agli altri illeciti nella PA, I Mappatura della corruzione, 2007.
TABLE OF CONTENTS Corruption indexes based on direct
experience of the phenomenon
20
Over the last twenty years, due to the limitations encountered in the use of judicial statistics, scholars and professionals have attempted to develop alternative measures of corruption. To this respect, indexes based on the perception of corruption, widely used in scientific works and benefiting of consistent public attention, are of some importance. Two of the most relevant indexes are the Corruption Perception Index, annually released by Transparency International (TI-CPI), and the Rating of Control of Corruption, published by the World Bank (WB-RCC). They provide a measure of the actual perception of corruption in almost every country in the world, but without distinguishing between different regions within a country. Both indexes use aggregate data derived from surveys designed to measure corruption or phenomena believed to be related to it. These surveys, where corruption is not always uniquely and consistently defined, are typically constructed and carried off by consulting firms and rely on interviewing experts of different fields, national and foreign businesspeople and other citizens. Each index use a peculiar technique to combine appropriately available data for each country, in order to obtain a single summary measure.
TABLE OF CONTENTS Corruption indexes based on direct
experience of the phenomenon
21
Not all surveys that contribute to each index are available for all countries. When computing TI-CPI, at least three sources are expected to be available for the index to be released, while WB-RCC can be computed when there is at least one source for any given country. That is also why WB-RCC index is available for more countries than TI-CPI, for every year of publication. Of course, the greater the number of surveys combined, the more precise we expect the resulting index to be. However, a fact has to be taken carefully into account: each of the indicators that contribute to the computation of the Corruption Perception Index has a measurement error and, whether the measurement errors of the different indicators are not independent, the use of a composite index (such as TI-CPI or WB-RCC) does not generate a significantly more accurate measure than the indicators used to compute it in the first place. Also, the indexes of corruption perception present various drawbacks. First, the actual reliability of the information provided by interviewed people is unknown. On the one hand, in fact, those who are directly implicated in corruption practices, despite the anonimity guaranteed to them, have an incentive to minimize the phenomenon. On the other hand, those who are not directly involved may provide inaccurate information. Moreover, being both TI-CPI and WB-RCC sort of averages of different measures, their reliability tends to be undermined when describing countries with poor or no data available, precisely the countries we expect to suffer more from corruption and corruption-related phenomena .
TABLE OF CONTENTS Corruption indexes based on direct
experience of the phenomenon
22
A few details: TI-CPI is available on an annual basis since 1995, and is computed using data from 13 different surveys and evaluations, from 10 independent organizations . Over the years, the coverage of TI-CPI has been greatly increased: from 54 countries in 1995 to 180 in 2009. The index takes value between 0 and 10, the lower values indicating a higher level of corruption. WB-RCC index, on the other hand, was published every two years from 1996 to 2002 and on an annual basis since then. The number of countries included, has been extended from 152 in 1996 to 215 in 2008 Both TI-CPI and WB-RCC compute the index of perceived corruption for each country, combining different information in different ways. In particular, TI-CPI compute a simple average of the basic measures, while WB-RCC uses a model with “non-observed components”. The methodological differences between TI-CPI and WB-RCC have a limited impact on final results, though. Values are very similar, apart from differences in the used scale. Being the two indexes very similar, their correlation (a measure of the degree of connection between two variables) is always very close to 1, the maximum For both indexes a measure of accuracy, computed on the basis of dispersion, the “average difference from the mean”, is also provided.
TABLE OF CONTENTS Corruption indexes based on direct
experience of the phenomenon
23
TABLE OF CONTENTS Corruption indexes based on direct
experience of the phenomenon
24
It is inadvisable to be over-confident in changes over time of the indexes of perceived corruption. The authors of TI-CPI themselves explicitly indicate that “given its methodology, the CPI is not a tool that is suitable for monitoring progress or lack of progress over time”, because “year-to-year changes in a country/territory’s score could result from a changed perception of a country’s performance, a change in the ranking provided by original sources or a change in the CPI’s methodology.” . This note of caution generates some doubts about the usefulness of TI-CPI for any type of comparison, even if the caution of Transparency International is to be attributed more to political than to scientific reasons. In fact, a punctual analysis of the TI-CPI index, in terms of its variation over time, could be easily considered as an actual evaluation of national governments policies and result in sharp unwanted criticism towards the German organization.
TABLE OF CONTENTS Corruption indexes based on direct
experience of the phenomenon
25
TABLE OF CONTENTS Corruption indexes based on direct
experience of the phenomenon
26
Country Forecast July 2010 © The Economist Intelligence Unit Limited 2010
Italy: Indicator scores in the business rankings model
a Out of 18 countries: Austria, Belgium, Cyprus, Denmark, Finland, France, Germany, Greece, Ireland, Italy, Netherlands, Norway, Portugal, Spain, Sweden, Switzerland, Turkey and the UK.
Note. A single asterisk (*) denotes scores based on quantitative indicators. Indicators with a double asterisk (**) are partly based on data. All other indicators are qualitative in nature.
ItalyRegionalaveragea Italy
Regionalaveragea
2005-09 2010-14
Political environment
1. Risk of armed conflict 5 4.4 5 4.5
2. Risk of social unrest 4 4.3 4 3.7
3. Constitutional mechanisms for the orderly transfer of power 5 4.8 5 4.8
4. Government and opposition 4 4.4 3 4.3
5. Threat of politically motivated violence 4 4.1 4 4.1
6. International disputes or tensions 5 4.3 5 4.3
7. Government policy towards business 3 3.9 3 3.8
8. Effectiveness of political system in policy formulation and execution 3 3.8 2 3.4
9. Quality of the bureaucracy 3 3.7 3 3.7
10. Transparency and fairness of legal system 3 4.0 3 4.0
11. Efficiency of legal system 2 4.2 2 4.2
12. Corruption 3 4.1 3 4.1
13. Impact of crime 2 4.1 2 4.1
Country Forecast July 2010 © The Economist Intelligence Unit Limited 2010
Italy:
Country Forecast July 2010
Italy
Editor: Robert O'Daly
Editorial closing date: 22nd July 2010
TABLE OF CONTENTS Corruption indexes based on direct
experience of the phenomenon
27
TABLE OF CONTENTS Corruption indexes based on direct
experience of the phenomenon
28
TABLE OF CONTENTS Corruption indexes based on direct
experience of the phenomenon
29
Global Integrity Report
www.report.globalintegrity.org/Italy/2008).
TABLE OF CONTENTS Corruption indexes based on direct
experience of the phenomenon
30
TABLE OF CONTENTS Corruption indexes based on direct
experience of the phenomenon
31
CITIZENS’ PERCEPTIONS OF FRAUD AND THE FIGHT AGAINST FRAUD IN THE EU 27
TABLE OF CONTENTS Corruption indexes based on direct
experience of the phenomenon
32
TABLE OF CONTENTS Corruption indexes based on direct
experience of the phenomenon
33
TABLE OF CONTENTS Corruption indexes based on direct
experience of the phenomenon
34
TABLE OF CONTENTS Objective measures of corruption
35
The indexes of corruption described so far suffer from numerous limitations. Therefore, a search for objective measures, based on facts rather than on people’s perception or answers, is not a useless quest. Until now, attempts in this direction have showed a non-systematic feature and have always been tied to the realization of specific analyses, mainly in the academia. For these reasons, the following considerations have a somewhat speculative nature. The main obstacle in building objective measures of corruption derives from both methodological problems and the difficulty encountered in finding the necessary data. However, with respect to the availability of data, we can expect that many data on public policies will be be in the near future directly available in digital format and, at least in principle, more easily accessible to researchers. More on this topic later on.
TABLE OF CONTENTS Objective measures of corruption
36
Someone propose a corruption index based on two distinct ideas of “public capital”. The first one is about measuring public capital by means of physical inventory – kilometers of roads, square meters of public buildings and so on. This measure is clearly a rough estimation of the actual public capital and also suffers from the incongruity of aggregating data that relate to different types of infrastructures. An alternative method consists in estimating how much has been spent over the years for public capital’s accumulation. This method, called of the “permanent inventory”, relies basically on adding the past and present investment flows, until the good’s average life is reached, for each type of infrastructure. For both the estimation based on the physical inventory and the permanent inventory method, it is possible to construct succint indexes in order to take into account the differences in population and size of each region. In particular, measures of network infrastructures, such as roads and railways, are divided by area, while infrastructures such as public buildings, schools and hospitals are divided by resident population. At the end of all these computations, each measure of infrastructure level is expressed with a percentage of the national average, made equal to 100.
TABLE OF CONTENTS Objective measures of corruption
37
Source: Golden e Picci (2005).
TABLE OF CONTENTS Objective measures of corruption
38
If we consider the measure computed with the permanent inventory technique, on the other hand, Lombardy appears to have spent only 9% more for its infrastructures than the national average. The difference would indicate a relative historical efficiency of Lombardy in transforming resources into finished works, compared to the national average. Certainly, the observed difference between the two indexes could be explained by economic considerations. For example, Lombardy could be enjoying lower costs of raw materials or labour, or benefiting from more favorable geographic characteristics than the rest of the country. Investigated this hypotheses, using for example the degree of seismic activity in the area, and constructed a more accurate measure of public capital, shown in column 3 of Table 2. The correction made does not lead to any significant change in the data, though. The last column of the table reports the ratio between the index of infrastructure based on physical capital and the index computed with the technique of permanent inventory, corrected for cost differentials. The differences are remarkable among regions and only for two Northern regions, Liguria and Val D’Aosta, characterized by mountainous land, the index is below the national average.
TABLE OF CONTENTS Objective measures of corruption
39
Such large differences can be explained by several factors. First, of course, by the presence of corruption-related phenomena, but also, more generally, by the existence of structural policy problems making the use of public resources systematically less effective in some areas. It is difficult to distinguish between these two aspects – poor governance is likely to be both the cause and the effect of high corruption – however, although with some caution, they also suggest to interpret the index of column 4 as a measure of the corruption level. This index has the advantage of relying on objective data, but it can be criticized for several reasons. First, if it is true that inefficiency and corruption are inter-related phenomena, it would be desirable to be able to distinguish between them. This is not allowed by the proposed index. Moreover, the index is computed for a given point in time while corruption is typically the result of a long-term process. Finally, its use shall be limited to areas providing solid, reliable data, which is not often the case.
TABLE OF CONTENTS Objective measures of corruption
40
The idea of building measures that are clearly disconnected from questionable individual perceptions, however, has intrigued other scholars and, in particular, the analysis of prices has been regarded with some interest. For example, someone measures corruption in a group of villages in Indonesia by observing the differences between the official prices of various projects and an estimation of the costs provided by independent engineers. Someone consider the gap between prices of some basic factors of production, purchased by hospitals in Buenos Aires, and prices indicated by a national investigation against corruption. Someone take into account the prices of standardized goods, purchased by the Italian government through Consip, thus succeeding in distinguishing corruption or “active waste” from simple inefficiency or incompetence.
TABLE OF CONTENTS Objective measures of corruption
41
An important feature of the “objective” approach, as we have already mentioned, regards the difficulty to collect reliable data. The increasing use of what insiders call the “administrative statistics”, coming directly from public bureaus, is likely assuage the problem in the future. Developments in computer engineering and science would also help, allowing computation of always more accurate indexes in an easier and quicker way. Always on data availability, it is important to draw a distinction between available but only human readable data and data which, thanks to a suitable structure and to the presence of accompanying information (“meta-information”), are also machine readable. The presence of high-quality machine readable data would permit to a variety of social actors, including private citizens and associations, to use them at a reduced cost and to automatically compute various indexes relating to public policies, some of these designed to measure corruption and identify situations that give the suspicion of corruption. This would certainly raise the awareness of the public and the media and allow the public administration itself to better focus its inspection activities.
42
Thank You for your kindly attention
TABLE OF CONTENTS Measures of corruption