The Determinants of Transitional Justice Demand: Avoiding False Acquittal and False Conviction

1
The Determinants of Transitional Justice Demand
Avoiding False Acquittal and False Conviction
By Monika Nalepa Rice University
Prepared for Presentation at the EITM workshop,
Ann Arbor, Michigan July 13, 2006
2
Lustration
What is Transitional Justice ?

• Legal institutions designed by new democracies to
  deal with members and collaborators of the former
  authoritarian regime

Reconciliation

• Procedures verifying whether persons running for
  office had collaborated prior to the transition
  with the former authoritarian regime

Willingness to participate in democratic
institutions with persons who prior to the
transition were on the opposite side of the
regime
3
DV Reconciliation (which cannot be measured
)

• Survey (face to face interviews)
• Poland (n1005)
• Hungary (n1003)
• Czech Republic (1049)
• (nationally representative samples)

Concept interpretation
measurement
Decrease in Transitional Justice Demand
Do you believe that at this point in time
lustration should be carried out? 1- Strongly
Disagree to 5- Agree Strongly
Reconciliation
4
The Question Why do citizens reconcile? (demand
less Transitional Justice)

• Political preferences matters
• Dissident past matters
• Threat perceptions of former autocrats matter
• Other things that might matter
• Age
• Settlement
• Education

5
Featured hypothesisFalse Conviction and False
Acquittal errors
Secret Police Files were destroyed by the
outgoing regime

False acquittal
Not all perpetrators or secret informers can be
punished
Files were fabricated by secret police officers
False conviction
Some Innocent people may be accused
As in the theory of hypotheses testing in
statistics, there is a trade-off between avoiding
these two errors
6
DV Do you believe that at this point in time
lustration should be carried out?
7
DV Do you believe that at this point in time
lustration should be carried out?
8
TJ Demand over time (in std. deviation) as
function of voting behavior (Poland)
9
Threat Perceptions and TJ Demand over time
10
Dissident vs. authoritarian past?
d(Pr(Y1)) d(Pr(Y2)) d(Pr(Y4))
d(Pr(Y5)) d(X)
d(X) d(X) d(X)
Czech R -.0193 -.0345
.0472 .0412
11
False Acquittal in Hungary and the Czech Republic
d(Pr(Y1)) d(Pr(Y2)) d(Pr(Y4))
d(Pr(Y5)) d(X) d(X)
d(X) d(X)
d(Pr(Y1)) d(Pr(Y2)) d(Pr(Y4))
d(Pr(Y5)) d(X) d(X)
d(X) d(X)
-.0552 -.0244 .0275 .0534
-.0132 -.0236 .0324 .0283
12
False Conviction in Poland and the Czech Republic
d(Pr(Y1)) d(Pr(Y2)) d(Pr(Y4))
d(Pr(Y5)) d(X) d(X)
d(X) d(X)
d(Pr(Y1)) d(Pr(Y2)) d(Pr(Y4))
d(Pr(Y5)) d(X) d(X)
d(X) d(X)
.0142 .02027 -.0104 -.0474
.0250 .0448 -.0613 -.0536
13
What do these results imply for TJ institutions?

• ITR Politicians
  Verification
• Declaration by
  Lustration
• 
  Agency
• ETR Search for Target
  banned,
• Evidence by from
  running for
• Lustration Agency office if
  evidence
• found

(Incentives Based Truth Revelation)
(Evidence Based Truth Revelation)
14
When does a ITR mechanism exist?
15
(No Transcript)
16
Incentive Compatibility
17
Ties
e
0
18
When one of the IC conditions is not satisfied
e
e
0
0
19
The idea of an ITR

• p is the probability with which evidence for
  collaboration (files) exist
• Payoffs C (Collaborator), N(Non collaborator), L
  (lustration liar)
• LltCltN

Nature
Files
Files
Eu(Admit)gtEu(Deny) when C pL (1-p)N or L
C/p
(p)
(1-p)
Agent
Admit
Deny
Deny
Admit
C
L
N
C