Measures of insularity

Measures of insularityy
Anna M.Pinna and Vania Licio
(University of Cagliari and CRENoS)
(University of Cagliari and CRENoS)
The Periphery and its Host Economy (Pemabo 2012)
4th Sept. Ísafjörður, Iceland
p
j
,
The background
The background
‘The costs of insularity’: tender financed by g
g
(
)
Sardinian regional government (CRENoS)
The project is focused on the case of Sardinia, Th
j i f
d
h
f S di i
but it is our interest to study the issue in a larger scale.
Project has several research lines.
Insularity in an international context
Problem of definition (choice):
bl
f d fi i i ( h i )
‐ insularity vs remoteness
unit of analysis (region/state/…continent)
‐ unit of analysis (region/state/…continent)
‐ dimensions to be used in order to measure insularity
‐ which other variables (objects) are we looking for? which other variables (objects) are we looking for?
firms (income, wages)? residents
id t as consumers? ?
people’s movements?
Objectives
• C
Consider
id the
th issue
i
off insularity
i l it in
i a comparative
ti set‐
t
up, at the international level.
• Micro/macro issues where geography plays an
important role (i.e. growth, trade).
• Measuring insularity as a state of nature (exogeneity)
with the aim of understanding its further interactions
with economic variables.
• Next step: test of our measures in the gravity model.
Available geography dataset
Available geography dataset
• Source: Sedac‐PLACEII
S
S d PLACEII
• Countries included: 228 Countries included: 228
• Variables included: 76
Variables included: 76
• Typology
yp gy variables: morphological
p
g
land characteristics
along several dimensions
–
–
–
–
Elevation;
Climate zone;
Climate zone;
Biome class;
Distance from coast.
Characteristics of insularity
Smallness (is it size the issue?)
Remoteness
(transport costs) =>> isolation
(transport costs) isolation
Vulnerability
(exposition to shocks)
Economic consequences of insularity (the BAD)
Armstrong and Read (2004a, 2004b)
A
t
d R d (2004 2004b)
1. the small size of the domestic market;
2. islands have to deal with a limited resource base;
3. a limited land area;
4. island and small states are often characterized by a strong migrating phenomenon;
migrating phenomenon;
5. islands are characterized by a narrow domestic output, exports and import market;
6 islands are often vulnerable (Briguglio 1995; Atkins et al 2000);
6. islands are often vulnerable (Briguglio, 1995; Atkins et al., 2000);
7. the domestic sector in islands tend to be extremely small;
8. in islands, especially small ones, the consequences of trade and fiscal policies may be exacerbated.
Economic consequences of insularity
(the GOOD)
Armstrong and Read (2004a, 2004b)
A
t
d R d (2004 2004b)
1. islands' high degree of structural openness to trade and high dependence upon exports may lead to
and high dependence upon exports may lead to export‐led growth strategies;
2. pursuing a niche sectoral
p
g
ggrowth strategy, may lead to gy
y
high standards of living as well as soon as adequate policies take place (Armstrong and Read, 1995, 2002; Armstrong et al 1998);
Armstrong et al., 1998);
3. being an island represents per se a factor of attraction for tourists;;
4. it is easier to build social capital.
Data and sources about islands (1)
Data and sources about islands (1)
• It doesn’t exist, at the international level, a database that includes for each country ALL its islands.
• However, there are four main websites that contain the
largest and most extensive amount of information, data
and statistics relating to the islands of the world.
• These websites are the four recommended by IGU
(International Geographical Union) Islands Commission.
Data and sources about islands (2)
Data and sources about islands (2)
•
•
•
•
http://www.worldislandinfo.com/ http://islands.unep.ch/ p //
p /
http://www.worldatlas.com/
htt //
http://www.globalislands.net/
l b li l d
t/
We mainly used the first two, because they are the
more detailed,
detailed and the third one just for some
elaborations.
Data and sources about islands (3)
Data and sources about islands (3)
• http://www.worldislandinfo.com
The web site is an attempt to provide the most accurate
information on the world’s islands.
Thousands of maps, reference works, books, web sites and
articles
i l have
h
i f
informed
d its
i creation.
i
• http://islands.unep.ch/
//
/
The web site is developed by UNEP (United Nation
Environment Program).
Insularity dataset Worldislandinfo and Unep (1)
• Putting together information from two islands databases
allowed to build a single dataset formed by 165 nations.
Î Information has been aggregated at the country level
(starting from data on more than 2000 islands).
islands)
• Info from “Worldislandinfo”:
Worldislandinfo : allowed us to built a dataset of
161 countries.
ƒ Number of islands by country;
y
y;
ƒ Total islands area by country;
ƒ Percentage of country area represented by islands.
Insularity dataset Worldislandinfo and Unep (2)
•
Info from “UNEP”: dataset composed by 144 countries. Variables
which refer to important characteristics for each island:
ƒ The same variables included in “Worldislandinfo”;
ƒ Isolation Index;
ƒ Coastal Index.
•
Diapositiva 14
•
For all 165 countries we also constructed a dummy set of variables:
ƒ Nation State Island (captures political geography) (i.e. UK and Japan);
N ti St t I l d ( t
liti l
h ) (i UK d J
)
ƒ Nation States which share an island (i.e. Borneo);
ƒ Countries distributed in an archipelago.
Since the purpose of the work is to study the costs of world countries insularity, by analyzing the flow of
international trade that they hold, it was necessary to adjust this dataset with trade data and, therefore, the
number of countries involved in the analysis was reduced to 151 because some nations have been merged
into others.
others
Landlocked countries
Landlocked countries
• Using the information about landlocked countries, we assumed that
the 34 landlocked countries in the geography dataset had insularity
measures equal to 0 (3 landlocked countries had islands, they were,
in fact, included in the islands databases).
• Using the information about not landlocked countries, we analyzed
their maps with the web site www.worldatlas.com e we founded
that the 52 not landlocked countries in the geography dataset had
no islands. So, as above, we attributed 0 to insularity indexes.
• ?
Insularity information
Insularity information
• Thanks to this assumptions we have insularity
information about 237 countries:
‐ 142 info on both insularity and geography;
‐ 9 included just in insularity dataset;
‐ 52 + 34 landlocked countries.
• Number of states included in the maps are 244.
Number islands by country
valori assoluti
(119 135] (4)
(119,135]
(99,119] (2)
(79,99] (1)
(49,79] (3)
(19,49] (18)
(1,19] (76)
[0,1] (126)
No data (14)
Nostre elaborazioni su dati Unep
Issues of measurement: number of islands (?)
number of islands (?)
• 0
0 and 1 are really different cases, but this is not d1
ll diff
t
b t thi i
t
reflected in this simple measure
• Insularity as a different condition depending on Insularity as a different condition depending on
‘geography’
Landlocked / coastal => having 1 island (2 3)
Landlocked / coastal => having 1 island (2….3)
• How to think BAD geography? Di
Dispersion of territory...
i
f t it
Landlocked
Archipelago
. l country isol_w if isl_3==1 & d_c_isl==1
country
isol_w
15.
20.
23.
28.
49.
Aruba
Bahrain
Barbados
Bermuda
Christmas island
18
13
46
91
79
50.
60.
64.
91.
100.
Cocos (keeling) islands
Cyprus
Dominica
Guam
Iceland
103
27
41
86
65
151.
156.
165
165.
166.
186.
Montserrat
Nauru
Ni
Niue
Norfolk island
St. helena
39
97
97
101
113
188.
197.
Saint lucia
Singapore
41
3
Our indicator
Our indicator
Put together two information:
1. dispersion of territory in islands
2. share of territory on island
# of islands in 95% land x share of island land
Weighted Isolation index by country
valori %
(80.89,113] (22)
(46,80.89] (21)
(31 77 46] (24)
(31.77,46]
(15.21,31.77] (22)
[3,15.21] (23)
No data (132)
Nostre elaborazioni su dati Unep
Insularity and GDP: all countries
Insularity and GDP: all countries
Insularity and GDP: not landlocked
Insularity and GDP: not landlocked
Insularity and GDP: insularity index>=200
Insularity and GDP: insularity index>=200 for State Islands
Isolation index and GDP: insularity index>=200 for State Islands
Isolation index and GDP
Isolation index and GDP
Percent land area, within 100 km of the coast - by country
valori %
(80 100] (92)
(80,100]
(50,80] (18)
(30,50] (23)
(7,30] (44)
[0,7] (46)
( )
No data (21)
Nostre elaborazioni su dati Sedac - PLACEII