Dear
My trouble is that I deal with non-synchronic elections. In the tutorial of
Ezi this table is present:
Turnout
Vote NoVote
+-------+---------+
Black| betaB | 1-betaB | X
Race +-------+---------+
White| betaW | 1-betaW | 1-X
+-----------------+
T 1-T
where we OBSERVE:
T = proportion of voting age people who Turnout to vote
X = proportion of voting age people who are black
N = number of voting age people
and we wish to ESTIMATE:
betaB = fraction of Blacks who vote
betaW = fraction of Whites who vote
my table considerin PCI party, one of the parties which partecipated to 1976
elections, is the following
Turnout
Vote '79 NoVote '79
+-------+---------+
PCI76| betaB | 1-betaB | X
Party +-------+---------+
1-PCI76| betaW | 1-betaW | 1-X
+-----------------+
T 1-T
where we OBSERVE:
T = proportion of voting age people who Turnout to vote in 1979
X = proportion of voting age people who are PCI in 1976
N = number of voting age people
and we wish to ESTIMATE:
betaB = fraction of PCI '76 who vote in 1979
betaW = fraction of 1-PCI '76 who vote in 1979
Then I yo on taking another party (in Italy there are a lot of parties). The
PSI 1976. Then I analyse the
coefficient of transition betaB between PSI 1976 and PCI 1979 so the table is:
Turnout
PCI79 1-PCI79
+-------+---------+
PSI76| betaB | 1-betaB | X
Party +-------+---------+
1-PSI76| betaW | 1-betaW | 1-X
+-----------------+
T 1-T
where we OBSERVE:
T = proportion of voting age people who Turnout to PCI '79
X = proportion of voting age people who are PSI in 1976
N = number of voting age people
and we wish to ESTIMATE:
betaB = fraction of PSI '76 who PCI79
betaW = fraction of 1-PSI '76 who PCI '79
Then in the same way I considerer all the other existing parties PSI76 on DC79
mtch. in successive steps
(proceding IMPORT in Ezi).
My first question is: If one less betaB and one less betaW are not exacty the
of 1-betaB and betaW is not
to complement consider because it's wrong? GHactual = 2 (example my Ezi output)
what does it mean?
How does this index change?
My second question is: taking the variables in couples must they actually come
out parts on vote or parts
on part too? Just PCI76 on Vote79 or PSI76 on PCI79?
Could yuo be so kind of giving me an example parties in non syncronic
elections? Considering the existence
in time 0 (1976) of at least 3 parties in time 1 (1979) of 3 parties as ever.
Best Regards.
Claudio Grasso
University of Trento
Trentino-Sùdtirol
Italy
Dear, I ask yuoif the following in the right my working EXCEL database:
>
> Party 1 in Don't vote Subscrive Party1 Don't
> vote
> to in to electoral list in t1 in t1
> in ((to+t1)/2)
>
> Council 1 50 50 100 40 60
> Cooucil 2 51 49 100 41 59
> Council 3 52 48 100 42 58
> Council 4 53 47 100 43 57
> Total 206 194 400 166 234
>
> For me: t=party1 in t0, x=Don't vote in time t1, n=Subscrive electoral list
>
> ((all people in vote age in t0 + t1)/2).
> Then as the example of Ezi sample.asc shows I consider once finished this
> first elaboration party one in theme t1 because its present in the
> successive
> elections in t1 so t = party1 in t0, x= Dont vote in time t1, n= Subscrive
>
> electoral list ((all people in vote age in t0 + t1)/2).
> I think that yuor expression vote other party means considering all the
> parties in t0 and t1 taking them as couples one to the time t0 and the other
>
> to t1 included the dont vote in time t0 and t1 completed the elaboration and
>
> obtained the result with the output Quantities of interest I consider t
> =
> party2 in time t0, x = Dont vote in t1, n = Subscrive electoral list
> and
> so on for the other eventual parties then I sum up the output of
> Quantities
> of interest and I abtain the single coefficient of transition.
> If the parties are 2 for example the result is a table 3 x 3 considering
>
> also dont vote in t0 and t1.
> Is this description correct? Thank. Best regards. Claudio Grasso
>
> > University of Trento
> > Trentino Suedtirol
> > Italy