Show Summary Details

Page of

date: 24 September 2017

The Economics of Oil Spills

Summary and Keywords

Vessel oil spills are very serious natural hazards that have affected coasts worldwide for many decades. Although oil spills from tankers are highly publicized, very little is known about the role played by the incentives and regulatory instruments in place to prevent them. In order to shed some light on these issues, data were collected worldwide on large oil spills from multiple databases, starting in the 1970s, and merged with other socioeconomic records. A crucial concern is that that large oil spills have been undercompensated over time with respect to the damages caused. A meta-analysis was estimated in order to assess relevant factors affecting the damage claimed in oil spills and the compensations received by the affected parties. Meta-regression results show that the legislation applied (strict unlimited liability versus limited liability) played a crucial role in both the amount claimed and the final compensation received. Also, time-trend variables are shown as determining factors for both the damages and claims that are finally paid. To correct the large gap between damage claimed and compensation scenarios, it is recommended to strengthen compensation funds, while carrying out more comprehensive assessment studies which apply valuation methods comparable with those proposed by green capital initiatives for marine ecosystem services, and which could be used successfully during the litigation process.

Introduction

Oil spills can be caused by vessels, industrial facilities, or oil platforms, or in other ways. Due to the lack of data related to many of these incidents, and in order to conduct a systematic review of the causes and consequences of the oil spills that have occurred since the 1970s, the focus of this chapter is on vessel oil spills. These kinds of incidents cause a major impact on the environment and the economy of the affected areas (Alló & Loureiro, 2013).

According to data from the International Tanker Owners Pollution Federation Limited (ITOPF, 2016),1 the total number of oil spills has decreased worldwide since the 1970s. The ITOPF database contains records on spills over this time period, and has found a significant decrease in both the number of accidents and the total amount spilled over the years (Alló & Loureiro, 2013). However, in spite of fewer spills being recorded recently, these are much larger in terms of the amount of oil spilt. Since 2000, a total of 223 vessel oil spills have been registered worldwide, resulting in 229,000 tons of oil being lost. Specifically, and with recent data, the ITOPF (2016) highlights that during the period from 2010 to 2015, 42 accidents were recorded with a spillage of around 33,000 tons, while emphasizing that 86% of total oil spilt was caused by only 10 accidents. The seriousness of the damage caused by oil spills has been highlighted by various impact assessment studies conducted after these catastrophic events, such as those from the Exxon Valdez (Carson et al., 1992), Erika (Bonnieux & Rainelli, 1993), Sea Empress (Moore et al., 1998), American Trader (Chapman & Hanemann, 2001), and Prestige (Loureiro, Ribas, López, & Ojea, 2006), among others.

The present chapter illustrates several key findings related to the economics of vessel oil spills since the 1970s. First, it provides a historical overview of the most important changes that have taken place in terms of compensation bodies and compensatory regulations worldwide. Second, via a meta-analysis, it assesses the causing factors of vessel oil spills, in order to understand the damage caused by these incidents. Third, it investigates the compensation received by the affected parties. It concludes with a number of recommendations in order to correct for differences in terms of compensations and damage claims.

Several key findings emerge from the review of vessel oil spills from the 1970s to the present. First, a review of the historical data reveal major discrepancies between damage claims and the compensation finally awarded. For the largest oil spills, in some cases the final compensation paid was nearly 10 times lower than the damage claimed. Historically, the revision and subsequent updating of compensation limits have usually come about as a result of political concerns and public awareness after very large oil spills, the majority of which were not properly covered by the existing compensation funds. Second, meta-regression results show that the legislation applied (strict unlimited liability versus limited liability) played a crucial role in terms of affecting both the amount claimed and the final compensation received. Third, time-trend variables are determining factors in both the damages and claims finally paid. As a future recommendation, and in order to correct the large gap between damage and compensation scenarios, the need to strengthen compensation funds is emphasized, while carrying out more comprehensive assessment studies which apply valuation methods comparable with those proposed by green capital initiatives for marine ecosystem services, and which could be used successfully during the litigation process.

The structure of this chapter is as follows: first, the data used are described in the following analysis, with a special section dedicated to damages and compensation; next, from a historical perspective, the legislation applied and compensation funds worldwide are disentangled; then the literature that discusses factors that cause oil spills is reviewed; next, a description of the empirical model in the meta-analysis is presented, followed by the results section. The chapter ends with a series of concluding remarks and recommendations.

Data Description

In order to collect data about vessel oil spills, a number of different databases were reviewed, including the French Centre de Documentation de Recherche et d’Expérimentations sur les Pollutions accidentelles des eaux (Cedre, 2016), the Damage Assessment, Remediation & Restoration Program (DARRP, 2016) database, and The International Tanker Owners Pollution Federation Limited (ITOPF, 2016). This latter database provides information on incidents that have taken place in International Marine Organization (IMO) member countries, providing data on claims filed by countries and individuals, as well as the final compensation paid by the ITOPF. In addition, information about the technical characteristics of the ships, which is highly relevant in order to assess the damage, was collected from the Center for Tank Ship Excellence (CTX-4). Information was recorded about the date and the place of the incident, the tons spilled, whether the vessels had signed flags of convenience (FOC), whether the ship had a single or double hull, and the cause of the accidents. We also registered the economic damages claimed and the compensation finally paid after each oil spill. All of the previous datasets were merged, creating a final database that contains information about oil spill incidents around the world from 1968 to 2013. In addition to the aforementioned variables, collected information included the characteristics of the country where the accident took place. Specifically, we identified whether the country has a strict regulation (based on strict liability), and the income level expressed by the Gross Domestic Product (GDP) in Power Purchasing Parity (from the World Bank, 2011). An earlier version of this dataset was used by Alló and Loureiro (2013).

It is also important to note that some of the observations collected were removed from the final dataset, due to a large amount of missing values in several explanatory variables. As a result, the final data used for this meta-analysis cover 114 oil spills worldwide.2 The information collected reports the damage caused and the compensation paid in different currencies ($, €, JPY, etc.) and different time periods (from 1968 to 2013 in the case of the damages, and from 1975 to 2007 in the case of the economic compensations). In order to homogenize all of the information, monetary amounts were converted to a standard measure ($2010). For this reason, the total damages of each accident in the year of the spill were inflated to their original currency at 2010 prices,3 transforming these amounts to $2010.4 With respect to the data that have been analyzed for the compensation finally paid, it was found that the mean economic compensation from an oil spill is around$106 million, and the mean quantity of oil spilled is around 8,975 tons. Around 85% of vessels had a single hull and about 86% were sailing under an FOC. In terms of the cause of the accident, 18.5% were related to grounding, 5.5% included various sources of error, 29.6% were collisions, 3.7% were explosions (in some cases the explosion could have happened after the accident), and 42% were due to other diverse causes (including loading, storm damage, or swamping). Table 1 presents a summary about the causes of accidents according to the size of the oil spill from the ITOPF (2016) oil tanker spill statistics. In terms of the damage claimed, around 22% of oil spills caused considerable impacts to the environment, 46% caused damage to fisheries, 20% caused damage to the tourism sector, and 26% caused damage in other areas.

Table 1 Accidents’ Causes According to the Size of the Oil Spill

Number of accidents causing spills of

Accident type

< 7 tons

7–700 tons

> 700 tons

Allision/collision

188

361

136

Grounding

240

270

150

Hull failure

577

101

60

Equipment failure

1692

207

18

Fire/explosion

174

47

52

Other (heavy weather damage, human error)

1815

175

30

Unknown

3188

203

13

Source: ITOPF (2016). Oil Tanker Spill Statistic. The data for the oil spills<7 tons covers the period 1974–2015, while for accidents of 7 more tons spilled the data covered is from 1970 to 2015.

Damage Claims and Compensation: A Historical Overview

Table 2 provides a historical overview of damages and compensation for all incidents, reporting both magnitudes. These are reported in 2010 prices. The average damages claimed per spill during this period amount to $231 million, while the compensation awarded stands at$107 million per accident. The discrepancies between both magnitudes are quite significant for the largest oil spills occurring in IMO country members. For example, in the Haven oil spill, total damages were claimed for $1604 million, while final compensation was granted for$178.97 million. For the Aegean Sea and Hebei Spirit spills, damages were claimed for $370.84 million and$362.55 million, respectively, while the final compensations were $90.68 million and$178.97 million, respectively. Several reasons justify these differences, but the most relevant are related to the compensation limits established by the different compensation funds ratified by the IMO countries. In addition, at times the lack of relevant assessments providing a clear link to the cause-effect relationship between the occurrence of the spill and the damages claimed prevented compensation from being paid. Due to the relevance of the legal environment on the economics of oil spills, the “Oil Spill Compensation Funds” section examines the main compensation mechanisms that exist, and how they have evolved during the study period.

Table 2. Damages Claimed and Final Compensation Received

Name of the ship

Year of spill

Claims (Damages)

Compensation

Year of payment

($2010 million) ($2010 million)

Aegean Sea

1992

370.84

90.68

2002

Agip Abruzzo

1991

37.2

18.69

1995

Akari

1987

2.45

0.18

1992

Al Jaziah 1

2000

3.67

2.03

2010

Amazon Venture

1986

5.97

2.39

1989

Amazzone

1988

6.47

5

1992

Antonio Gramsci (second time)

1979

8.72

0.65

1980

Apex/Galveston Bay/Shinoussa

1990

2.5

1.91

1994

Baltic Carrier

2001

12.07

9.31

2003

1986

3.26

2.24

1988

Cape Mohican

1996

6.82

4.53

1996

Dainichi Maru n5

1989

0.08

0.04

1991

Daito Maru no. 3

1990

0.1

0.07

1992

Diamond Grace

1997

23.11

6.75

1999

Erika

1999

239.36

240.9

2013

Exxon Valdez

1989

4262.06

4484.14

2004

Haven

1991

1604

178.97

1999

Hebei Spirit

2007

362.55

25.45

2013

Hinode Maru no. 1

1987

0.03

0.02

1989

Honam Sapphire

1995

82.04

9.31

1999

Hosei Maru

1980

2.73

2.05

1982

Jan

1985

3.67

1.8

1988

Katja

1997

0.22

0.22

2008

Kazuei Maru no. 10

1990

0.63

0.6

1991

Kifuku Maru no. 35

1982

0.05

0.01

1983

Koei Maru no. 3

1983

0.4

0.33

1986

Koshum Maru no. 1

1985

0.35

0.31

1990

Kriti Sea

1996

6.42

6.42

2009

Kyung Won

2003

5.98

2.95

2004

Maritza Sayalero

1998

28.2

5.23

2001

Mebaruzaki Maru no. 5

1979

0.12

0.1

1981

Nakhodka

1997

373.18

221.9

2002

Nissos Amorgos

1997

262.12

96.06

2013

Ondina

1981

18.8

7.89

1984

Pacific Colocotronis

1975

0.7

0.7

1975

Patmos

1985

125.56

6.96

1994

Portfield

1990

0.87

0.44

1995

Prestige

2002

5103

127.68

2003

Rio Orinoco

1990

32.9

19.67

1995

Rosebay

1990

0.07

0.07

-

Sea Empress

1996

63.15

27.22

2003

Sea Prince

1995

257.26

18.17

2003

Shinkai Maru no. 3

1983

0.04

0.02

1984

Shiota Maru no. 2

1982

1.05

0.86

1982

Showa Maru

1980

0.59

1.24

1981

Taiyo Maru no. 13

1988

0.11

0.08

1989

Take Maru no. 6

1986

0.04

0

1987

Tanio

1980

218.12

54.25

1988

Tarpenbek

1979

6.33

1.4

1986

Thuntak 5

1986

6.86

3.66

1992

Tolmiros

1987

24.62

0.11

1992

Tsubame Maru no. 58

1989

0.27

0.24

1991

Volgoneft263

1990

4.48

4.11

1992

Sources: Cedre (2016); ITOPF (2016); DARRP (2016).

Oil Spill Compensation Funds

In terms of regulating oil spills, there are currently two quite different regimes: one is a common international regime applied by all IMO country members, and the other is the strict liability regime applied by the United States and regulated by the Oil Pollution Act enacted in 1990 (OPA-90). Most other countries in the world apply the international IMO regime. The IMO is a specialized agency of the United Nations which deals with marine affairs and pollution. This organization was set up in 1982 and has drawn up numerous Conventions and restrictions in order to increase marine safety and reduce pollution. However, the IMO has no enforcement capacity over signatory States of the Protocols, and the safety standards are voluntary for each nation. Since 1992, two Protocols, known as the 1992 Convention on Civil Liability for Oil Spill Damage (1992 CLC) and the International Fund for Compensation for Oil Pollution Damage (1992 Fund Convention), have been applied.

The international regime emerged after the Torrey Canyon oil spill in 1967, while the US regime was established after the Exxon Valdez oil spill in 1989. Both dramatic incidents confirmed the need for legislation on liability and financial responsibility in oil spill accidents and on compensation regimes (Kim, 2003). However, the US regime imposes unlimited liability and compensates damage to natural resources, while the international IMO regime imposes a system of limited liability with similar characteristics to a negligence system, and excludes compensation for environmental damage.

The overall international compensation regime for damage caused by spills of persistent oil from laden tankers was based initially on two IMO Conventions: (1) the 1969 International Convention on Civil Liability for Oil Spill Damage (1969 CLC), and (2) the 1971 International Convention on the Establishment of an International Fund for Compensation for Oil Pollution Damage (1971 Fund Convention). These two Conventions were amended in 1992 by two Protocols which increased the compensation limits and widened the scope of the original Conventions. The 1969 CLC came into force in 1975 and laid down the principle of strict liability (i.e., liability even when no fault is proven) for tanker owners, and creates a system of compulsory liability insurance. However, this strict liability was linked to the tonnage of the tanker causing the pollution (IMO, 2006). The 1971 Fund Convention provided for the payment of supplementary funds to those who could not obtain full payment from the International Oil Pollution Compensation Fund (1971 IOPC Fund). This 1971 IOPC Fund Convention came into force in 1978.

In 1992, a diplomatic conference adopted two Protocols amending the 1969 CLC and 1971 Fund Convention. The 1992 Fund came into force in 1996. As in the case of the original Conventions, the tanker owner and the principal and interest (P&I) insurer are liable for payment of compensation under the 1992 CLC and oil receivers in countries that are party to the 1992 Fund Convention. The 1971 Fund Convention was terminated completely on May 24, 2002. In October 2000, the contracting States to the 1992 CLC and 1992 Fund Convention approved a proposal to increase the amount of compensation available under the terms of these Conventions by about 50% (to around US$260 million). This Fund, which came into force on November 1, 2003, was reinforced by other additional Supplementary Compensation Funds that increase the international regime for the compensation of victims of oil pollution from oil tankers. This Supplementary Fund does not replace the existing 1992 Fund, but will make additional compensation available to victims in the Convention States. In total, about$1,158.5 million ($2005) are available for compensation for each incident in the States which are Members of the Supplementary Fund. This Supplementary Fund covers accidents that occur after the entry into force of the Protocol on March 3, 2005. Figure 1 shows the limits of compensation funds over time. Although the deterrence effects associated with monitoring and liability regimes were studied on several occasions in the economics literature (see, e.g., Opaluch & Grigalunas, 1984; Wayne & Kite Powell, 2005), to our knowledge there is no evidence of whether these international Conventions (which involve different levels of compensation based on the strict liability of ship owners) have a deterrence effect on oil spill occurrence. In the following empirical model , the effects of the unlimited liability regime versus the limited liability IMO regime are explicitly considered. Click to view larger Figure 1. Compensation limits. Although we acknowledge that oil spills are seriously underreported internationally, due to diverse errors and omissions, (particularly for spills in remote areas and minor spills), in recent years, the majority of the largest oil spills occurred in European waters. In particular, there were three large oil spills in European waters, including the Erika oil spill on the French coast (1999), the Baltic Carrier in Denmark (2001), and the Prestige oil spill in Spain (2002). The following analysis assesses whether the strict liability had any deterrence effect on the occurrence of oil spills. The answer to this question is of vital importance, since if IMO international Protocols do not have a deterring effect in reducing oil spills, then additional or alternative measures should be considered by IMO country members. The approval of the OPA-90 has strengthened accountability for vessel oil spills occurring in US waters. Moreover, the OPA-90 has increased penalties for polluters, broadened the response of the authorities, and preserved State authority to establish laws governing oil spill prevention and response (US Environmental Protection Agency, 2011). Table 3 presents a summary of the different Conventions and dates of the various regulatory changes behind these international limited liability Protocols in place in all IMO countries. Table 3. OPA-90 and IMO Regimes OPA-90 IMO Type of regime Strict liability Common international regime Where is it applied? USA 110 countries Since when has it been applied? 1989 1967 Type of liability Unlimited Limited to the tonnage of the ship Recognized damage Economic and environmental damage Economic and only regeneration damage History of the regimes OPA-90 1969 CLC 1971 Fund Convention (1971) 1992 CLC 1992 Fund Convention (1992) Supplementary Fund Convention (2005) Maximum compensation$1000 million per incident + unlimited amount

1971 Fund Convention

1987–2002: $91.4 million 1992 Fund Convention 1996–Nov. 1, 2003:$205.6 million

Nov. 1, 03–2017: $309.1 million Supplementary Fund Convention (2005) After March 1, 2005:$1159 million

Thébaud, Bailly, Hay, and Pérez (2005) have shown that there is a significant difference between the amounts claimed as damages and the payments received as compensation from the CLC funds. In particular, the maximum amount of compensation paid by the 1992 Fund Convention with respect to an incident that occurred prior to November 1, 2003 was set at US$205.6 million, which increased to US$309.1 million in November 2003.5 However, due to a shortage of funds for compensation, and in the wake of the multiple complaints following the Prestige oil spill in Spain in November 2002, this Fund has been reinforced by an additional Supplementary Compensation Fund that increases the total compensation of victims of oil pollution from oil tankers to US$1,159 million. This Supplementary Fund covers accidents occurring after the entry into force of the Protocol on March 3, 2005 (ITOPCF, 2011). In the following analysis, we assess the marginal contribution of the various causing factors of accidental oil spills to the total damage. Causing Factors of Oil Spills White and Molloy (2003) reviewed the factors affecting the costs of oil spills in a database containing more than 300 cases. They show that in addition to the type of oil and the characteristics of the location where the spill takes place, other factors, such as the amount spilled, the weather and sea conditions, as well as the time of the year, are crucial determining factors of the costs the oil spill. In terms of the causes, Ogus (1999) found that when an accident happens due to collision or grounding, the spill size is on average smaller. In terms of ship characteristics, Alló and Loureiro (2013) found that single-hull ships cause larger oil spills than double-hull ships. Furthermore, the effect of the flag of convenience (FOC) is another aspect of interest. An OECD report states of many of these countries allowing FOC have generally been less rigorous in their surveillance of high safety standards (OECD, 2001). Also, Levantion (1982) stated that vessels sailing under an FOC are responsible for most oil pollution in the high seas, posing a major threat to the environment. In terms of surveillance efforts, Goodstein (1992) observed that certain types of spills happen much more frequently over the weekend (on Saturdays) due to the lack of surveillance. The explanatory variables for the following empirical results were selected following this review of the literature. Table 4 shows the description of the variables contained in this meta-dataset and their respective summary statistics, for each of the independent regressions, modeling (1) the damages claimed and (2) the compensation received. Model Specification and Research Hypotheses The aim of this section is to bring together information on the factors affecting both the damage claimed in vessel spills, and the corresponding compensation paid in oil spills worldwide. In doing so, we focus our attention on the role of legislation, which has proven to be crucial, as well as other structural characteristics of vessels, while including the weather and time-related variables. Furthermore, we control for institutional performance, including the GDP per capita. In order to assess the effects of these explanatory factors, we model damage claimed in each incident and the compensation received as a function of the countries’ characteristics, the ship’s characteristics, the characteristics of the accident, and the liability regime applied, among other variables. First, we estimate a baseline function model through an ordinary least squares (OLS) meta-regression, and a Tobit model in order to check the robustness of the results. We run two independent regressions, the first modeling the damage claimed in each accident and the second assessing the total compensation paid. Dependent variables are labeled as ($log(Y)$), while the vectors of explanatory variables ($Xi$) (see Table 4 for description variables) are grouped into five categories, including the country’s characteristics $Xc$; the circumstances of the accident $Xa$; the vessel’s characteristics $Xv$; the claims presented $Xc$; and whether the country where the accident occurred has strict liability in place $Xli$. The baseline meta-regression OLS models correspond with the estimation of the following equation: $Display mathematics$ (1) where $β0$ is the constant term, the corresponding $βj$ are the coefficients associated with the explanatory variables to be estimated, and $εi$ is a vector of error term, independently and identically distributed (i.i.d). In order to correct for spatial heteroskedasticity, we estimated the OLS model with the Huber-White adjusted standard errors clustered for each country where the accident took place. Previous meta-analyses have adopted a similar approach (Alló & Loureiro, 2013; Barrio & Loureiro, 2010). Table 4. Summary Statistics for the Damages and Compensation Regressions Damage Compensation Variable Mean SD. Mean SD. Log (dependent) Logarithm of the dependent variable (expressed in$2010)

15.322

2.637

14.541

2.912

Characteristics of the country

LogGDP

Logarithm of the GDP (ppp1)

9.931

0.461

9.926

0.328

Characteristics of the accident

North America

1, if the accident took place in North America; 0 otherwise

0.254

0.437

0.092

0.292

Europe

1, if the accident took place in Europe; 0 otherwise

0.342

0.477

0.425

0.499

Asia

1, if the accident took place in Asia; 0 otherwise

0.342

0.477

0.444

0.501

South America

1, if the accident took place in South America; 0 otherwise

0.053

0.224

0.037

0.190

Before80

1, if accident occurred before the 80s; otherwise 0

0.114

0.319

0.055

0.231

1, if accident occurred in the 80s; otherwise 0

0.386

0.489

0.444

0.501

1, if accident occurred in the 90s; otherwise 0

0.325

0.470

0.388

0.492

1, if accident occurred in the 00s; otherwise 0

0.149

0.357

0.111

0.317

1, if accident occurred after the 00s; otherwise 0

0.026

0.161

-

-

Log oil spill

Logarithm of the tons discharged

6.438

2.886

5.940

3.082

Saturday

1, if the accident occurred on a Saturday; otherwise 0

0.166

0.374

0.074

0.264

Winter

1, if the accident occurred during the winter; otherwise 0

0.193

0.396

0.203

0.406

Characteristics of the ship

Single hull

1, if the ship is single hull; 0 otherwise

0.851

0.358

0.851

0.358

Flag

1, if the ship sailed with a flag of convenience; 0 otherwise

0.833

0.374

0.867

0.341

Characteristics of the claims

Environmental

1, if claims for environmental damage were filed; 0 otherwise

0.210

0.409

0.222

0.419

Tourism

1, if claims for tourism damages were filed; 0 otherwise

0.149

0.358

0.203

0.406

Fisheries

1, if claims for fisheries damages were filed; 0 otherwise

0.351

0.479

0.463

0.503

Other damage

1, if claims for other damages were filed; 0 otherwise

0.254

0.437

0.259

0.442

Grounding

1, if the cause of the accident was grounding; otherwise 0

0.289

0.455

0.185

0.392

Error

1, if the cause of the accident was an error; otherwise 0

0.088

0.284

0.055

0.231

Explosion

1, if the cause of the accident was explosion; otherwise 0

0.053

0.224

0.037

0.191

Collision

1, if the cause of the accident was collision; otherwise 0

0.254

0.437

0.296

0.460

Other cause

1, if the cause of the accident was not grounding, sinking, explosion, collision or an error; 0 otherwise

0.316

0.467

0.426

0.499

Strict liability

Strict liability

1, for the US from 1991; otherwise 0

0.088

0.284

0.018

0.136

(1) PPP is referred to the Power Purchase Parity

In addition, we tested the presence of multicollinearity through the variance inflation factor (VIF). This indicator measures the proportion of multicollinearity of the jth explanatory variable with the rest of explanatory variables in a regression model (O’Brien, 2007); an average value of 1.81 and 2.39 was obtained, and it was concluded that multicollinearity is not present in our empirical specification.

In order to check the robustness of the results, we estimated a Tobit model with the same specification. In this setting, we established a left limit equal to the smallest accident in terms of the economic damage, while the same criteria are used for modeling the economic compensation.

Research Hypotheses

As stated, our aim is to test whether the application of the strict liability regime is linked with a negative impact on the damage caused and with a positive impact on the amount of compensation paid after an accident. Previous studies, such as those of Alberini and Austin (1999, 2002), or more recently Bernard (2011) and Mondello (2011), have referred to the same idea. The logic of this argument is that under this regime, polluters are strictly liable for all the damage suffered as a consequence of their activity. Therefore, it is expected that they work with higher levels of care. Furthermore, the establishment of unlimited liability is usually accompanied by a more aggressive negotiation strategy, given the odds to obtain higher compensations derived from additional litigation efforts.

In order to assess these relationships, we test whether the coefficient representing the strict liability regime has a negative impact on the amount of damage claimed, and a positive impact on the amount of compensation paid:

Impact on the amount of damage claimed:

$Display mathematics$
(2)

Impact on the amount of compensations paid:

$Display mathematics$
(3)

Empirical Results

Main Results from Damage Claimed Regression

The meta-analytical results confirm the deterrent effect of the application of strict liability, reducing the resulting damage claimed by oil spills in a statistically significant way (Table 5). Furthermore, and as expected from previous literature, the amount spilled also has a positive and significant effect on damage claims (in both OLS and Tobit models). Also, if environmental claims have been accounted for, this significantly increases the damages being claimed. Environmental damage and damage in the tourism sector also seem to significantly increase the amount claimed (in the OLS model). Moreover, in the Tobit model, we find that damage claims in fisheries also increase the total amount of damages claimed from these accidents. Time effects also seem to be relevant, and even though the number of oil spills has decreased in recent years, the amount of damage caused seems to be on the rise (particularly for the decades of the 1980s and 2000s in the case of the OLS model, and during the 2010s attending to the Tobit results). This may be a consequence of the fact that although the number of accidents has decreased, the accidents that do occur these days are larger. Also, the accidents that occur in the winter season are related to less damage. In the case of the Tobit model, we also obtain a negative coefficient for the GDP variable, which may indicate that countries with a higher GDP suffer less damage from oil spills than others. This may be a consequence of the fact that poorer countries have fewer resources and tools to fight against these disasters, and therefore the losses suffered are greater. Furthermore, wealthier countries are expected to have a better institutional performance that may allow them to be more efficient. Other structural ship variables, such as whether they had a single hull and were sailing under an FOC, are not statistically significant.

Table 5. OLS and Tobit Results for the Damages and Compensation

Damages

Compensation

OLS

Tobit

OLS

Tobit

Dependent variable

Coef.

Robust

P>|t|

Coef.

Robust

P>|t|

Coef.

Robust

P>|t|

Coef.

Robust

P>|t|

SE

SE

SE

SE

Characteristics of the country

LogGDP

-0.850

0.500

0.101

-0.850

0.456

0.065

-0.530

0.656

0.433

-0.557

0.540

0.309

Characteristics of the accident

North America

0.394

0.517

0.453

0.394

0.472

0.406

1.984

0.643

0.008

2.076

0.525

0.000

Europe

0.101

0.590

0.865

0.101

0.538

0.852

0.461

0.623

0.471

0.477

0.497

0.344

1.326

0.757

0.091

1.326

0.690

0.058

0.784

0.765

0.322

0.785

0.645

0.231

1.211

0.874

0.177

1.211

0.797

0.132

0.887

0.916

0.349

0.891

0.715

0.221

1.861

0.855

0.038

1.861

0.780

0.019

1.151

0.781

0.163

1.162

0.625

0.071

2.054

1.333

0.135

2.054

1.216

0.094

-

-

-

-

-

-

Log oil spill

0.411

0.120

0.002

0.411

0.109

0.000

0.597

0.090

0.000

0.565

0.077

0.000

Saturday

-0.021

0.662

0.974

-0.021

0.604

0.972

0.624

1.210

0.614

0.422

0.962

0.664

Winter

-0.802

0.417

0.065

-0.802

0.380

0.038

-0.140

0.344

0.690

-0.004

0.304

0.988

Grounding

0.421

0.296

0.167

0.421

0.270

0.122

1.176

0.467

0.025

1.130

0.372

0.004

Error

-0.147

0.849

0.864

-0.147

0.774

0.850

-2.029

0.976

0.056

-2.085

0.778

0.011

Explosion

0.344

0.686

0.620

0.344

0.626

0.584

1.262

0.829

0.150

1.219

0.645

0.067

Characteristics of the ship

Single hull

0.085

0.432

0.845

0.085

0.394

0.829

0.077

0.787

0.923

0.059

0.621

0.924

Flag

-0.035

0.876

0.969

-0.035

0.799

0.966

1.151

0.677

0.111

1.172

0.545

0.039

Characteristics of the claims

Environmental

1.731

0.417

0.000

1.731

0.380

0.000

0.691

0.639

0.298

0.785

0.488

0.116

Tourism

1.067

0.519

0.050

1.067

0.474

0.027

0.571

0.470

0.244

0.608

0.383

0.121

Fisheries

0.788

0.464

0.101

0.788

0.423

0.066

1.342

0.485

0.015

1.284

0.401

0.003

Strict liability

Strict liability

-2.037

0.587

0.002

-2.037

0.536

0.000

2.955

0.794

0.002

2.677

0.681

0.000

Constant

19.111

5.032

0.001

19.111

4.589

0.000

12.886

7.203

0.095

13.358

5.872

0.029

Sigma

1.796

0.101

1.185

0.145

N

114

114

53

53

Log-pseudolikelihood

-228.533

-81.115

R-squared

0.532

0.829

Regression Results on Compensations

Once again, the results show the crucial effects of the strict liability variable, which in this case increases the compensation received by the affected parties (for the OLS and Tobit models). The amount spilled is also statistically significant, increasing the compensation received, as in the case of claims related to losses suffered by the tourism and fishing sectors. Claims related to environmental damage do not increase the compensation received, but claims in the fisheries sector increase the amount of compensation finally paid. This is a logical result given that under the IMO regulation, environmental damage is not compensated. This is an important result, since the contribution of environmental damage is important when we explain the total damage claimed by a large oil spill. In terms of geographical differences, and as expected, compensations for spills occurring in North America are larger in magnitude, due to the effect of strict liability in the United States. Attending to other variables related to the causing factors, explosion and grounding seem to be related to higher levels of compensation, while the various error causes are related to smaller payments. The result related to the grounding cause contradicts previous results found by Ogus (1999). With regard to time trends, we have observed that the decade of the 2000s is correlated with a higher level of compensation, which may indicate that over time the compensation tends have increased (although there are still major differences between the damages claimed and the amount paid). Finally, attending to the Tobit model, we find that the FOC shows a positive coefficient, indicating that the courts or compensating bodies may consider these types of ships as substandard vessels and therefore provide higher levels of compensation.

Conclusions and Implications

Oil spills result in very serious societal costs and environmental consequences. In order to provide incentives for controlling and properly managing the international bodies involved in oil transportation, the IMO has approved a series of Conventions over the years on liability limits and compensation packages to victims. The objective of this chapter was to assess the evolution of compensations and damage claims, in order to obtain evidence from a meta-regression on the causing factors that may alleviate or reduce these environmental accidents. Although the empirical analysis is simple, the results obtained indicate that strict liability acts as a deterring factor on the impact of oil spills, while other additional explanatory factors can also reduce the impact of the damage and increase potential compensations.

Future studies should investigate whether current compensation schemes are optimally designed. In the past, the compensation of damage caused by large oil spills has been quite limited, and considerably delayed. In the case of the most recent European oil spill, caused by the Prestige in 2002, the short-term economic costs have been estimated to be €743.73 million by Loureiro et al. (2006), while a comprehensive study of the mid-term effects has shown that the damage may amount to more than €4,000 million (Loureiro, 2012). However, the maximum compensation obtainable under the 1992 Fund is approximately €165.76 million, which was subsequently increased by the Supplementary Fund up to \$1159 million (not applicable to the Prestige case). However, there are still many questions to be answered regarding whether the full costs associated with large oil spills can be covered by this Supplementary Fund. Judging by the Prestige oil spill, the maximum compensation obtained is well below the amount of damage caused.

Unfortunately, the establishment of a maximum economic compens ation which is well below the actual damage does not generate sufficient incentives for marine safety. This has been particularly true until recent years in Europe, where single-hull and old vessels were allowed until 2003, whereas these substandard vessels have been forbidden to enter US waters since the 1990s.

Not only are the compensation funds limited; the data obtained also show that economic assessments for past oil spills have been based on only partial valuations, leaving aside important amounts or categories, not limited to environmental damage that was excluded from compensation funds applicable in IMO countries.

We believe that the application of a generally acceptable valuation methodology will reduce the underestimation of damage assessments conducted after oil spills, and will help with comparability issues across international settings. Better estimates of catastrophe magnitude could also be a next step in order to obtain more accurate compensation, or at least, to ensure that the compensation is more in line with the damage caused.

References

Alberini, A., & Austin, D. (1999). Strict liability as a deterrent in toxic waste management: Empirical evidence from accident and spill data. Journal of Environmental Economics and Management, 38, 20–48.Find this resource:

Alberini, A., & Austin, D. (2002). Accidents waiting to happen: Liability policy and toxic pollution releases. Review of Economics and Statistics, 84(4), 729–741.Find this resource:

Alló, M., & Loureiro, M. (2013). Estimating a meta-damage regression for large accidental oil spills. Ecological Economics, 86, 167–175.Find this resource:

Barrio, M., & Loureiro, M. (2010). A meta-analysis of contingent valuation forest studies. Ecological Economics, 69, 1023–1030.Find this resource:

Bernard, S. (2011). Transboundary movements of waste: Second-hand markets and illegal shipments. (Working Paper No. 2011s-77). CIRANO Working Papers.Find this resource:

Bonnieux, F., & Rainelli, P. (1993). Learning from the Amoco Cadiz oil spill: Damage valuation and court’s ruling. Organization and Environment, 7, 169–188.Find this resource:

Bureau of Labor Statistics. (2016). Databases, tables & calculators by subject.

Carson, R. T., Mitchell, R.C., Hanemann, W. M., Kopp, R. J., Presser, S., & Ruud, P.A. (1992). A contingent valuation study of lost passive use values resulting from the Exxon Valdez oil spill. A Report to the Attorney General of the State of Alaska.Find this resource:

Center for Tankship Excellence. (2016). Casualties of oil spills.

Centre de Documentation de Recherche et d’Expérimentations sur les Pollutions accidentelles des eaux (Cedre). (2016). Dossiers sur les grandes pollutions accidentelles des eaux et sur les moins grandes. Cedre, Brest (France). Available from http://wwz.cedre.fr/en/Our-resources/Spills.Find this resource:

Chapman, D., & Hanemann, M. (2001). Environmental damages in court: The American Trader case. In A. Heyes (Ed.), The law and economics of the environment (pp. 319–367). Chelthenham, U.K.: Edward Elgar.Find this resource:

Damage Assessment, Remediation & Restoration Program (DARRP). (2016). Information about incidents. Available from https://darrp.noaa.gov/oil-spills.

Goodstein, E. (1992). Saturday effects in tanker oil spills. Journal of Environmental Economics and Management, 23(3), 276–288.Find this resource:

International Marine Organization (IMO). (2006). http://www.imo.org/en/About/Conventions/ListOfConventions/Pages/Default.aspx

International Oil Pollution Compensation Funds (ITOPCF). (2011). About IOPC Funds.

International Oil Pollution Compensation Funds (ITOPCF). (2016). Oil tanker spill statistics, 2015.

International Tanker Owners Pollution Federation Limited (ITOPF). (2016). Fate of oil spills. Technical Information Paper.

Kim, I. (2003). A comparison between the international and US regimes regulating oil pollution liability and compensation. Marine Policy, 27, 265–279.Find this resource:

Levantion, B. T. (1982). Protection of the high seas from operational oil pollution: a proposal. Fordham International Law Journal, 6(1).Find this resource:

Loureiro, M. (2012). Expert report conducted for the Prestige oil spill.Find this resource:

Loureiro, M., Ribas, A., López, E., & Ojea, E. (2006). Estimated costs and admissible claims linked to the Prestige oil spill. Ecological Economics, 59(1), 48–63.Find this resource:

Mondello, G. (2011). Hazardous activities and civil strict liability: The regulator’s dilemma. (Nota di Lavoro no. 21.2011.) Fondazione Eni Enrico Mattei, Milan (Italy).Find this resource:

Moore, L., Footiff, A., Reynolds, L., Postle, M., Flyod, P., & Virani, T. (1998). Sea Empress cost-benefit project. (Technical Report P119.) Environment Agency’s Project Manager. Bristol.Find this resource:

Oanda. Exchange rates.

O’Brien, R. M. (2007). A caution regarding rules of thumb for variance inflation factors. Quality & Quantity, 41(5), 673–690.Find this resource:

Ogus, A. (1999). Estimating the size of oil tanker spills.

Opaluch, J., & Grigalunas, T. (1984). Controlling stochastic pollution events through liability rules: Some evidence from OCS leasing. The RAND Journal of Economics, 15(1), 142–151.Find this resource:

Organization for Economic Co-operation and Development (OECD). (2001). Spills from tankers, 1975–2000.Find this resource:

Thébaud, O., Bailly, D., Hay, J., & Pérez, J. (2005). The cost of oil pollution at sea: An analysis of the process of damage valuation and compensation following oil spills. In Economic, Social and Environmental Effects of the Prestige (pp. 187–219). Consello da Cultura Gallega. Ed: Albino Prada, Maria Xosé Vázquez. Santiago de Compostela.Find this resource:

US Environmental Protection Agency. (2011). National Service Center for Environmental Publications. Available from https://www.epa.gov/nscep.

Wayne, K.T., Kite-Powell, DJ. (2005) Post OPA-90 vessel oil transfer spill prevention: the effectiveness of coast guard enforcement. Environmental and Resource Economics, 30(1), 93–114.Find this resource:

White, I. C., & Molloy, F. C. (2003). Factors that determine the cost of oil spills. International Oil Spill Conference Proceedings, 2003(1), 1225–1229.Find this resource:

World Bank. (2011). World Bank open data.

Notes:

(1.) The International Tanker Owners Pollution Fund is a not-for-profit organization involved in all aspects related to ship-source spills of oil, chemicals, and other substances in the marine environment.

(2.) To analyze the compensation paid in each oil spill, the data is taken from 52 accidents from the period 1975–2007, due to the existence of missing values.

(3.) We have used the OANDA currency converter, where there are historical data available.

(4.) The inflation calculator is available online.

(5.) This information is available online.