THE POTENTIAL FOR POLITICAL VIOLENCE IN ARGENTINA, ETHIOPIA, AND THAILAND APRIL 1975[SANITIZED] - 1975/04/01

• Document Type: SPECIALCOLLECTION
• Collection: Argentina Declassification Project - The "Dirty War" (1976-83)
• Document Number (FOIA) /ESDN (CREST): 06628560
• Release Decision: RIPPUB
• Original Classification: U
• Document Page Count: 10
• Document Creation Date: April 3, 2019
• Document Release Date: April 12, 2019
• Publication Date: April 1, 1975

Approved for Release: 2018/10/02 C06628560
Research Project
The Potential for Political Violence in
Argentina, Ethiopia, and Thailand:
Report on a Quantitative Analytical Model
Progress Report No. 4
ntial
gonflitthal
3.5(c).
April 1975
Approved for Release: 2018/10/02 C06628560
Approved for Release: 2018/10/02 C06628560
Approved for Release: 2018/10/02 C06628560
3.5(c)
3.5(c)
Approved for Release: 2018/10/02 C06628560
CO TIAL
Progress Report No. 4 0 April 1975
The Potential for Political Violence in Argentina, Ethiopia, and
Thailand: Report on a Quantitative Analytical Model
This is the fourth in a series of reports on the testing of an analytical model
of political violence adapted from Ted Robert Gurr's frustration-aggression theory
(Why Men Rebel, Princeton University Press, 1970). The Gurr theory is based
on the proposition that political violence is the product of group frustrations
reinforced by the belief that violence is justified and by the capability to turn the
resulting politicized anger into collective aggression. The form of the ensuing
violence�turmoil (riots or demonstrations), insurgency (terrorist acts or small-
scale guerrilla operations), conspiracy (attempted coups), or internal war ( large-
scale revolutionary actions or civil war)�will depend upon the kinds of relation-
ships which are found to exist among various types of groups in the country
under examination. A complete description of the theory and of the procedures
devised to operationalize it as an analytical tool is contained in an earlier OPR
report of an experimental, ex post facto test of the model in the Chilean situ-
ation of mid-1973 ( OPR 502, Revised, November 1974).
The purpose of this second phase of the project is to test the value of the
Gurr model as a technique for assessing the nature and potential for political
violence in societies of varying stages of development and with quite different
cultural heritages and political institutions. The three countries chosen�Argen-
tina, Ethiopia, 'and Thailand�seem to meet our requirements: Each represents
a different stage of development in a distinct geographic area, and each is con-
fronted with domestic unrest or political conflict.
For each of the three countries, a panel of five CIA analysts assigns numerical
evaluations at regular intervals to the model's key variables�relative depriva-
tion or collective frustration, belief in the justification for violence, coercive
force, and institutional support. The evaluations are made for each group or
"actor" which, in the panel's judgment, represents a significant political force in
the country. The country's actors are also assessed in terms of their identifica-
tion with four basic actor-categories: Pro-regime, mass-oriented (PR-MO);
pro-regime, elite-oriented (PR-E0); anti-regime, mass-oriented (AR-M0); and
anti-regime, elite-oriented (AR-EO). At each point of assessment during the
test period, the panel's evaluations are combined statistically, using comput-
erized procedures devised in the Office of Political Research, to produce over-
all evaluations of the Potential for Political Violence (PPV) and of the conditions
conducive to particular types of violence in the country under observation. The
three panels' evaluations are made on a monthly schedule, and the entire test
series is expected to run until mid-1975.
NOTE: Comments or questions on this project will be welcomed by its author
Office of Political Research, code 143, ext. 4091.
3.5(c)
Approved for Release: 2018/10/02 C06628560
Approved for Release: 2018/10/02 C06628560
CO TIAL
The summary below presents the principal findings from the fifth set of
assessments made by the country panels in mid-April 1975. The graphs on suc-
ceeding pages display trends based on the five sets of assessments made to date.
Subsequent monthly progress reports, based in each case on mid-month evalu-
ations, will indicate trend changes in the panels' scoring patterns. In this way,
the model's projections of the potential and form of political violence can be
tested against actual developments in each country during the test period.
Summary of Principal Findings
( NOTE: The data cited below and in the following graphs are keyed to
each country's Potential for Political Violence or PPV, as reflected by the
median PPV score of each five-member country panel. The figures should
not be regarded as representing either probabilities or absolute quanti-
ties; they should be interpreted merely as indicating relative status or
strength, compared to other variables in the model assessed by the same
country panel or to corresponding variables assessed by the other two
country panels.)
General:
1. In mid-April, assessments of all three country panels produced shifts
in trends which contrasted sharply with previous months. For the first time
since mid-January, the potential for political violence in Ethiopia decreased,
though it remained the highest of the three countries. At the same time, the
potential in both Argentina and Thailand jumped sharply, ending downward
trends in those two countries over the past two months. At mid-April, Thailand
ranked somewhat below Ethiopia in potential for political violence, with
Argentina now a distant third.
2. Within the limits of Argentina's relatively lower potential, the conditions
for the four basic types of political violence were roughly equal. In contrast,
conditions in Ethiopia continued to show a strong tendency toward internal
war and conspiracy, with the emphasis shifting from the former to the latter.
In Thailand, within the limits of its considerably enlarged potential, the condi-
tions for turmoil clearly outweighed those for insurgency, conspiracy, and
internal war.
3. In Ethiopia, anti-regime actors, particularly anti-regime elite groups,
continued to display the greatest potential for political violence. In the other
two countries, particularly in Thailand, anti-regime mass actors appeared to
have the strongest potential.
4. While in Ethiopia politicized frustration subsided somewhat as a moti-
vating force for violence, it rose markedly in both Argentina and Thailand. The
rise in Thai politicized frustration was particularly marked, reaching a level
60 percent above that of mid-March. Though Thailand displayed a perceptible
increase in institutional support and coercive force needed to turn politicized
frustration into violence, the capability to incite actual political violence re-
mained relatively stable in the three countries. In mid-April, Thailand displayed
the largest amount of institutional support and coercive force, Argentina the least.
NTIAL
Approved for Release: 2018/10/02 C06628560
Approved for Release: 2018/10/02 C06628560
.5..C).1F111
5. The range of variation in the assessments of the Argentine and Ethiopian
panels remained relatively constant in the March-April period, while that for
the Thai panel increased markedly. At mid-April, the range of variation was
approximately the same for all three panels.
Argentina:
The potential for political violence in Argentina showed a marked increase
in the March-April period (a jump from 37 to 48 in median PPV score). Within
the limits of that increased potential, conditions favorable to turmoil (12),
conspiracy (12), and internal war (14) all increased during the period. Increases
in PPV were evident among all types of actor groups, but the increase for anti-
regime mass actors was the largest ( a gain of 6 to 15). Politicized frustration
widened during the period ( an increase from 21 to 30), while the capability
to channel that frustration into actual political violence remained relatively
unchanged ( a gain of 1 to a PPV score of 9 for both coercive force and institu-
tional support). The range of variation in assessments among members of the
Argentine panel remained constant ( a range of 43 around a median PPV score
of 37 in March, compared to a range of 42 around a median PPV score of 48
in April). The pattern of variation, however, showed a continuing divergence
between analysts from OPR and DDO at the low end of the scale (Analyst A
from OPR at 23, Analyst B from DDO at 16) and those from CRS, OER, and
OCI bunched at a considerably higher point ( Analyst C from CRS at 52,
Analyst D from OER at 58, and Analyst E from OCI at 48).
Ethiopia:
The potential for political violence in Ethiopia showed a slight decline from
March (from 78 to 73 in median PPV score), reversing the uptrend in PPV over
the past two months. While the conditions for turmoil and internal war declined
(a decrease from 16 to 13 for turmoil, a drop from 24 to 19 for internal war),
the conditions for insurgency remained the same (at 18), and those for con-
spiracy increased (from 20 to 23). While in March conditions seemed most
favorable to internal war, conditions in April appeared most conducive to con-
spiracy. The potential for violence among the four types of actors remained
relatively unchanged; anti-regime mass and elite actors (at 20 and 27, respec-
tively) continued to hold a strong edge in PPV over pro-regime mass and elite
actors (9 and 17, respectively). Politicized frustration in the country declined
slightly in the March-April period (from 56 to 52), while the capability to
turn that frustration into actual political violence remained stable (10 for
coercive force, 11 for institutional support). The range of variation among
Ethiopian panel members remained constant during the period (40 around a
median PPV score of 78 in March, 41 around a median PPV score of 73 in April).
Thailand:
In the March-April period the potential for political violence in Thailand
jumped sharply (from a median PPV score of 44 in March to one of 67 in April).
This is the largest monthly increase recorded for any of the three countries
since the present series of assessments began. Of the conditions conducive to
particular types of violence, those for turmoil showed the largest increase (from
C9054:EgrIAL
Approved for Release: 2018/10/02 C06628560
Approved for Release: 2018/10/02 C06628560
COyDL'�AL
17 to 30), the conditions for internal war and insurgency also grew (from 9 to
17 for internal war, from 13 to 16 for insurgency), while those for conspiracy
remained relatively low and stable (an increase from 5 to 6). The greatest
increase in PPV was seen among anti-regime mass actors (an increase from 13
to 28); smaller increases were recorded for anti-regime elite actors (8 to 10),
pro-regime mass actors (7 to 11), and pro-regime elite actors (16 to 18). Fueling
the relatively large increase in PPV, politicized frustration rose sharply in the
period (from 24 to 40), while the capability required to transform that frustra-
tion into actual political violence increased moderately (from 9 to 14 for
coercive force, from 11 to 13 for institutional support). Variation in assessments
among Thai panel members increased during the period (from a range of 34
around a median PPV score of 44 in March to a range of 41 around a median
PPV score of 67 in April).
4
CONENTIAL
Approved for Release: 2018/10/02 C06628560
Approved for Release: 2018/10/02 C06628560
CO
IAL
Conditions Conducive to Particular Types of Political Violence
ShOdetl�Oras,,reprp,isnt:Ai(ithilf....the'llaiitt of ea;h.coun,fry's Turmoil
for political violence (PPV) the degree : to ':liliiClf:that countryInsurgency
median scores on key variables fulfill the conditions hypothesized by
the Gun theory:ine'...celidutiVe;:tO eacli::lipe-:ifitteritical violence
-:� �. �
, 1W Internal War
CO TIAL
Approved for Release: 2018/10/02 C06628560
'PPV
100
BO
60
40
20
Approved for Release: 2018/10/02 C06628560
Sjablfttl.
The Potential for Political Violence (PPV) for Each Type of Actor.
ARGENTINA
� 0
Dec 74
PPV
100
40
60
40�
20
Dec 74
PPV
100
80
48
PR-MO 10
PR -E0 11
AR-MO 15
AR-012
Jan 75 Feb
Mar
Apr May J110
ETHIOPIA
73
PR-MOO
PR -co 17
AR-MO 20
60
20
O.
bee 74
AR-E0 27
Jan 75
Feb
Mar.
Apr
May
- Jun
. THAILAND
67
PR-611111
PR-E0 18
AR-MO 28
AR-ED 10
Jan 75
Feb
, �..
,-:Bhidefl:areos;roprepont each country
panel's median scores for PPV for
'..;��-1.:eaOti�OOfOr category. '������. �
�
CONFI IAL
Apr
May
Jun
PR-MO-Pro-Regime, Moss-Oriented Actors
PR-E0- Pro-Regime, Me-Oriented Actors
AR-MO-Anti-Regime, Mass-Oriented Actors
AR-E0- Anti-Regime, Elite-Oriented Actors
6
CONFI�NTIAL
565722� 4,75 CIA
Approved for Release: 2018/10/02 C06628560
Approved for Release: 2018/10/02 C06628560
SS)Nff6grITIAL
The Potential for Political Violence 0311V) in Terms of Key Variables in the Gurr Model
PPV
100
BO
60
40
ARGENTINA
20
0
Dec 74
PPV
100
80
60
20
0
Dec 74
PPV
49
IS 9
CF
PF 30
Jan 75 Feb Mar Apr May .Jun
ETHIOPIA
73
IS ll
PF 52
100-
Jan 75::
Feb
Mar
Apr
May,
Jun
.THAILAND
67
15 13
CF 14
20
PF 40
0
Dec 74 Jan 75 Feb Mar Apr May Jun
C0N)IDENTIAL
Shaded areas represent sums of cant), country
panel's median.sccrei.'for�each key variable .
all actor categories:
7
CONFIDENTIAL
IS - Institutional Support:
CF - Coercive ForcePF �
- Politicized Frustration
. .
(relative deprivation x.
justifidiitioli for imilence).
565723 475CtA
Approved for Release: 2018/10/02 C06628560
Approved for Release: 2018/10/02 C06628560
CON TIAL
Individual Analysts' Assessments of the Potential for Political Violence (PPV)
BO
60
0
ARGENTINA
1.0000
e,m, ..... _......
....,. ..m
ow4.219.agw""""
, .52
.a.....er �"18
WO-
00
....woo
--------16
23
1.
.0
.....
.....
Dec 74
20
Jan 75
Feb
Mar
Apr
May Jun
ETHIOPIA
rx.���''-
,,,,,, oo-
,CINMISE
-
--e....,,,,,, 0..0
.--
����� NIB
....
...�
85
7733
65
s
0�-----o.-
ro# so
�
...
.1. . .7...? 6 .es.., 8 6 ......4. ::
*4'
...0
s'
.zz, �
...
Dec 74 Jan 75
.60
40
20
Feb
Mar
A r
May
THAILAND
.Jun
et
89
*o�
%., e ,..
.
4'0
MCIEPIOVIIIWIFIVV,ie...6
. �=p
...;
.4
�
48
worm ss.... ..�
..-
.."""
....... -..."*-
�'
*. .....
..��
�".r.... .. ms
Dec 74
O)�DENTJAL
Jan 75
Feb
Colored lines represent the range of individual -
analysts' scores for PPV around each. Country
panel's median score for PPV.
Mar
8
CONFID AL
Apr
May.
Jun
AnalyttA
�
Analyst-8
Analyst-C
Analyst-D
Analyst-E
565724 4456A
Approved for Release: 2018/10/02 C06628560