Askew Lecture Series: 2023 - Coins Cruises Crashes


The Reubin O'D. Askew Department of Government - University of West Florida - College of Arts, Social Sciences and Humanities, presents the 2023 Askew Lecture Series Biased Coins, Cruises and Crashes: A Study of Elections in Four Countries. 


Using concepts from statistics (the normal curve), physics (exponential decay), and politology (cost of ruling), professors Alfred G. Cuzán and Richard J. Heggen developed models that approximate the results of 2,000 state or provincial elections in Australia, Canada, Germany and the United States. Professors Cuzán and Heggen have a long history of unusual collaboration across two disciplines, political science and civil engineering, respectively. It began in the 1970s, when they were teaching at universities in New Mexico, Heggen in Albuquerque and Cuzán in Las Cruces. In close to half a century, they have made original contributions to the study of socially efficient government, the Nicaraguan Revolution and U.S. presidential elections (a subject in which professor C. Mike Bundrick of UWF also collaborated). In the last two years, Heggen and Cuzán have turned their attention to “the cost of ruling,” the loss of support that the political party in control of government incurs during its time in office. 
During this talk, they presented partial results of these on-going investigations. Professor Josep Colomer of Georgetown University critiqued the presentation in light of his own work.


*The following Automated transcript has not yet been corrected. The corrected version is in progress. Check back again for updated versions.


0:11
all right good evening everyone welcome please  uh go ahead and take your seats let me remind  
0:19
you to uh turn off silence your phones right now  I'm David Ramsey I serve as chair of the Ruben o  
0:27
diasu Department of government it's my pleasure to  introduce tonight's speakers but first a few words  
0:34
about the name at the top of your programs Ruben  ODI ascu served as governor of this state from  
0:41
1971 to 1979 he was the first Florida Governor to  serve two full four-year terms in office though  
0:50
born in Oklahoma we claim the governor as our  own here in Pensacola because he moved here  
0:57
during childhood and after he' studied law in  Gainesville he moved back and started his law  
1:03
practice many of you are familiar with his firm  which Bears such a prominent place in our Skyline  
1:10
today Levan Pap anonio Governor ASU was a close  friend of the late Fred Levin's Father David like  
1:19
so many in this area ASU valued military service  serving as a naval intelligence officer during the  
1:26
Korean war after serving in both houses of the  legislature culminating in a two-year term as  
1:34
president protim he was elected governor in 1971  winning 57% of the vote his accomplishments in  
1:43
office were wide ranging I can name a few uh he  oversaw the desegregation of public schools and  
1:51
implementation of busing he secured passage of the  state's first corporate income tax and increased  
1:59
Homestead exemp iions at the same time he oversaw  the reapportionment of slate of State Legislative  
2:05
districts following the Supreme Court's ruling in  Reynolds versus Sims an advocate for transparency  
2:12
in government ask you secured passage of the  sunshine Amendment an early attempt to shine  
2:18
a light on the role of dark money in our state  politics this was the First Amendment to the  
2:26
State Constitution that was ratified by voters  rather than the legislature and it secured more  
2:32
than 75% of the popular vote ask you cultivated  a public reputation as a t- totaling Church  
2:39
attending politician who didn't smoke didn't cuss  and opposed gambling in all of its forums he was  
2:47
known by friends and adversaries as Ruben the good  we find in ask you's record of Public Service a  
2:56
welcome example of a successful politician who  did not not set aside his Integrity in order to  
3:02
secure political power in the words of one of  his political adversaries he is established a  
3:08
kind of morality and office that causes people  to have faith in government perhaps it's for  
3:14
this reason that a study of Harvard's Kennedy  School of government ranked asku as one of the   10 best American governors of the 20th century  we commemorate his legacy tonight with this talk  
3:28
now in just a moment I will welcome to the stage  our speakers but first I should introduce them   Dr Alfred kuzan is our distinguished University  professor of government he's published articles  
3:40
and book chapters on a variety of topics during  his career here at WWF most recently a book the  
3:47
laws of politics their operations in democracies  and dictatorships after serving his department  
3:55
chair for two decades he now teaches regularly  in both our grad graduate and our undergraduate  
4:01
programs and in 2016 received a full bright  to teach in Estonia at the University of
4:08
tartu next to him is Richard Hagen professor  ameritus of civil engineering at the University  
4:16
of New Mexico with whom Dr kuzan has collaborated  frequently over a 40-year career to analyze  
4:23
questions of government scope fiscal policy and  elections we'll hear more about their research in  
4:29
just a moment the batting cleanup as it were  is Joseph kir who holds dual appointments at  
4:36
Georgetown University's School of Foreign Service  and the autonomous University of Barcelona's  
4:42
Institute of political and social sciences  during his distinguished career Professor kir  
4:49
has published widely in the field of comparative  Politics the record of his published books and  
4:54
articles book chapters uh runs to more than 200  his next book out this July studies the American  
5:01
constitutional order as a source of our current  political divisions following the presentation of  
5:07
professors kuzan and Hagen Professor COA will  provide some commentary and Analysis of their  
5:13
work each of you also have important work  to do this evening it's our practice in the  
5:21
department of government when we invite speakers  on occasions like this to require them to stand  
5:27
for questioning after they're done so right  now as you prepare to listen to their remarks  
5:32
I encourage you to think as you're taking in  the speeches what is my question what would I  
5:37
ask if I if I could these speakers to address  maybe their comments provoke something in your  
5:43
mind um and so remember you have work to do once  their work is done my work is done I've completed  
5:52
introductions of our speakers please join me  in welcoming to the podium our three speakers
6:07
thank you all very much for coming I asked my  colleagues um in the department of government  
6:15
uh when I first uh did the first draft of my  presentation um I said what do you think of  
6:21
that title does it tell you anything I it can  you figure out what it means and I never got a  
6:27
response um and you know we I I always teach my  students pick uh pick uh a title that is sort  
6:39
of like uh katchi you know that will catch  people's attention and they give a subtitle  
6:44
that actually substantive that actually are you  hear me okay uh that is that is substantive okay  
6:52
uh that is substantive that actually describes  what the paper uh is all about so I I did I did  
7:00
hopefully I I I succeeded in my in my first um  with I succeeded in the purpose of my tile all  
7:09
right so this is U the product one product of  a collaboration that I've had with uh Richard  
7:20
Hagen for a number of years uh like Dr Ramsey said  over instead of 40 years every 10 years or so we  
7:28
kind of get together and you know get going and  then you know I do I go my way he goes his way  
7:34
and then 10 years later he can do it again um and  so but I think this time I think it's probably um  
7:44
in my biased opinion uh I think the best work  we're we're doing uh the best work I think uh  
7:51
but but but it's not finished we it's kind of  work in progress so how did an engineer and a  
7:59
political IAL scientist or politologists as I  like to call myself um get together well what  
8:06
happened was that uh we were both teaching in New  Mexico he was at in Albuquerque at the University  
8:12
of New Mexico and I was at New Mexico state in  Las Cruses and I interest I was interested in  
8:17
water and he's a engineer who studies also Water  Resources economics so we happened to go to the  
8:25
same seminar uh in Austin and on the way back back  uh from the um seminar you know we happen to sit  
8:34
one of those random events uh that uh brought us  together sitting next to each other and so what am  
8:42
I to do the next hour so I pulled out an article I  was working with I said what do you think of this   and I said well that look kind of something like  I've done you know it's suggest some of the things  
8:51
that I have done in Water Resources economics  so that's it uh that launched the number of   Articles uh that got published and the other day  he was telling me how he he took these articles  
9:02
when he was applying for professorship and he  took the Articles to uh the political science   department at um New Mexico and say what do you  think of this are this legitimate the GU said
9:12
yeah all right so um over the last 10 years  or so I I began to get interested uh in in  
9:25
elections worldwide and uh it's one of those weird  things that happen you got to get interested in  
9:31
something and you say something in class and and  the next thing you know you're and and I don't  
9:36
really can't remember how these things came to me  it's kind of weird uh I don't know if anybody else   has that experience but the next thing I know  I'm writing an article and I got published in a  
9:46
in in a journal that is not the most prestigious  but it's the most widely read in the discipline  
9:52
and then I did another one and another one and  another one and um and so then um after some TR  
10:00
time to several tries to get uh Richard interested  in in working with me and some more on this he he  
10:06
finally he finally picked up and so we did so what  am I going to do today well there is something in  
10:12
the study of of Elections um that is titled but it  goes by the phrase the cost of ruling or the cost  
10:20
of governing and this is the cost in votes that  the party in office incurs in in office office in  
10:29
other words certain you might say it pays a price  in votes for exercising power and um the Pioneers  
10:38
in this work or two two European uh economists  actually political Economist perhaps uh uh Peter  
10:46
nanad and Martin paam um and they published a  paper actually and eventually became it sort of  
10:53
uh was republished as as a chapter in a in a book  and they came up with the they study like all all  
11:03
well about 18 uh developed democracies in Europe  Japan Australia uh and North America and they came  
11:11
up with a a number they said that on average the  incumbents uh lose 2.25% points percentage points  
11:22
uh but it's per term per term so that's you know  a lot of people have followed up on that a lot  
11:29
of estimates a lot of estimates coincide more or  less so another political scientist from England  
11:35
his name is Ian budge uh he kind of summarized  it all by saying uh the most uh widely tested  
11:43
uh inductive law in political science is that  governments lose two to 3% percent two two to  
11:50
3% of the vote percentages of the votes points  of the votes per term so it's kind of a well  
11:56
established so the first thing we need to do  is is find out whether we can reproduce What  
12:02
Nan pman said so uh so uh this is an attempt to  uh reproduce Nan empowerment's work and we now  
12:13
have more countries more elections over a longer  period of time and here we go between two and 3%  
12:19
of of the this is this is what they lose per  term average all the votes that they the net  
12:25
vote right starts at vote and all the through and  then average the loss over all the terms that in  
12:32
office and you come up with two to 3% that at the  national level where they tested it and then here  
12:38
is at the subnational level is a little bit higher  right so the overall the the across average there  
12:44
you can see it's about well yes it's 2.7 so 3% and  here is a little a little more 3% so that's how we  
12:52
start right but we think that uh this is accurate  as far as it go is correct but we think it is not  
13:02
a true representation uh a useful representation  of the phenomenon we think that the it hides more  
13:09
than it actually shows so um so what we did and  this is uh I give all the credit to Richard here  
13:19
because I said to Richard I want you to look  at this data here because I I I see that the  
13:25
function it appears to be the little bump uh the  party is elected and if it's reelected it gets  
13:32
a little bump in the vote and then and then  eventually eventually the Dam breaks and and  
13:38
it kind of you know they they lose so he looked  at that he said you know I don't think that is a   bump and and neither do I think that is kind of  an incremental because the idea of a two two to  
13:48
3% per term gives the impression that there is  kind of an incremental loss right of the vote  
13:55
over the time of the the incumbent are in office  so I don't I think that's the way it works so we  
14:00
began to back and forth back and forth you know  how collaborative work goes and so what we found  
14:08
so uh so that's the next thing I'm going to show  you is well this is this tells you all the all the  
14:14
election that we have studied okay Al together  is over 2,000 elections 1,00 elections at the  
14:23
subnational level uh 500 elections at the national  level now spells uh spells is basically um the  
14:34
series of terms consecutive terms that a party  is in office right so for example uh president  
14:42
Ruth well it's not a it's not a person it's a  party okay so for example to give a long one  
14:47
it would be like the Democratic Spell election in  1932 and they were in office until 1952 when they  
14:55
lost so 20 years five terms okay okay so that's a  spell now a spell can go from only one term which  
15:03
is actually the most the most frequent uh to two  three four and up to many more terms okay up to  
15:09
10 even more in some in some cases some unusual  cases okay so we got a total of um almost 900  
15:19
around 900 spells and almost over 2,000 election  so we have a lot of data but we concentrate our  
15:27
analysis on the subnational because we have so  many data points and if you're going to establish  
15:33
relations if you really have to have a lot of a  lot of data points that is be right so our next  
15:41
slide okay so here's um on unlike the the first  slide I showed you that is a per term loss this  
15:55
one is a total loss okay total loss and so this is  the five countries we work we we have been working  
16:02
with the most and this at the subnational  level state or provincial elections and we  
16:08
began with Australia Canada and the USA then we  added Germany and then we add in India and this  
16:17
is what we find uh we find that um we find two  two things you you see in this in this slide the  
16:24
green represents elections re-elections I should  say re-elections the red represents defeat defeat  
16:35
comes at the end of the term at the end of the  spell that's what ends the spell when they lose   the election right they lose the last attempt of  reelection and so we see uh that uh in in a couple  
16:48
of cases the incumbents actually gain a little uh  during re-elections uh but the bulk almost all of  
16:57
the laws occurs when they lose the election  so one thing we had done with this to begin  
17:04
with is there is no such thing as an incremental  loss you know in a stepwise sort of step by step  
17:10
step fashion uh over the course of the spell they  basically they go and then and Rachel we're going  
17:16
to show show you more of that kind of St stable  and then boom like this that's why Cruise the  
17:22
election the the incumbent Cruise right and  then they crash they cruise and they crash
17:33
now here um in this slide we're showing you a  little bit different because first was the total  
17:42
uh cost now uh I want to show you the fact is  that different countries and different systems of  
17:51
government the two principle being parliamentary  versus presidential or as I call them uh cabinet  
17:59
versus executive they have different votes at the  beginning so a 10% Point loss when your average  
18:08
40% is a lot more than a 10% Point loss when you  get 50% and the presidential governments they get  
18:16
higher votes than the Parliamentary ones so what  we have to do well we have to standardize the  
18:23
total loss by uh their their their the average  of their wins okay so what we find uh is that  
18:33
basically when you do that on the average across  the countries is about very close to 20 20% they  
18:42
lose when they crash 20% of their average win  vote okay now that's when you hear what 20% as  
18:50
opposed to two to 3% Point per term when you hear  20% now now you get to see a real amazing you know  
18:58
it's a much more impressive number than two to  3% per term you're losing 20% of your of the  
19:05
vote that you won when you first got elected  or the average of the win vote when you get
19:10
reelected now this uh slide show you it's the  same proportional change right but now what we  
19:22
want to show you because the other one was kind  of static right they just show you the total and  
19:29
now we're showing you what happens uh when during  the course of a spell so spell like I say consist  
19:37
of a number of terms or elections right so in  each of these vertical panels you see that it  
19:45
says election from election 0 to 5 10 15 and  some cases that's about the maximum but very  
19:52
close to the maximum so what we what we're showing  here let me see if I can use the the the pointer  
19:58
now I think I can do this right there sorry  all right so see this line here this line is  
20:09
the average okay the same average that I showed  you in the previous slide roughly about 20% um  
20:17
they're a little different of course but roughly  the same 20% but what I'm showing you here is not  
20:22
just the 20% what I'm showing you here is that  there is no no difference on average uh between  
20:31
losing uh your cost when you when you uh lose  one election when you last only one election  
20:38
or you last 15 elections it's flat this is flat  there's no relationship if there were a relation  
20:45
if there were a relationship you will see this  going up the numbers the the dots going up or  
20:50
if there were negative relationship it will be  going down but notice some go up some go down  
20:56
so on average is flat what does that tell  you that there is no on average there is no  
21:03
different there it doesn't really matter when you  lose on average you you Cur the same cost about
21:09
20% okay now this one is the the last slide I will  show you I'm taking two turns the first turn then  
21:21
Rao takes over and then I come back so in this  one it's a little bit more complicated but and   I want to by the way I want to thank uh Michelle  Williams my colleague because she introduced the  
21:32
whole department to Tableau but I was the only one  who kind of you know really enjoyed it and and I  
21:40
have used Tableau you know it's a great it's  a great program to analyze electoral data so  
21:46
you know thank you Michelle it's taken me several  years because it's thr and error thr and error thr  
21:51
eror but anyway so what I'm showing here uh is um  a number of a number of things are shown in this  
21:56
very complicated but really complicated but really  illustrative uh chart all right so the blues  
22:03
represent the number of terms okay so you can  see and this is for the US only because we have  
22:10
the most data for the US so uh basically uh us um  governments us administrations at the governatori  
22:19
level they they basically the most frequent uh uh  terms are either one term or two terms so so you  
22:28
either only uh last one term or you last two terms  and then that's it then after that you just declin  
22:33
it like this okay and this this decline here is  something that Richard will explain uh with a with  
22:40
somewhat different chart so I won't I won't get  into it now okay so uh here that number there is  
22:47
the vote uh at the last election where they lose  right because you you serve one term that means  
22:53
you lost and when you lost you lost with 40% of  44% of the vote and and you lost 7.8 points what  
23:02
does that tell you that tell you that they they  had when they went into the last election they had 51% right so this is what they had when they  lost this is the the actual number that they  
23:15
that they lost that the crash uh which is uh  in in in in points not in percentage points  
23:22
uh and so so look look how look how this these  numbers don't vary very much right right so uh  
23:30
they when they lose they lose pretty much  about the by the same number of votes uh  
23:35
and when they get reelected about the same as  and as she and Richard will show you this but   here hints there are hints of that when you  look at this when you look at this chart so  
23:44
in the blues uh you're showing how many this  is this is the number of terms right so how  
23:50
many administrations lasted one term how many  lefted two terms and then how many lefted three   four five and so on and at the top what vote  what vote did they take in when they lost the  
24:03
election they they the spell closed and and how  many points they had how many points they lost  
24:09
so that's the vote they had they they came up  with that vote and that represented an eighto  
24:14
loss a 13o loss from the uh for the mean of of the  previous wins now what about this right well this  
24:23
shows you uh the uh the the proportional cost  so you can see that for the most part uh they  
24:31
go that was a little bit less that's a little  bit less but 20 19 24 22 18 20 so very very  
24:39
similar right so again there's no relationship  right no matter how many times how many election  
24:44
you survive when you when you when you crash you  crash by about the same percentage of your vote
24:52
okay and I did this for the cabinet uh for  the parliamentary government the same thing  
25:00
you can see the same thing here not only  the Parliamentary governments you can see  
25:05
that one term is the most common right  uh but note that in the Parliamentary  
25:12
governments they they uh they lose with 30%  of the vot 31 32 33 very very similar very  
25:19
very similar there there's no relationship  here there are you up and down up and down   but basically uh and and again here the the  the crash this is the this is the the percent  
25:29
Point crash is right there and that is the uh  the proportional crash right uh the the the  
25:35
one that we think is that the the represent  the best way represented the true cost of of
25:41
governing and this will come later so  uh I'll stop now and I'll come back  
25:50
later when richer has done his  presentation thank you for your attention
26:05
I'd like to thank Al for inviting me uh as you'll  quickly figure out uh I'm a civil engineer I know  
26:11
nothing about political science and uh but I  think that's to my advantage in this situation  
26:16
because the less I come in with maybe by some  kind of odd coincidence the more I might see so  
26:22
why don't we have the first slide that'll just say  that I'm doing there we go part two my name down  
26:28
there continue on El and let's have the next  slide right away uh this looks formidable I'm  
26:35
going to simplify this way down okay I'm going  to ask two questions which were the questions  
26:41
Al was talking about winning the next election  and then one of those you went by I'm going to  
26:46
employ two statistical distributions which will  be very simple you say oh my goodness I didn't   you don't need to understand these things the  data I'll talk about where we got all this stuff
26:57
yeah technology challenge here let's try this  thing work okay uh where did that get all that  
27:06
stuff he tolds you we'll see uh what data do  we need to pull out of it we'll see fa simple  
27:13
statistics uh how we're going to test it so you  don't need to worry about this but this is kind   of my uh plan of attack so why don't we have next  slide forward one oh have the first spells I'll  
27:25
find show spells so you your I want you have a  picture I don't want you to remember what he said  
27:31
I'll talk about exponential distribution would  you say what's that I'll say if you ever heard   of radioactive decay said oh yeah you know  that one already something that we developed  
27:43
that would be extremely engineering L is take  a complicated question part of it and bring it  
27:50
down to a small diagram to focus on what you  want to see okay so and you'll see what that   is immediately uh and then uh I'll progress as  kind of allows how we turn that into actually  
28:01
a count okay next slide okay this first slide go  clinical science B it's a clinical science slide  
28:11
a picture uh nothing wrong Ben Fronton stole it  okay this was actually drawn by Socrates and a  
28:18
student a Dre this these are this re discovered  underneath in the pyramids I Believe by Harvard  
28:25
up archaeologist okay Socrates was one of ideas  and so his idea was you have his P oops we don't  
28:34
want to get that yet we have this complicated life  okay uh let's link all these things together so he  
28:42
was big on don't worry about reality as much as  the real thing it's just the best you do is get   the ideas together um his student Aristotle  had the same problem same question here but  
28:56
he was much more the first he wasn't a scientist  yet but he was much more let's look at some data  
29:01
let's look at the reality let's cut some little  pieces and see what they look like let's find out   why toads are different than caterpillars okay  that bother howas and all okay so uh there's Al  
29:15
right there join it together put it in a big  picture uh and if you say the big picture the  
29:20
details will come no argument on my part at  all okay but here's the engineering side of  
29:26
the thing is to do that it can be beneficial  to back away from it chop into little pieces  
29:32
not under picture study little pieces and then  see what big picture looks like so we're going   it takes two to two to time it takes two to make  research work it's not one and not the other okay  
29:43
let's next slide uh I brought several show and  Tails uh first which I'll show the right right  
29:53
time I brought a weighted coin you don't know but  I glued a washer on one side okay and I bought a  
30:01
dice I made two colors Two Reds no no two Ls  and four RS okay that's for part of it after  
30:09
I do that that's for like part A going to get  into some statistics uh to do that I brought a  
30:18
bell-shaped curve so you all know what I'm  talking about I'm going to cut it what you   don't know about is that I'm going to cut it in  half okay anding back there that's never been  
30:29
done before never heard to anybody why would you  want to do that to me do it and then a balance   dayti so that'll be care statistics okay but  to talk about spells I went back to that uh  
30:42
uh definition I gave you and let's and and so uh  let's remember this guy I I actually brought one  
30:50
of these and I brought a meat cleaver I cut it  into sections but the thing got loose and so it's
31:01
okay next slide okay these are next alls okay  these are the sub Nationals would be provinces  
31:11
or the equivalent in whatever country you're  in uh you know Queensland or something Canada  
31:18
Alberta Okay and like like that snake I saw these  are the spell histories with the that you can't  
31:27
read the uh uh 20 election so in so whatever this  says I 1 2 3 four five they have six spells in  
31:37
about 20 elections okay Canada is okay just for  what it's worth if our spell started and it was  
31:50
open to the history before we don't use it okay  because we don't know how long it was there before   it's better just not use that spell than to say  us started today okays were in at our data set  
32:02
if if somebody was in office and hadn't had been  there and hadn't lost we don't say well you lost  
32:08
now we're going to ually we just throw that out so  my home state of Oregon uh lost a Bunches of data  
32:14
because we haven't had a republican governor  for since 1978 okay so we don't get to count  
32:19
that as a long run okay Florida loss because  whenever florid had started and I what it was  
32:24
you had a long string of previous of the same  party Democrats I don't okay so we could have  
32:34
more SP want let have next uh we're going to see  Germany here in a second uh a few German few of  
32:44
these are pretty bizarre some of these countries  just some of the provinces in German especially I   think the ones that have more traditional kind of  conservative V to them they just sat these guys s  
32:55
with the same party for 15 uh 15 years one spell  for 15 years okay let's go to the US my treasure  
33:04
this is half of the US states and you'll see  I back the snake picture you'll see a lot of  
33:10
snakes this snake well it'ss better U maybe this  snake that looks kind of like that red and black  
33:17
I had uh because a lot of are pretty back and for  look have an excellent second half like you uh
33:27
well last show showed the rest of the states a  lot of data in those States because every spell  
33:36
has its own history okay uh those were Subs we  did the same for Nationals the presidents the  
33:43
Prime Minister sort of winning party things  for 23 countries uh four of which we also  
33:49
had the previous subnationals so we'll compare  those so a lot of data let's have the next one  
33:54
this 23 develop uh countries okay I pled it out  okay I I'll show you a picture in a second but  
34:04
I'll show you uh what I did take all the Spells  let's put all the US spells together how many  
34:14
two how many one spell governments did we have a  whole lot okay how many two spell governments did  
34:24
we have before it changed not quite as many how  many three spill well not more number gets down  
34:32
okay now when I ploted that data I'll show it  you in a minute I got a curve that looks like  
34:38
this in the back of my mind I an exponential  exponential well actually if I smarter I would  
34:44
said geometri geometri that's these are not  continuous data but that's a small Point uh  
34:51
you know how that curve works and we'll see we'll  see if it works for our data I'll show you in a   second uh r activ know I know it from hydrology  I work in the my rad and butter is is interface  
35:03
between hydrology and hydraulics give me a good  flash flood I'm just happy so are if I was here  
35:10
I would be working heavy duty on floods from  per but in the Mexico flash down people few  
35:21
uh this is a this is this is a uh overtime  of the rate of infiltration over here this  
35:27
is like Ines hour so if you watch a tu suck the  ground this go slower and slower over time it's  
35:35
called a Horton equation it's a great example  because there's no theortical basis for this  
35:41
there's no natural law that says it's supposed to  be that way it's not that's that's the in that's   the mix of a whole bunch of physical laws  with r okay uh but as we're going to make  
35:54
the point we don't have to have a EXP why we  have a curve the curve shows what's going on
36:01
uh uh games of chance okay uh classic how many  times are you going to have a chance of getting  
36:14
a red card if you just figure that out how many CH  you get a chance of getting a full house all all  
36:22
those are just and they're not all exponential  but the the red part of them would be or here  
36:27
the Family Planning what your chance of having  your second kid the same gender and then your   third kid again the same gender it goes down to  chapter two half okay U one paper that I found uh  
36:40
written in Italians I don't know what they said  seemed have bit on the subject and that's all we  
36:45
found so I'm simply going to say uh there wasn't  really this interest in this expon exponential  
36:52
distribution in political science Arena slide okay  uh first the math all you need is this equation  
37:03
here you don't need this is if you're going to  do more mathematical derivations forget it you   don't you're me to do that okay and it just says  give me a k 7 you say okay if I have 100 today  
37:16
after one one time unit which is an election  cycle I'll have 70 left if I want do another  
37:23
uh time cycle turn two out I'll 7 7 I'll have 49  left if I go another go three times Cycles I'll  
37:31
have 7 to 49 which it's a tad under 35% left and  then down into the 20s that's cutting down and  
37:41
so the point7 I gave you which which is a number  that we're going to sort of see ex like that the  
37:48
bigger the K the slower the K this says you have  90% left at one time out this down here says you  
37:55
have the 20% less okay don't worry about this I  worry about it you so if you understand what pay  
38:04
is it's proportions step to step um go other way  uh you don't understand that on that let's see  
38:12
the next one more okay here's the data these are  the US subnationals I like that data set it's a  
38:19
big data set okay this gets really iffy when you  start looking at these little runs of seven data  
38:25
points you hundreds six uh when I made this there  was 600 uh uh spell start to th this yet okay  
38:37
everybody wins the first election that's what we  call ezo okay how election two or the after your  
38:45
first reelection is the second bar okay this many  one that okay so everybody is still there but some  
38:56
now away one more election this many want it that  this is real data okay what you can't do is just  
39:03
compare these I just want to put a little teaching  moment here never take this data and just say well   let's just all the the ratios between Bar number  seven and Bar number eight you'll get garbage like  
39:13
this okay the reason it's garbage is because when  you get down here you're talking about very few   points okay and so one ER one get B this guy uh  uh uh this guy this this docum have four values I  
39:28
think 45% so we don't do that okay just point out  that's your temptation to do that let's have the  
39:36
next slide U this is that same data you just saw  but now it's been normalized so it accounts for  
39:43
100% area under here that's just so we can compare  different countries or different governments  
39:50
okay uh that's called a PDF probability density  function uh uh the the distribution function is  
39:58
this thing you don't need to worry about it but  you often see it uh uh polished out this is the  
40:06
one that starts at zero and go up to one okay next  slide we don't we don't need that we can use it  
40:11
but we can get by uh if you take that data and if  you plot if you just all do hit the log function  
40:19
on your calculator put log in Excel and you  plot that same data of the remaining if it falls  
40:26
line you don't know you do it by do you got an  exponential distribution you hit it you're lucky  
40:33
I made a good guess is by luck I didn't know okay  that's the K that does it's the slope of this line  
40:39
okay again you don't need to really do the math  but it's extremely straightforward if you go back  
40:45
and plot what we predict with this equation which  is the smooth line against our data points we did  
40:53
pretty well so and and we did that for everybody  okay the US I like this is p on because there so  
41:00
many data points there's not a lot of mess in it  let's next slide okay that's what that's what the  
41:07
US that c I'll show it he had a bunch of bars look  like this it showed the story but I got it down to  
41:15
one equation this just says every step out every  uh for the second spell uh you you're going to  
41:22
have 59% of what you had in the spill for okay so  your surviv rate is 59% okay election to election  
41:32
does it change your let's find out okay let's  keep going uh so what I'm calling the K is just  
41:38
the slope the 79 whatever 52 OR7 example this is  what we got for everybody we looked at now I don't  
41:48
really first of all there's obviously some odd on  at the end I'm happy just to say away because some  
41:56
of these countries had some odd histories okay  but it is kind of interesting they do kind of all   F in a family and our subnationals all the brown  ones are National Data some Nationals which have  
42:07
a lot more data actually fall in pretty well so  it just makes me kind of feel good that we're on   to something that isn't just good for National  and not subnational just subnational is it is  
42:17
kind of a feeling and if if if you want to argue  well this one here says USA is less than Canada uh  
42:27
probably true uh but I some like Al say well why  that like this is what it is okay so that's that's  
42:34
the the proportionality proportionality constants  is what those are right over there next slide if   you please okay this is plotting them out okay  now they're just ploted out alphabetically down  
42:46
here so that's our significance the bottom four  are the four subnationals they have much tighter  
42:53
degree of confidence because we could squeeze it  in because we have so many more data points and  
43:01
the more data points you can have the more sure  you are about squeezing in if you have seven five   data points you get some of this strange stuff uh  they are real erratic data Bo so that's just an  
43:14
issue so this is you know these are all the this  you can't read it that goes Australia you know and  
43:21
just going down Canada Germany USA India to late  add would be another line there okay it's good for  
43:31
comparing uh is this bigger than this because if  the confence intervals start to overlap you start  
43:37
to think they could be from the same uh data set  and the cont intervals here is a tough one because  
43:43
these are exponential and you you can't go by  the old uh 68% plus plus or minus one standard  
43:50
deviation 68% 95 plus that doesn't work this  is not this this is that slow down for shape  
43:58
but at least gives you a sense for where we have  those confidents next slide uh okay well maybe I  
44:06
don't want to do the curve fitting maybe I just  want to look at L spreadsheet and pull out the  
44:11
spell length and see what how that agrees with  okay it's actually a pretty good agreement it's  
44:18
curve in here okay but it's that that's false uh  math okay because the reason is uh the for a given  
44:28
well just put this way they're not independent  of each other so I don't take any I take conation  
44:34
that I didn't lie to you but I don't think I  proven it let's try the next one uh okay how  
44:40
often did they win that' be another way to guess  what this proportionality is that's not a bad way   to go about it but how often you want over history  doesn't really say what your rate was for this is  
44:50
what you average over history so whereas it falls  on line here uh knowing win ratio somebody W 70%  
44:58
of their doesn't mean win in their relection  with a 70% chance maybe the chances increase   and then decrease we wouldn't know that's just  another way to get K I'm just saying K could be  
45:08
I garnered several ways okay but I I didn't that  let's next slide please okay let's look at that  
45:16
snake anybody seen it it's it's red and white  okay I I did bring a cleav so I got scissors  
45:24
so we can go out okay let's chop the case apart  that's where it gets interesting to me okay six  
45:31
I had to look up them backs they're okay this is  a madeup spell okay but just for illustration okay  
45:40
the com vote uh there's got to read this there's  50 right across 50% right there so this guy came  
45:48
in this part in Democrats of Utah in 1941 whatever  uh came in with 54 % then they went up they went  
45:58
down and they kind of bounced around what Al was  talking about as the cruise but this way we can  
46:05
say what the cruise was and then crash with my red  line uh two ways to look at it uh the red line is  
46:14
how much they crash from the last election which  I say it's bad statistics the other is how much  
46:22
they crashed from the mean election the reason I  say it's bad statistics to uh go for that one is  
46:28
that everything including this everything here  is basically randomly bouncing around and this  
46:34
could just easily this this this point could  have been down here and this point could have   been up there it looks like a Bigg crash much  better work from the bean the mean has also a  
46:42
smaller uh variance itself because this this mean  is the mean of six numbers so it's it's better by  
46:51
that two two and a half squeeze down VAR so it's  just better okay so this is what I call the I this  
46:59
is what I would call an engineering diagram of a  election cycle okay let's look this startom line  
47:07
okay I was talking about the average if you  just said what you came in with what you went   out with I came in at 54 I went out at 48 I did  it in seven spells that's a long long long last  
47:20
I had but I finally went out you know by the  the difference by and you did two and a half  
47:27
points per spell fake number you didn't lose half  s per spell what you might think if you had gone  
47:35
by the standard understanding okay so we didn't  like that quoted in his book I said you got it  
47:46
wrong chapter the next revise that chapter um  here it is I just thought this out this is say  
47:57
okay we had concept spells of this is say t uh  crashes I'm sorry of T and didn't very too much  
48:04
I'll show you some of that stuff okay well if you  lose it all if you go out the first election you  
48:10
went down at n or this is the US day you went  down to 8% per ter because you lost 8% of your  
48:20
your votes uh you lost 10 wait okay you you lost  all at one time this this way if you didn't lose  
48:29
it for until the second one it halfs your average  okay so now you're losing four points per turn if  
48:36
you lose here this turns out that if you get  into two or three which the common rule says  
48:42
you're probably about right if you lasted in this  Zone here three four three or four ter and that's  
48:50
what our rule really says it is when we look  at the average length of a government and we  
48:55
find out how much they changed overall we get two  and a half that's how they did their calculations   I don't like never right okay it's what I call  a 2 or 3% law okay it's by no means a law okay  
49:09
okay it's it's not only a law it's a detrimental  rule because it makes people think there's a law  
49:17
there is maybe a law but it's it's the law prison  and crashing it's not about constant rate let's  
49:23
have the next slide okay and this is what we  concluded what Al said uh the cost I'll say 10  
49:31
where changes the country is independent of how  long you're in office how many spells you goise  
49:36
along about the same rate bounce around for sure  but you're always winning so they you don't bounce  
49:41
too low and they probably don't bounce gantly high  you know 95% okay uh and then at that point they  
49:49
say the same crash same PTY as the ones that crash  before even those that crashes uh SE so that's the  
49:56
Val right here and this this is probably our first  finding it's actually a meaningful finding not the  
50:03
number we get let's have the next slide okay so  let's look at the wins and the losses those two  
50:11
numbers for sake of argument I'm going to say  you cruise at about a 2% above the required 50  
50:18
or 49 or whatever it is 4% find your remember you  crash at maybe 10% or 20 well that uh okay Vats  
50:29
are uh uh fairly well winning sorry incum votes  are pretty well normally distributed not really  
50:39
well because there's a lot of curtosis there sub  uh not things are squeeze this this should but a  
50:45
little quite so high and then come down put a  normal distribution that'll B us a little bit   of the end with it okay if you P all that out  this is I'll have this think blocks okay so I I  
50:58
certainly don't assume you figure that out just  thinkig about what each block was I try to get   each I still have over here it's discre I have it  on the even or on theer spells uh have awful lot  
51:14
of one spell elections uh a smaller number of two  smaller number three down to and I if this is this  
51:21
might of New Hampshire or somebody uh somebody  lasted for 10 SP but not okay if you plot a slope  
51:29
through this you might say you see something  okay well maybe you do but let's do this let's
51:39
take my normal curve that thing up right  there and let's just say I actually I made  
51:51
a note for my own self at 52 it could  be any place uh you you want to get the   winners are on this side the losers are on  that side so I'm going to change a different  
52:01
52 I'll make it more like 7 okay just chop  them in two same distribution use two hous
52:12
of these are the losers these are the winners  nobody wins if they're down here nobody loses if  
52:21
they're up there each of these are distributions  they just happen to come together like this but  
52:26
this is almost like brary distribution this is  sort of like a distribution I guess I want to  
52:34
look at these two distributions okay so I want  to look at the two Hal so it took us a long time   to to figure out except X we ought to be look  our statistics at just the winners and just the  
52:46
losers well here's where you really see the uh  stationarity of it okay that slow is basically  
52:56
de FL the significance of significance of the  slope is n there's mathematically there some  
53:04
little number there but beta you put that in  is zero 7 x okay and the losers all lost but  
53:14
again there's a awful lot more than lost at one  point by one point by 10 points if if you run a  
53:21
agression through all those points it's veryy  flat the winners win the same the losers lose  
53:29
the same no matter what the election okay again  I'll set the question is how can we turn that  
53:40
into uh trying to get votes because I've never  used the word I've got have said vote I've never  
53:47
used a vote in anything I've said so far than  just explain winning and losing votes I haven't  
53:53
given any numbers okay let's have the next slide  okay there's here's here's that same mail curve  
53:58
a little more explicit there's a normal curve  statistics at least social science statistics  
54:04
has a great dependency uh everything and I don't  have any argument with it but so it's probably not  
54:13
quite a wellsh that's the curtosis I was talking  about this is just what history has given us in  
54:18
US subnational elections okay a lot about 50  okay uh there there's the split I was just  
54:26
talking about I think I split it at 52 because  that's what it was so that's the two sides that  
54:31
I just showed you winners and losers let's have  excellent let's look at the winners okay so the  
54:39
question is what's this word involes you know this  is these are standard deviations this is just a a  
54:50
standardized let's say V down here uh you got to  know what the ciation is in the boats okay so what  
54:57
is this how to solve that well you might have  ideas and you might be right on track okay but  
55:03
I'll tell you how an engineer would solve it we  would take this thing we' say I won't do it very
55:10
well close good close where's there I go where's  a centroid of this shape I'll just move my thumb  
55:22
it's right there that is the centroid and you  how I found now I can find it mathematically  
55:27
no problem finding mathematically but if all else  failed I can bounce a thing on my finger and that   would be the CID that's all you need to know  okay you know what the centroid of a circle is  
55:37
it's the midp point you know what the centroid  of a rectangle is centroid of a triangle is a   third of the way over from the long side this  is a weird one okay because this curve here  
55:47
is the math is is more complex but it's system  equation okay likewise so so like there a CID  
55:56
on okay this is that other side of the curve find  where the centroid is mathematically I'll get the  
56:04
right answer with my finger okay right there okay  found two cids this will take a leave of faith on  
56:13
your part to say why trust an engineer on this  stuff okay fair question okay but you trusted an  
56:20
engineer presumably because you're all sitting in  it to design that beam right there above you okay  
56:27
same principle let's have the next slide okay  equipent the point being I can take all the  
56:33
weight know I'll to stick with the green line the  winners on the green side on the winner side I can  
56:41
make all this which before I had my balance beam  let's just do just you want to see it so let's do
56:51
it there how we're going to do it okay  just let's see let's get like this okay  
57:00
just assume it's a balance being but I I don't  have the rest of the balance there so I bing uh  
57:06
you want to see that's all pushing down it's  like a whole lot of people sitting upstairs  
57:12
here they all go to the far end to look at  a the pradey going by it's going to change  
57:17
load on that be okay well you want to design  that beam and you don't know okay what you do  
57:25
is take that area which went from where my nose  is to where my nose is with all the way you move  
57:37
it over again it's all mathematical if you do  it mathematically but you don't need to do it  
57:43
mathematically rather than go from here to here  you Mo move it over to makes the same balance  
57:50
okay that's the equivalent location you do that  on both sides okay this is what I'm calling the  
57:56
equivalent vote equivalent meaning Engineers  call the equivalent Force okay so we'll call  
58:02
the equivalent vote because I don't know manyal  science use that term I don't know but no they all
58:09
will this is equivalent vote okay so well do right  here the equivalent vote of the winners is about  
58:20
um6 standard deviations over the equivalent vote  the losers is about one cation the other way well  
58:28
if you know what cations are and you know where  the middle is you know those points okay so at  
58:40
numbers okay actually go go back one okay if you  please okay my question now is how far is it from  
58:53
here here from average this is the main average  win okay think all the winners what they win with  
59:00
what's their average take all the losers what's  their average what's the difference here if you   want the two points that's good too but the big  question is probably what's the difference okay  
59:10
this is just statistics this is not votes this  is just statistics okay from here to here is a  
59:16
known distance when you know where this line is  and you found this croid and that croid okay so  
59:23
now the next curve which you just have to take  my oh we going there okay where I call the B  
59:30
differential curve okay and all says is if you  solve the thing I didn't Numerica because I do  
59:35
not think a post form solution exists but sorry  my tell me it does okay if you put in a c find  
59:45
the two centroids okay and then subtract the  big one the small one from the big one you you  
59:54
get 1.7 well move down here midrange okay  you get these differences okay this is the  
1:00:06
differential between the losers and the winners  okay uh and so so those are just translated so  
1:00:15
with that and the math here is just substituting  uh standardizer numbers for uh actual constants  
1:00:23
for your dat which are the mean and the ciation  get so let's see how that work now let's go next  
1:00:30
one almost done here okay we take the course  subnationals okay we have the real data okay  
1:00:37
the dots show the real data okay our model  shows what I just showed you with equipent  
1:00:44
votes okay well you know I didn't get it great  uh but I didn't fle anything either uh and the  
1:00:52
OB is not to hit it exactly it's to see if you  got the right idea this looked like we were we  
1:00:57
probably closed in pretty well on the uh uh  losing votes but we were underestimating the  
1:01:04
winning votes okay this said the winning votes  for for the US was like 58 57 but really our  
1:01:12
our model only showed 54 okay we missed okay  but that's fixable okay at this point you're  
1:01:18
just adjusting your model and the way to there's  several ways to adjust it uh the sophisticated  
1:01:24
is beond me but I could figure out how to do  it would be go back and change my distribution   it's not bell shaped it's really cized Bell or  whatever you might call that there's probably  
1:01:34
name for it uh I could probably find a work for  the other is just apply a Conant to it and say  
1:01:40
well we're off by this many points but we have  to H and all sorts of equations starting points  
1:01:46
let's do it let's look those constants oh this is  this a dimensionist it basically it shows simply
1:01:58
our model doesn't doesn't quite do okay so skip  okay if you put in those correction factors  
1:02:11
we'll hit it exactly I'm saying let's not go  there our objetive is not to hit it exactly  
1:02:17
our objetive is to see if we're on the  right track and then think of because I   don't believe it's rational irrationally  finding the close it it looks like if we  
1:02:26
kind of subract or two from our means and uh  oh subract two from everything to start with  
1:02:33
we might do a little bit better I I AR those  numbers those numbers make it work okay and  
1:02:39
that's where we're at okay and I don't really  think we're going to push it to make our model  
1:02:44
hit everything this is po countries so we at in  end here something else weird about them um but  
1:02:51
uh that's that's where I'm going to leave it we  started out we chopped up the snake is it must be
1:02:58
somewhere SL right around your foot c s looked  at the eye curves and you say well eye Cur was  
1:03:10
so obvious well obviously the the team that came  up with 2 and a half% on average never looked  
1:03:17
at I curs they never chopped it up into little  pieces they just said oh put them all together   that's what we get okay you look at iurs then you  start to see that Cru crash if you start to wonder  
1:03:28
why Cruise so much of why you crash we Cho up a  bell curve to explain that did pretty well uh we  
1:03:36
might do better we we might not even publish this  this last bit it's pretty spec but if we do uh uh  
1:03:44
I just asked Professor bu back there he says I  think he's ever seen that done before ask you've   seen it introduce there'll be all statistics  book they go through all the hle curves of  
1:03:55
distribution so may add one to that but I'm not  to count on that yet okay so let me turn it over
1:04:00
to yes I tell you I had never seen Richard lecture  before uh we have been on the phone countless  
1:04:11
hours we have gone through innumerable drafts uh  and argued sometimes as I think I it going ATT  
1:04:20
test but um but I never seen him lecture before  uh Richard is is just really great uh he he ask  
1:04:28
humor and so on anyway I just want a a couple of  more things uh as you can see we have the cruise  
1:04:35
and we have the crash and we have the flipping  of the coin right the bias coin the bias coin is  
1:04:41
K which varies from about a little bit about  say 045 little bit less and half to about 75  
1:04:50
that's a very by VAR by country so that's that's  the the play on the coin now here just a couple  
1:04:56
more things I want to show you uh this is actually  looking again the subnationals uh divided between  
1:05:02
the parliamentary or cabinet and executive  and this is the um spell length again back  
1:05:09
to the spell SP spell by uh party ideology near  subnational and party ideology is shown here by  
1:05:23
sorry uh party ideology is shown here by colors  so uh a party that is on the left of center all  
1:05:29
the way up to minus one or right of Center all  the way to one sometimes I have to kind of 0  
1:05:35
25 and so on 050 and anything in between like  for example Sometimes some Regional parties or  
1:05:41
parties whose ideology is ambiguous uh so that  would be a yellow so you can see that basically  
1:05:48
parties alternate through the entire um time  here uh right right left right left right left  
1:05:56
and this conforms to something that Ian budge  said in one of his books he said that this is  
1:06:02
what democracy requires it requires that there  be frequent alternation between parties one on  
1:06:07
the right and one on the left and this then on  balance it gets the policy right down more or  
1:06:13
less the middle okay now this is the last thing I  want to show you here and that is basically weend  
1:06:19
began to extend this to the U to the to Latin  American cases we found three cases Costa Rica  
1:06:26
Dominican Republic and Uruguay that have a  good democracy again very few observations  
1:06:32
but nevertheless you can see the uh Costa Rica  uh pretty much online with the others and so is  
1:06:39
Uruguay and uh Dominican Republic has a higher  thing all right uh well I think we have covered  
1:06:47
quite a bit of territory uh I just want to uh  say a couple of final words I want to acknowledge  
1:06:54
it is sort of um it is uh sort of U conventional  right customary to to thank people who have helped  
1:07:02
you and of course uh Richard uh is um uh above all  uh but also I have uh uh Mike bundrick who's back  
1:07:11
there somewhere uh Mike what do you you please  stand up please uh Mike uh was wonderful uh he  
1:07:16
and I worked together uh many years uh on presid  American presidential elections which got me into  
1:07:22
elections and eventually uh to this to this work  and then I want to acknowledge the uh the comments  
1:07:29
and and so forth by Ian budge KB Clark John carry  Joseph Coler who comes next Adan lipart rain taga  
1:07:39
uh and Christopher valis of the University of  Texas and I also want to thank I always thank  
1:07:46
my colleague Michelle Williams but also graduate  students uh we've had I've had a very fortunate  
1:07:51
to work with a several cohorts of graduate  students over the last five to 10 years and  
1:07:57
the I have acknowledged all of them uh in previous  publication but uh this year is Carter Edwards and  
1:08:04
and Grace wheer and they have done really a really  excellent job really a lot of drudgery you know   sort of proof reading and proof reading and proof  reading data and finally I want to I want to uh  
1:08:15
acknowledge my nephew Alfred Luiz who did the two  pictures of the cruise and they crash the cars so
1:08:26
good evening and thank you for this invitation  I've been told to make some comments about  
1:08:35
Professor Kuan and hean findings and also some  suggestions based on my own work in order to  
1:08:45
try to First understand what they did we had  a lot of conversations before Emil in person  
1:08:54
I'm not yet completely sure that I underst  everything but I think more or less and  
1:09:01
then some suggestions on my own okay on the  same topic following up from their findings
1:09:06
findings so uh let me say first of all  there's a lot of interesting empirical  
1:09:18
findings here okay so the I may I I read a few  versions of your paper so perhaps I'm quoting  
1:09:27
something all but the total mean percentage  points lost during a spell in office is near  
1:09:32
9 9 percentage points at some moment which  means on average 1/5 of it its initial vote
1:09:42
uh the rection rate at some moment in your papers  was 60% which means that as I understand well 60%  
1:09:53
probability to be reelected means that the SEC  for the third term would be 60 * 60 so 36% and  
1:10:01
so on right as an average so which means all  of this means that as you said we can expect  
1:10:08
between two and third and three terms on average  for 8 to 10 years or something you you made much  
1:10:17
more precise uh numbers here but this this ve and  then uh what what T my attention more is first  
1:10:30
that there is no significant differences between  parliamentary and presidential regimes that's  
1:10:36
very interesting because I work institutions and  this means that you must look at something which  
1:10:43
is not exactly constrained by institutions to  explain this regularities okay and another even  
1:10:50
more interesting perhaps is that ideology does  not matter so for both left parties or right  
1:10:57
parties they have about the same numbers same  probabilities and same performances regarding  
1:11:04
winning or losing elections which is still even  more interesting actually um so you say regular  
1:11:11
alternation between parties of the left and  the right about once per decade on average   is a feature of democratic politics well that's a  good finding I mean important then when professor  
1:11:23
told me first time by email the paper Cru and  and crash I I thought that I understood that  
1:11:32
I didn't at the first my first reaction was well  this looks sounds like Hemingways coming about a  
1:11:41
friend uh who go back R and no he ask him how did  you go backr and the said two ways first gradually  
1:11:52
and then suddenly and then and they said no no  no it's not this uh it's uh it's more constant  
1:12:01
and then crash okay and then and then I said  okay so that's like dying everybody dies uh  
1:12:09
when you die you dies completely right uh and  I said no yes but but because this would imply  
1:12:15
some deter gradual deterioration more or less  right uh and he no that's not the finding okay  
1:12:22
uh then is when I understood the metaphor about  car crash a car crash can happen at any moment  
1:12:30
at the beginning of the Tre or at the middle or  at the end of the Tre right so that's the idea okay these are inductive empirical laws
1:12:41
regularities so you found some statistical  regularities out of so many cases uh about  
1:12:52
how parties lose wi elections but uh I think  if you want to be scientific so to speak we  
1:13:01
call no we call as political scientist do  political science and science has the same   rules for every field okay uh in addition to  empirical data you need some logical hypothesis  
1:13:16
to explain what why it is happening or how  why it's happening this way and uh uh so I  
1:13:27
think you have to so some people say uh the  scientific study of politics should work on  
1:13:34
two legs logical models and empirical test  about the consequences of the the expected  
1:13:41
consequences of those model to see whether they  fit or not or they are helpful or not to explain  
1:13:48
real facts so uh my idea is that your empirical  laws inductive laws as you say uh are still in  
1:14:01
searge of an explanation that's my feel about  this uh so an hypothesis that from which you  
1:14:12
can derive deductive implications to be tested  empirically your data that's that's what I can  
1:14:20
say so the point is uh we Ober for advance of  mat democracies in periods of Crisis recently  
1:14:32
they are not performing performing very well and  you can observe a lot of political instability  
1:14:38
frequent government or NOS I would provide a  few observations about it that I think should  
1:14:44
be included in this discussion to make sense of  the findings because of course so everybody can  
1:14:54
crash okay but it's not the same to crash after  the first attempt to be reelected that to crash  
1:15:01
up after 10 reelection you say it's exactly  the same crash okay but why it happens after  
1:15:10
one term or after three terms or after 10  terms that's a very interesting question I   think and even more generally if you want to be  a little more philosophical uh why government
1:15:25
fall at all why are not reelected indefinitely  if you say democracy is the less bad form of  
1:15:34
government uh well the Coman government is  doing something that people can appreciate  
1:15:40
should be reled again again again uh so something  happens that is does not fulfill the expectation  
1:15:49
about how democracy should perform right so  my my point is uh the government's fall but  
1:15:59
the question is not only how and when but  another question must would be why uh why  
1:16:08
some governments crash at the beginning or  after a few terms or very long term and I  
1:16:15
think it's an interesting question question  and then uh as you uh in your invitation said  
1:16:22
well you introduce some uh discussion about my  own work I'm going to make a few suggestions  
1:16:31
about it about this question basically my  idea is that governments fall because they
1:16:37
fail and what means they fail that basically means  that they don't fulfill people's expectations okay  
1:16:51
which may mean that they do a lot of things  but in some cases the expectations were too   high in some other cases expectations very  very low and then it has different effects  
1:17:02
on the chances of the government to get Reed  and then you have some hints in your papers  
1:17:09
in that way that I think you could explode so  one is ideology does not matter at I mention  
1:17:17
that's very interesting so regular alternation  between left and right views is independent
1:17:24
on regularities at once per decade to average  which means that so all parties fail somehow  
1:17:35
once in government they don't fulfill their  expectations as completely in order to avoid  
1:17:44
losing elections so you mention this would  be interesting and even more interesting I  
1:17:50
think you didn't see you mention in the paper I  today in the presentation but it was one of the   last slides that uh no Professor made this first  attempt to examined other democracies in Africa  
1:18:04
Asia Latin America newer democracies and you may  you show a couple of three countries right but as  
1:18:13
I remember your drafts you said that in those  countries the cost of ruling uh is on average  
1:18:21
two or three times higher than in these M those  mature democracies that's extremely interesting  
1:18:30
we should look at this very well because makes  a big difference right two or three times higher  
1:18:36
cost of losing of ruling in order to lose election  attemp made so means that there is a significant  
1:18:44
difference about the performance of governments  in developed countries and in nework democracies  
1:18:49
and less developed countries that's another  thing and the other thing that I'm not sure   I understood completely Professor h k letter  the rate of exponential decay uh you said that  
1:19:04
I understood the meaning but uh but you said  that different in different countries price  
1:19:10
across countries then well let's look at it let's  look at the countries which countries is the rate  
1:19:18
of Decay is higher than other countries so  if you enlarge the the this observation for  
1:19:27
other countries and and you make it try to make  a distinction between different periods I think  
1:19:33
you can try to understand more things so why  losing elections happens in different ways for  
1:19:43
different parties in different countries and  different moments and that's my question let   me add a little more evidence that can help to  discuss these just two points based on my own uh
1:19:57
research I have a book here which is democracy  and globalization republics the first half of  
1:20:08
the book is about this more not exactly this but  I mean it's the crisis of them second par is has  
1:20:15
be more optimistic about things that can go  well in this complex globalized world world  
1:20:22
world uh but in the first part uh there's a lot  of dat about two things one uh one it was the  
1:20:30
traditional explanation about how governments uh  perform in elections included the idea that the  
1:20:40
the incumbent Advantage meaning the government uh  so the party in office has an advantage with they  
1:20:48
run in elections because basically because they  can use information or manipulate information  
1:20:55
about their performance uh so uh giving sence  to their successes and not talking much about  
1:21:05
their failures uh and and then in any case what  they promise to do next time is waited by the  
1:21:15
previous performance and people can believe that  is some positive likelihood that they going they  
1:21:21
they going to to comply with their promises right  in contrast the opposition can promise anything  
1:21:30
but is also an hypothesis because we don't know  whether this is going to be real or just blah   blah blah so this was the idea no the incumbent  government running for election has a has an
1:21:40
advantage well this Advantage is very weak  right now I would say the last 15 years in  
1:21:52
people can discuss this it may be longer May  a little shorter but I was um I was observing  
1:22:01
all these events especially since the Great  Recession which is exactly 15 years now much  
1:22:09
and then in this period in National Democratic  elections the incumbent government Advantage has  
1:22:17
almost disappeared let me just mention one  obvious case that you know uh Trump Donald  
1:22:25
Trump failed at winning an election from the  White House in contrast with nine of his 11  
1:22:35
immediate predecessors so the rule for 70 years  right or more was uh more 80 years or so it was  
1:22:49
the inum is elected and then two terms limit and  then the other party wins the following election  
1:22:58
well it didn't happen this time right um but the  US is not a singular case in this respect so many  
1:23:07
voters many places are dissatisfied with the  government performance and they are voting for  
1:23:15
The Dil unnown rather that the devil now okay just  to put it in this traditional terms which was the  
1:23:23
idea of in Advantage so even if the new party is  not now for any previous experience perhaps not  
1:23:31
very reliable many people are so despair about  the previous performance that they refer anyone
1:23:38
whatever uh say in Europe for several decades  the government parties won elections two to  
1:23:47
three times in a row on aage which is more  or less what you say but since the Great  
1:23:53
Recession of the last 15 years theic and now  only one4 of governments have been reelected  
1:24:02
that's once uh instead of being reected two  or three times now so only the 1/4 are Rel  
1:24:11
wats okay 34 are out of Government after one  10 the last and I met this in my but I met  
1:24:22
more recent counting uh for the last two years  say the pandemic period more less there was a  
1:24:30
elections in 17 Democratic countries and the  leading party of the eum government lost it  
1:24:38
B for election 12 times out of 17 so about 34  right which is new very different from people  
1:24:49
were used to see uh and then there were a few  that were running for fourth or fifth term and  
1:24:57
it was expected that they may do after so  long but there were uh eight that were run  
1:25:04
running just for a second term and six of them  lost okay so even stay room just for this is  
1:25:14
one thing so inum Advantage is Vanishing second  observation so add there's more political party
1:25:25
fragmentation uh some government parties are  in crisis new parties have emerged it's more  
1:25:34
difficult to build a parliamentary majority  Parliament regimes then the formation of Po  
1:25:41
government majorities becomes more and more  difficult I can give you examples of cases  
1:25:48
one measure that I tried to do for the last 17  that you mentioned in the last vide years uh  
1:25:55
is something that is called effective number  of parties we know which weights the number  
1:26:02
of parties by their sides so larger parties come  from small par well in the 17 countries haven't  
1:26:12
held elections in the last two years the average  for the previous three decades for the prev 30  
1:26:19
years was 4.7 parties votes and that's my for  the last two years has to 6.2 so 13 more which  
1:26:33
means more fragmentation right and then I give you  lot of examples Ireland Romania Spain my original  
1:26:43
country in Spain there were four elections in four  years 2016 to uh 2019 because after every election  
1:26:54
the the parties were not able to for majority  in Parliament to appoint the government and for  
1:27:00
another elction just try was even worse actually  and then they had to form a coalition that just  
1:27:08
one of the previous elections the the leading  party said that they couldn't sleep if he had  
1:27:15
to make a polition with that party and finally  he had to do it okay I was making F of this but  
1:27:23
this was a few years ago but now the record has  been beaten so in Bulgaria there have been four  
1:27:31
elections again in a year and a half okay for the  same reason and they couldn't make a majority and  
1:27:37
you we know Israel had been five elections in  a little more than 3 years and it's still not  
1:27:43
clear that this to pass pretty much right so  uh inreasing fragmentation ining political part  
1:27:50
fragmentation even the best govern countries  have problems so there there are some list  
1:27:58
of the best govern countries uh Germany was one  uh for 60 years or more uh I'm talking about 17  
1:28:12
years actually the rule in Germany was two party  Coalition governments out of now there are six  
1:28:21
parties in par but uh there were fewer before but  always two parties were sufficient to make they  
1:28:28
were all possible combinations the Christian  democrats by with the fre liberals the uh the  
1:28:37
social Democrats with the free liberals the social  Democrats with the greens or the grand Coalition  
1:28:44
between the Democrats and the social Democrats  which was the formula for three of the last four  
1:28:50
okay with back they have with the last election no  capit of parties brought a majority together which  
1:29:01
was more fragment and then they for for first  time they have to do a three party coalition
1:29:08
to is not clearly connected Social  Democrat but then the and some issues  
1:29:19
are on the it's not consistent  okay so even inmany they have
1:29:27
problem the other case could be they are used to have four
1:29:40
party and then the last time they  record it took almost a year after  
1:29:52
election will be able to form a maity in so  was okay and then if you go to presidential  
1:30:03
regimes no recent elections withu Chile  colia new ex candidates right left went  
1:30:12
to the second round new parties were  elected pres the Challenger Alles so  
1:30:22
fewer spells in your shorter terms what is Nots  uh inability which means that your rules about
1:30:36
bash the question is howly after thect and seem
1:30:57
so question is why differ countries
1:31:07
differents for the duration of
1:31:14
sucess well my IDE they they fail and they failed
1:31:21
because we are still more than we were used to in  
1:31:29
ofici and socially accept polies  on numerous issues that are not
1:31:34
settled for a long period of  stable democracy after World War
1:31:40
I agreements and consensus about some basic  policies in the US Special Economic Policy  
1:31:51
for policy s but in the last period uh  especially because the foreign policy for  
1:32:01
the world has not been dominant as dominant  as it Wasing the war now we see but issue but  
1:32:12
in the last 15 20 years at Le more there  have been a lot of domestic issues that  
1:32:19
have Ed I the public controversy uh course  about climate change about integration about  
1:32:29
TR about issues about gender voting rights gun  control many issu that are not settled there's  
1:32:40
no policy consensus between the and public  debate uh that those already existed some
1:33:00
but now in terms of relations  the Dom all Dom issu take much  
1:33:16
much more and and there no some of  these ISU is not clear which is the
1:33:24
of then uh so that's my idea uh governments  fail more a whole more often than they were  
1:33:47
used to because they failed due to a new  agenda new set of new issues that were not  
1:33:57
previously manag to create policy consensus  and then uh and then the crucial thing is uh  
1:34:07
many people many vs had better expectations  about the government's performance and now  
1:34:16
they are disappointed and that many  people are ready to vote for anybody
1:34:22
except so then let me go to my own findings
1:34:31
uh can you can you this first I assume many of  you didn't read my book if you some of you did  
1:34:48
it don't don't answer my question okay uh so the  the question is is not exactly the same it's not  
1:34:56
how governments all is uh a question question  that I think has some logical relationship with  
1:35:03
it can you okay that's don't and the question is  uh how satisfied are you with the way democracy  
1:35:14
works in your country this a typical question  in in some sorts that are do periodically in  
1:35:21
almost every country at least in Democratic  countries which are right right and then uh  
1:35:31
so especially the PE Research Center about  it right that that has done this this survey  
1:35:39
several times and then uh then there are many  different answers so how how much satisfied out  
1:35:46
of 100 are you with the way democracy works  in your country so my assumption is that  
1:35:53
being dissatisfied might be related to voting  voting against the government inum government  
1:36:02
okay that's and then uh I Collapse the sample  I mean number of countries in this three groups  
1:36:14
uh so when more than 2/3 of population are  satisfied with the way democracy Works would  
1:36:23
be number one group one is my group my group  when about half of the population between 1/3  
1:36:30
and 2/3 are satisfied is satisfied and when  less than 1/3 of people are disatisfied with  
1:36:38
the the way democracy works so let me ask you  a few countries whether you can guess which is  
1:36:46
group one group two or group three the group one  is the more satisfied people right group two is  
1:36:55
uh and then group three is by far a large large  majority of people this okay uh let me remember  
1:37:05
the list so let's start with the US United  States what do you think one two or three one
1:37:16
stre is true okay this a couple of years ago  that's before the last election okay still
1:37:29
president that might be little  B POS not sure uh let me see
1:37:38
Sweden that's clear Germany  I mention what' you say
1:37:55
is one I said one of the best uh
1:38:00
Canada Canada no it's not OB no it's not common  sense yeah no then uh Spain they know
1:38:17
well 25% only people satisfied  Greece if you remember the recent
1:38:28
history India what do you think  three right that's what you
1:38:39
think so that's my point expectations  in many countries is say in this list  
1:38:54
us Som how but Spain Greece Italy those  countries there were the expectation that  
1:39:00
they are developed countries everything  is going well and our children will be  
1:39:05
better than us and our grandchildren  better than our children and this that   happen so there's a lot of disappointment  high expectations are disappointed okay
1:39:23
but in India I'm serious after Independence in India  there were 40 years 1952 the first election until  
1:39:38
19 9192 for 40 years the rate of growth of the  per capita income of India was Zero the country  
1:39:52
didn't grow at all at the time was a lot of jokes  about it the Indian rate of growth mean but in the  
1:40:02
last 30 years since the early 90s until now is the  first country in the world with highest rates of  
1:40:10
of economic growth now even more than China in the  last few years an average about 7% every year for  
1:40:20
for 30 years never happened anywhere mean in mod  times well even less in previous times uh so then  
1:40:29
of course after this they are extremely satisfied  as a new middle class so poverty has been reduced  
1:40:36
dramatically Etc a lot of things to do but you ask  about it how do you feel about the way democracy  
1:40:45
work in your country very well so that's my point  so it's expectations matter more than actual  
1:40:52
substantial result and then much of the problems  of the of disappointment with democracy is the  
1:41:02
meaning of the word disappointment you expect  something and you don't get it and that's why  
1:41:07
many people are dissatisfied and they are angry  and they are ready to vote for anybody that is not  
1:41:13
in government or the traditional parties because  they don't perform as expected and they are proing  
1:41:19
very which is not always a realistic description  of reality so many countries still work well in  
1:41:31
economic social terms but not as well as before  or not as well as was expected so then I try to  
1:41:39
do a correlation I'm going back to the papers  but I'm going I Tred in the book per simple  
1:41:46
correlation which is not final proof of anything  but timey to correlate the the dissatisfaction  
1:41:53
with the way democracy works with economic growth  and well this this is the number that I selected  
1:42:01
a few these are longer list of the countries that  I I selected just a few for for Europe questions  
1:42:07
okay but you see at the top is uh satisfied on the  right side Canada Sweden the ne Germany are the  
1:42:16
best right even but you but the second group you  see Philippines Indonesia South Korea India most  
1:42:24
people satisfied right and in contrast in Europe  dissatisfaction in France UK Italy Spain Greece
1:42:33
Etc uh and then this the correlation that I try  to do is not the final words but you see average  
1:42:47
percentage of growth per per capita income uh  in the last period in a very recent period uh  
1:42:55
and this is the the degree of satisfaction with  democracy world you see India here right and  
1:43:01
you see Greece here Italy Spain uh so that's my  idea so I mean just a suggestion no uh there is  
1:43:12
a correlation between the government's performance  which can be measured in different ways one this  
1:43:18
is one of the ways of how the Govern goes for the  progress right and the the degree of satisfaction  
1:43:29
with the the way democracy Works which I think  might be related with the support for theing  
1:43:38
governments to Beed that's anyway so uh this  because you ask me to use my own findings to  
1:43:48
discuss but basically about Professor K's  findings I would say great findings uh I  
1:43:59
would encourage you to develop those other  countries that you mention in Africa Asia   Latin America to see the differences I would say  if you have a very long term period try to take  
1:44:13
it in different of period especially the last  15 years or so whether it's different we talk  
1:44:19
about it yesterday you said yeah there's less  rections in the most recent period apparently  
1:44:25
okay that's what my so and then uh even more so  try to make a formal hypothesis to explain why
1:44:36
governments so the idea is you need an hypothesis  in order to to uh have some implications that  
1:44:54
can be tested with your empirical data  that very and I'm sure that we eluminate
1:45:02
thank all right we have time for questions
1:45:31
uh well you mentioned National governments right  and and actually right Japan India but we could  
1:45:39
we could mention a few more we could mention  the mapai party in Israel we can mention the  
1:45:44
Social Democratic party the Christian Democratic  party in uh in Italy several parties like that  
1:45:50
that really were very high and then they they  kind of crashed some have to around a little   bit but um so why well we don't know why the  simple answer right so we know that K varies  
1:46:05
from country to country but also even between  periods that's one thing that you know my my T  
1:46:11
had run out so I didn't want to go into that but  that's I thing we want to do we want to explore  
1:46:17
say Italy uh Israel and other countries where  you had that India at the national level even  
1:46:22
even on the uh uh subnational level the same thing  happened the uh National Congress ruled for a long  
1:46:28
time and then you know basically other party  took over so why why that's the question that  
1:46:35
uh that Professor ker asked me many times right  why why why and and you know I I wrote him an  
1:46:44
email this morning I woke up early and I wrote  him an email about this I said you know uh and  
1:46:52
you know we he and I don't agree in in everything  right uh rich and I don't agree in everything but  
1:46:58
most things we do but um why why why I said well  you know uh in in political science my opinion we  
1:47:08
have too many theories we don't have any hardly  any laws or any laws at all right we talk about  
1:47:14
dver J law we talk about uh the the Democratic  peace hypothesis and uh um Richard makes fun of  
1:47:22
me because because I do think that we have laws  and but see because he studied physics you know  
1:47:27
well you know social doesn't have any laws well  you know I disagree with him we do have laws  
1:47:34
but um so I I'm interested in establishing the  facts right the fact that different countries  
1:47:41
have different KS and and how this behaves  and everything that we said here my my side  
1:47:47
and Richard side and I'm I am less interested  in why right now you know I want to get I want  
1:47:53
to get the how does the function work how does  it really work uh and also another thing that Dr  
1:47:59
col and I were talking about today I said to him  you know political scien are always looking for  
1:48:04
explanations outside of political science outside  of politics you know the economic system the  
1:48:10
sociology the psychology said I'm more interested  in finding explanations within politics itself  
1:48:16
right the work the very workings of politics and  the state uh how can that help us understand how  
1:48:22
it behaves you the various mechanisms and I  think that's what Richard and I are doing here  
1:48:27
uh we're not looking for explanations right  right now outside of politics we're looking   at how it behaves and and and understand his  behavior without seeking an explanation outside  
1:48:37
eventually of course you have to look at the  whole system after all state and politics are   part of a larger system you know economic so  social International so the more you do that  
1:48:47
but let's understand the system first and then we  can expand so the short answer is we or no uh the  
1:48:55
second answer to to borrow something that my my  mother-in-law once said uh when she was you know  
1:49:01
sort of declining uh I said Ruth I said uh what  year is it and she said I don't know and I don't
1:49:08
care actually that's a good question the only  turn limits are in presidential systems or us  
1:49:26
governatori systems but um we we don't look at  the person as the incumbent we look at the party  
1:49:33
as the incumbent okay but but your question  is it's worth asking and and trying to think  
1:49:39
about it because I think there is an effect uh and  that's another question that Richard and I argue  
1:49:47
about uh because in in the governatori systems uh  not only I but other Scholars observe a bump uh  
1:49:56
with the first uh rection uh when the incumbent  person in of the party it gets reelected and but  
1:50:04
then after that he cannot re be he canot go back  and so then it kind of seem to dissipate right  
1:50:09
so he and I work you know disagree with this we  work on that and uh and and but I think there's  
1:50:14
something to it uh so yes I think I think you're  right there persons do matter uh but not not  
1:50:21
they don't matter as much as you know people we  tend to focus to much on Persons and not on the
1:50:27
party any questions for me or for Dr col or Dr
1:50:36
Hagen not many economic collapses like  that that one uh 20 you can say 1970s  
1:50:54
early '70s okay um two oil crisis but not  for many countries 2008 was a big disaster  
1:51:04
I can tell you about many countries in  Europe I know Spain better than other   countries but also Greece Italy Etc  the 15 years after the the financial
1:51:16
crisis they are not yet at the level of per capita  income that they were before the number of people  
1:51:26
employed still lower than it was before even  the population is bigger so those countes didn't  
1:51:33
recover yet they are better now than 7 years  ago but still if you look at the whole period  
1:51:42
is kind of flat so was down down and up but now  still more or less the same level that it was in  
1:51:50
well lower than it was in 2007 so this has been  a big chock uh for politics also because most  
1:52:00
many parties have just disappeared you look  at Greece no party survived Italy even more  
1:52:08
spectacular no the current government is a new  party never been in government never have more  
1:52:14
than 5% of votes Now isn't leading the leading  the government uh Etc uh so it could be so you  
1:52:24
could try several periods especially after C  War I think that things began to change uh with  
1:52:33
Clinton basically right uh in the period U but uh  the last 15 years has been very significant for  
1:52:40
many democracies for many Advanced democracies in  econom has been significant in economic terms and  
1:52:49
then political as well I can tell you many stories  even France uh no France is supposedly to be more  
1:52:57
solid country well look at the situation so the  the last two presidential elections the survivors  
1:53:04
at the second round were from two parties that had  never been candidates at the second round in the  
1:53:11
previous 50 years uh or more 60 near 60 years okay  uh and the two parties that were alternating in  
1:53:22
the presidency uh for many decades are in shamol  so the Socialist Party almost is less than 10%  
1:53:31
of votes and the Republicans a little more but  never going back to presidency probably so I  
1:53:40
think it's a general crisis in that in the most  recent period much bigger than in the previous
1:53:45
decades I think we have time for one more no I  just want to uh give a give give l a more serious
1:53:57
answer I I leave it to the party specialist to  to to kind of inquire into that right uh so we  
1:54:08
look at the system as a whole so to speak  uh the general patterns but when you ask   us about specific part although I'm intrigued  too why these parties you know the mapai party  
1:54:18
the Christian democratic party and so on why did  they kind of you know kind of basically shriveled  
1:54:24
up and died or or almost died uh but I leave it  up to the people who studed parties like we have  
1:54:29
a colleague here who studed parties right uh  Michelle Williams right so maybe you can ask her it's impossible I've been in more or less in  this country or close enough my first American  
1:54:57
election was 88 it was in University of Chicago at  the time and I I follow very one of them and from  
1:55:04
the very beginning I I was coming from Europe  and I was not bad at betting on elections in  
1:55:12
Europe okay and even sometimes when this party is  going to be the third or this party which is the  
1:55:20
parliament is going to re-enter I was doing so  well and then uh in this country I realized that  
1:55:27
was impossible it's the most difficult prediction  that you can do in any country and it makes sense  
1:55:35
because so let me put a little more academic but  this a kind of theory social Choice theory is more  
1:55:43
formal mathematics about this kind of rections  Etc and one of the clear rules is the the higher  
1:55:51
the number of uh no is how is it the the the  more the higher the number of Voters and the  
1:55:59
most disperse their preferences the more the more  difficult to predict the winner because there are  
1:56:06
many potential winners that could form different  majorities with different sets of Voters right  
1:56:14
and the best example is the United States so you  say f r whatever so there are many possibilities  
1:56:22
to form a majority that is not the traditional  democratic party majority or the traditional  
1:56:28
Republican party majority and now we changing  that way know you know from apparently the last  
1:56:34
couple of Elections because I'm simplifying too  much but many Latin voters move from Democrats  
1:56:43
or Republicans apparently not yet a majority  but significant numbers many Suburban women  
1:56:51
made the opposite travel uh so there are changes  okay and which means that different majorities  
1:56:58
can be formed by aggregating different sets of  Voters okay then is very difficult to predict  
1:57:05
the point is that the the political system is  very restrictive extremely restrictive it's  
1:57:11
the most restrictive in the world in democracy  I mean because they have two formulas that no  
1:57:17
other party have together which is single member  districts for congress which is the formula for  
1:57:24
Britain of course but still no former British  colonies like the US right uh which permits only  
1:57:32
two parties to compete for the seat basically to  candidates work for the seat so very restrictive  
1:57:39
and the direct presidential elections which that  not exist in Britain okay which are also more  
1:57:46
polarizing because only one winner can get it  right uh which is not the case in parliamentary  
1:57:53
regimes in which several parties go to Parliament  and they try to form a majority and there several   possibilities for different Prime Ministers  candidates right so uh think about it there  
1:58:03
is no other country with this combination there  are many presidential elections but mostly in  
1:58:09
large countries especially middle-sized countries  with multiple parties whether because they have  
1:58:15
proportional representation or some other rules  that permit several parties to like in France for   instance but also in Latin America and um but  another the single member dist is in Britain  
1:58:27
in India Etc are with parlament Canada Australia  somehow is with parliamentary regimes not direct  
1:58:35
presidential elections which means that after the  election there's a lot of room for negotiation and  
1:58:41
agreements to make it possible majority right so  this country is very restrictive in this sense we  
1:58:47
talk about it this morning also but uh I can  say it again so I think that if you if I can  
1:58:54
be positive for a moment this is the cost of  having been the first so if you look at the I  
1:59:03
I been study the the US uh Philadelphia Convention  recently which it's extremely interesting and then  
1:59:11
was a big experiment they didn't know exactly what  they were doing okay and then uh the inventing  
1:59:18
from scribe right with misunderstandings about the  British system they thought they were copying the  
1:59:24
British system they didn't do it well because  they thought that the King was still the chief   executive and it was not anymore uh they didn't  they didn't expect parties parties were so corrupt  
1:59:36
factions right for all the delegates you remember  Washington's firewell speeech don't have parties  
1:59:44
because this is going to be disaster no M all  parties are the worst thing Etc no and then after  
1:59:51
30 years or so when all the funding fathers or  whatever disappeared no the first five presidents  
1:59:58
were either authors of the Constitution or the  I mean of the independence of the Constitution  
2:00:03
of both and then uh uh for a while they kept this  idea okay the best people are going to be elected  
2:00:12
okay the most honest and uh competent people are  going to be the winners but after 30 years or so  
2:00:18
when those people disappeared uh Andrew Jackson  though the first introduced partisan presidential  
2:00:25
elections which was the beginning of this conflict  polarization between two sides all the time right  
2:00:32
uh so then so it's very difficult to make a  prediction because the system is too conflict  
2:00:42
BR okay and then any any site can invent something  to typ to create a winner uh and even more if you  
2:00:53
want a simpler answer you ask me who is going to  win the election well tell me who is going to be  
2:00:59
the candidate you you going make a prediction  about who's going to win this football match  
2:01:04
who is winning who who is playing you know who  who are the teams because is this is this team  
2:01:11
going to win the championship depends on the  others right so we don't really couldn't know
2:01:22
all right please join me in please  join me in thanking our speakers tonight post to don't crash on the way
2:01:39
home