the college-noncollege elasticity, that's equation for college and noncollege has been as many times and there's a certain consensus of what elasticity is, between 1.5 and to work three. i'm going to take a 11.75. in reality the estimate about substitutable or how different ouare people with high school ad people who are dropouts is more contentious. some people think there are similar to other people say they're relatively dissimilar. in these exercises the most negative potential of elasticity or the smallest and most negative center with the smallest elasticity saying that people without a degree and people with a high school degree are not very substitutable. they are as different as college and noncollege. i want to point out i can show you for each decade from 1972 1972-2014, by how much increase the growth of high school and by how much? they change the relative supply which is directly affecting the relative wages. the first called on the effects i attribute to immigration with this relatively simple model. college educated started doing much better and high school started fa

this is the college/noncollege group. if you take the high school drop out versus high school graduate, again, this is the percentage of drop outs. they're due toim grags. this will be the wage -- and, again, you see that, again, in two of the three decades between the 80s and 2000, immigration either has the wrong sign so it wouldn't explain the increase in drop out r you would explain variable, only in the 199 # 0s explain some of them. and either very recent is no explanatory. so this exercise done and i think summarized a little bit and going to say, even if you take the most sort of negative estimates here of this effect, for the college/noncollege, you simply don't have the number to generate the negative effect, because a lot of immigration was college intense. so it worked the other way to reduce this. so for drop out high school graduates, you have some action, but only in the '90s. the '90s look different. it's to say that really where a lot of this researchers pointed out. they've been shape that are effected b

a depression of the noncollege wage relative to the college? and then we can do the same exercise between high school dropout and nonhigh school dropouts and is the relative supply can and how much generate this depression of wage of high school dropouts. the important parameter here that we need to do, estimate, which is as i said that elasticity. i'm going to take the parameter that people who are in favor of finding a negative impact on immigration are going to argue is the correct wasn't. so i'm going to bias this model as much against as myself and so in favor of finding a negative effect on immigration as i can. in this parameterization the college and noncollege elasticity and there's a certain consensus of what the elasticity is. even between 1.5 and 3. in reality, difference people with high school and people that are dropouts are more contentious. i'm going to take in this exercise the most negative potential elasticity or the smallest, most negative scenario saying that people with no degree and people with a high school degree are n

. >> the gamble of trump cam in that he is realigning by 'tracking nor noncollege educated white voters and democrats. we haven't seen him do. that all the other groups are stable. >> you mention him, he has a model of election returns that predicts based on based on the variables, trump should win, clinton should lose. but he doesn't think that will happen because trump has run such different kind of campaign that scrambled the electorate in that way. >> dickerson: do you think going back to this point that molly was making about realigning the electorate, if that's your take wouldn't you not need to do the softening because you're realigning? in other words, do you see in the softening undermining this notion that there's a huge group of people out there. this group of people isn't there, or not large enough to make up for the voters you're losing. i think the original theory was trump could have this hard line message be pull noncollege educated whites and white men from democrats, those who aren't voting pull them booty electric tort. what happens basically, yeah, you have this acti

a poll of likely voters shows noncollege educated white men pick trump. in 2012 exit poll on election day found 46% non-educated white men picked romney is that margin big enough to help trump on election day? joe trippi is here. joe, he is doing well with that particular subset of the electorate. >> doing amazing with it. >> there is another poll that shows he is beating romney in that category by 29 points. that one says 12. ultimately what does it mean? >> i think it's 29. i think he is doing -- i don't think there is any doubt that trump is pushing that number as high -- higher than any republican has ever done. his problem is that as he does that he is losing college educated whites. which is the first time in polling history that a republican has lost that group. and it locks like it's either offsetting. i mean, the gains are offset by his losses among college educated whites. there is even some significant evidence in some polling that more college educated whites are leaving him than he is gaining with uneducated white men. so the problem here, th

donald trump's doing well with noncollege whites. and that's a key split, but the african-american vote, overwhelmingly for hillary clinton. hispanic vote overwhelmingly for hillary clinton. she needs both of those to be the case for her to get to a n winning number, but i don't think it's about economic philosophy in the end. >> as part of the policy, john, it's interesting. he promises 3.5% annual gdp, the goal of four mr. trump saying he can do better than that. 25 million jobs over ten years. at the same time, you have oxford economics saying the trump presidency would cost the economy a trillion dollars over the next five years from a loss of jobs and spendinging and possible trade war. on the note of trade, he took china on once again, head on, saying that he would say that they are and label them a currency manipulator. >> he would. even though of course the value of the chinese currency has been aappreciating. certainly, they have manipulated their currency in the past as part of an export promotion policy. that's changed som

iowa: she is vulnerable in because almost half the population is white, noncollege, .nd that is trump'sliwick in nevada, she is him what vulnerable because they had been through a terrible recession, longer and deeper than the rest of the country. carolina isnorth incredibly close. it always is. i tend to think it has a little bit of a blue tint this year. florida and ohio are highly competitive, but i still give clinton the edge. nicolle: what do you think -- what do you attribute the tightening this week to? nowary clinton's numbers look a lot like what they looked like after the comey press conference. it seems like when there is a scandal, she gets a tightening. when she had a couple good weeks, she widens the spread. do you think this is a pattern we can expect? larry: i think it is simply the gradual evaporation of the convention bounce. donald trump had a very convention bounce. it wasn't very impressive because, to be blunt, his convention was not very impressive. the democratic convention was excellent from a number of different perspectives. hillary clinton got a big ounce out

trump does great with noncollege educated white people.with the noncollege educated whites, but romney wasn't plus 18 with them. he was plus 26 with them and romney lost. trump is plus 55 with conservatives. that's great. but romney was plus 65 and he lost. trump is plus 77 with republicans. woo-hoo, plus 77. but you know what? romney was plus 87 with republicans and he lost. so broadly speaking, you look at the polls right now, the poll numbers nationwide and in a bunch of the swing states they're all looking like they're tightening up or tilts towards trump. but in the earliest actual voting data, not polling but voting data and in the demographic data that we've got nationwide, trump looks like he is still right now very well on track to lose and to lose significantly worse than mitt romney did. hillary clinton and the democratic party and the clinton supporters of our country cannot be psyched in general about how the polls are going, but when you drill down and look at the real data that we've got, it's good for her and bad for him. n

that would produce a great jobs boom in america -- both for people with college educations and noncollegeeople. we need an infrastructure program that rebuilds the roads, infrastructures, airports and the water systems. there are a lot of babies all over the country drinking water with too much lead. if you took them all up and put things down that were clean and new, you would create a kind of jobs and give our children a healthier future. we need affordable rapid broadband in every rural community in america. there are 5 million children in this country, some living not far from here, who get homework assignments that require them to get on the internet and they cannot do it. she is the only person that has talked about this. we need a plan that recognizes that we can bring back manufacturing jobs. you know how many jobs have been brought back to america since the bottom of the crash under president obama's presidency? 900,000. [cheers and applause] mr. clinton: you know what the net migration is from mexico to the united states was? talk -- all wall thi hims him wall talk. all this wal

that's a pretty good approximation of the us labor market if you defined unskilled as noncollege and skilled as college so basically college educated workers compete with each other and noncollege educated workers compete with each other for jobs so what matters in this data is how much immigration affects the ratio of unskilled to skilled workers. that's when you can have an impact on the labor market. and it turns out there's a formula for that and sorry for being matted this stuff but it's a pretty simple formula and it's right there, if you take the number of unskilled immigrants to unskilled natives and subtract the same ratio for skilled immigrants and skilled natives in this set up, another way to put this is so those ratios represent how much immigrants grow the workforce of that type. let me illustrate the formula with three examples. imagine all immigrants were unskilled . that first ratio would be potentially a large number. that second ratio would be potentially zero and so at least we would have alarge positive impact on the unskilled to skilled ratio, makes sense. they

something that would produce a great job boom in america but for people with college educations and noncollegeple. we needed infrastructure program that rebuilds the infrastructure. and the water systems. there are a lot of babies all over the country that get water with too much lead. if you took them all up and put things down that were clean and new, you get a -- would get better jobs and a better future. we need affordable rapid broadband and every rule community in america. -- rural community in america. there are 59 children in this children- 500 million in this country who get homework assignments that require them to get on the internet and they cannot do it. she is the only person that has talked about this. we need a plan that recognizes that we can bring back manufacturing jobs. you know how many jobs have been brought to america since the bottom of the crash under president obama? 900,000. [cheers] you know what the net migration is from mexico to the united states? all of this while talk, from , -15to the end of 2014 ,000. they have this open economy now. it is going to get worse

iowa: she is vulnerable in , because almost half the population of iowa is white noncollege, and thatp's bill would. nevada has been through a terrible recession, much longer and deeper than the rest of the country. and we all know that north carolina is incredibly close and always is. i think it has a blue tint this year. florida and ohio, obvious to everybody, are highly competitive, but i still give clinton the edge. nicolle: what do you attribute the tightening this week to? hillary clinton's numbers now look a lot like what they looked pressfter the comey conference. it seems like she goes through a scandal, there is a tightening, when she gets a couple of good weeks, she widens the spread. is this a pattern we can expect if the contours of the race status same? larry: no, i think it is the evaporation of the convention bounce. donald trump had a very brief convention ounce that was not very impressive, because to be blunt, the convention was not very perspective. the democratic convention was impressive from a lot of different perspectives. clinton bounce has declined as we appr

he has an enormous margin on noncollege educated white workers, blue collar workers. you can't please everybody. you're not going to get every vote but we will see how that changes after the first debate or the series of debates. jenna: one of the reason that is we are taking a look at the polls and battleground states and not only shift in numbers, hillary clinton lost ground, donald trump has gained some ground and his vice president, vice presidential candidate, governor pence was on with fox and friends this morning and they were asking him about what's working for the campaign. here is his response. >> i don't think really anything has changed. as donald said on your program this morning, the message has remained the same, make america great again. jenna: has nothing changed, larry? >> no, what has changed is hillary clinton has had problems. her problems have cut some of her support. she's been off the trail, i've already mention it had recent controversies and, you know, she had tremendous advantages starting out in solid leads in almost all of the swing states

donald trump has faced in the general election is that he built himself a base in the primaries of noncollegeducated white voters who respond to his core message about economic declierngs about cultural change, and are inspired by that notion. but there's a different set of people that he needs to reach in the general election. they are suburban, more upscale. they are white college educated, college graduates. and they want a more display of gravitas, of sophistication of fluency in policy and i think donald trump didn't give him that last night. doesn't mean he can't do it in the next couple of debates but i don't think he got that done. >> john thank you very much for that analysis. john harwood in new york. that brings to us our twitter and facebook question of the day. has the debate made you decide how you'll vote. the options yes, clinton yes, trump or no still unsure. get in touch with us throughout the show. we'll bring the results of that later and continued coverage on cnbc today of those debates last night. >>> the labt department is launching a review of all complaints of wells f

whether it's suburban women, college, noncollege, young women. his negatives aren't good. they are all in the 60s. calling a woman miss piggy, having the story come out, how does it help him appeal to this demographic? >> it is not helpful. there are some things that can be less than helpful. i'm certain hillary clinton right now is thinking maybe i should not have been so vindictive toward some of those women. maybe i should have sought to be more helpful through the entire process. there again i think you have to be more judicious with the words you choose. i am pleased to have a presidential nominee for our party who is the first one to come forward talking about the issues of child care and women in the workplace. >> you know that's interesting. i was speaking to a reporter earlier. i got off the phone with a republican operative in wisconsin who reflected that. she said she liked his child care plan, mentioned his high profile female surrogates, ivanka trump, kellyanne conway. hillary clinton has a surrogate out today, michelle obama. >> there are those who question a

rry: she is vulnerable in iowa because almost half the population is white, noncollege, and that is trump'sliwick. in nevada, she is him what vulnerable because they had been through a terrible recession, longer and deeper than the rest of the country. we all know north carolina is incredibly close. it always is. i tend to think it has a little bit of a blue tint this year. flida and ohio are highly competitive, but i still give clinton the edge. nicolle: what do you think -- what do you attribute the tightening this week to? hillary clinton's numbers now look a lot like what they looked like after the comey press conference. and conference. it seems like when there is a scandal, she gets a tightening. when she had a couple good weeks, she widens the spread. do you think this is a pattern we can expect? larry: i think it is simply the gradual evaporation of the convention bounce. donald trump had a very convention bounce. it wasn't very impressive because, to be blunt, his convention was not very impressive. the democratic convention was in excellent from a number of different perspectives.

educatedwhites mitt romney won by 18 points trump up 19 points, noncollege educated whites. >> thereence in the room not enough for him to win. >> okay. let's talk about this -- this action, this effort that he just went through to go down to mexico meet with mexican president, about talk about that wall, he says they don't know they are going to pay for it but they are. talking about in his speech clarifying his immigration position, talking specifically about his plans to deport illegal immigrants currently in this country he said he was softening stance then sort of doubled down in original approach. >> i think path to presidency much more than visit in mexico president a.m. image cutting through bombastic donald trump to be a lowered american people can see him in the oval office the problem i have with proposals when you say you are going to tell every single person in america came here illegally you need to get out of the country turning off so many could vote republican it is not practical not going to happen, i met somebody from -- a waiter at restaurant said donald trump doe

noncollege white men. you can see where they are.tered around the pittsburgh media market and a smaller degree around erie and scranton. >> i cut you off -- >> it is a totally different situation than ohio. if you add those up, she could win without winning any persuadable voters in the persuadable universe in pennsylvania. it is so much smaller than it is in neighboring ohio, about three times larger in ohio. trump could do everything he could to turn on his base and his gotv's. he could win persuadables and it would not be enough for him. that is why people say it is such an uphill climb. but it is still a really important state for him to win. he has to appeal to democrats and these are the ones we think are the most vulnerable for him to convert. john: sasha, talk to me about michigan. if pennsylvania is hard for him, michigan must be well out of his reach. sasha: we see similar economic dynamics across these states, but the electorate dynamic is really difficult for him there. there are about 189,000 we found of these working-cl

secondly, donald trump is not building the margins among noncollege white voters that he needs. he's doing more poorly among that group for all the attention it's gotten than mitt romney did four years ago in 2012. he needs to overperform romney with that group. and finally, hillary clinton is doing better among seniors and voters age 30-44. she has done less well than obama among the younger voters, 18-29. but in that middle range, 30-44, she's doing better. and among seniors, she's doing better. guys. >> all right, john. thank you. john harwood with the latest numbers there. so dan, does this fuel your thesis? >> yeah, i really do. listen, i'm not trying to be partisan. i really think if she regains some of that steam she had in august, i think it's all clear for stocks. especially if we get that earnings bump that we could get. we did get that out of july, guys. i mean, listen, i was very pessimistic. and they weren't as bad as expected and that was really what people needed. and then when it became unlikely the fed was going to raise in september, despite the chatter -- >> s

second, almost half of the electorate in iowa is noncollege educated. and we've seen a tremendous split between those who are college-educated. they're pretty heavily for clinton and those without a college education. they're pretty heavily for trump. >> larry sabato, always good to have you. thanks, larry. >> thank you, brit. let's go out to iowa now and get an upclose sense of the race there from the des moines register's political columnist cathy. you heard this last segment. >> the poll is quite remarkable. what are your thoughts about it? >> yeah, brit. it is remarkable. because the race in iowa has been really, really close. the last -- over the summer it really since the -- both candidates have locked down the nomination, it has been neck in neck in iowa. and an 8 point lead as you said largest lead that donald trump has had iowa to date. so, you know, i think the question is why is it suddenly, you know, is he opening up a big lead in really a short period of time. and when you look at when the poll is taken, monday through wednesday, just like

she she's, head in the average in virtually every swing state and this weighs noncollege-educated whites and that could be wrong and hillary has said she's going to campaign as an underdog whether she's ahead or not. that's what she's doing and the more people see of her and the more they like it and the more they'll press her and she'll win the case. >> trust has not been going in that direction. and kayleigh, this is the beginning of a big momentum. just last week cnn reported that cnn's poll of polls and the average of polls showed that he made up half the grounding of hillary clinton and that was confirmed by our independent polling today and he's ahead by two and it's worth mentioning he made up 11 points in four weeks. he's leading 20 points among independents and that's something romney won by five points and married women, 17 points and it is not good for hillary clinton. this trend is definitively going. >> we always hear the women -- it's just, forget about it for donald trump. when you look at married women, more than half of married women, 53% support donald trump and that is

. >> rose: noncollege educated. >> right. the degree to which he had tapped into the resentment, the grieveances they have, the fact that they have done far less well with this recovery than people better educated and certainly more affluent. and yet it's like it's flipped now. and he needs either a huge turnout, a bigger turnout from that constituency than we have seen in the past, or he's got to solve the problem with college educated. >> rose: one or the other has to be big. >> yes. and perhaps some combination. >> rose: and we don't know how to measure the discontent. >> no, we don't. we know it's out there. one measure is that you know, the right track wrong track which is a standard measurement in survey research has consistently been, people say we're on the wrong track. but that's been the case for years, for many years. and barack obama was re-elected with that. i think you have to take that measurement with some grain of salt. but one of the things we did in our 50 state poll is we asked people, do you think that th

white voters, and especially strong on average by all the polls, close to 30 points, with white noncollege degree people. host: based on that, which sector is in play of the electorate? ken: so, the sector in play is right?ollege-educated, we sent out clinton is doing well with college education, trump is doing well with noncollege education, so that white college, which is a real swing group, and then, the other age groups.it is a broader range , but younger voters are for clinton, but the 35-49 slightly for clinton. host: you are calling this the bloomberg politics poll decoder, what is the methodology, and with some indifferent surveys, for example the l.a. times has donald trump ahead, the latest, from nbc news shows and requested of the head so how do you make sense of all this? ken: so, i think when you hear from a lot of folks is not paid attention to any one poll. i agree with that. there is lots of great sites that aggregate all the polls, p ollster.com, real clear politics, a bunch of other news positions in aggregate as well. basically, why don't we do the same thing, but for th

we said how clinton was doing well with college-educated and trump is doing well with noncollege educated. so it is that spot of white college which is a real swing group. and then the other age group sort of in that. it's a broad range but younger voters for clinton and older for trump but the slightly. [inaudible] >> you're calling this the bloomberg poll decoder what is the methodology and with so many different surveys out there, for example the l.a. times has trump ahead, the latest from nbc news shows hillary clinton is ahead. how do you how do you make sense of all of this? >> i think what you hear from a lot of foxes do not pay attention to anyone poll. i grew that. don't pay attention to anyone poll. there are are lots of great sites out there that aggregate all of the polls. there's 538 and pollsters.com and other news organizations that aggregate polls as well. so i basically thought well why don't we do the same thing but for the demographics of a survey for what people in politics are sometimes called the internals. winning elections if your campaign managers about maximizing

. >> trump is doing well with the noncollege educated so there's the spot of white college and then the other age groups -- they'll get the broad range but youngest voters for clinton and oldest voters for trump but the 35 to 49-year-old for clinton and the 50 to 64-year-old slightly for trump. >> you're calling this the bloomberg politics code decoder. what is the methodology and with so many different polls for example the la times has donald trump ahead. the latest from nbc news shows hillary clinton is ahead. how do you make sense of all of that. >> i think what you hear for a lot of folks is don't pay attention to any one poll. and there's lots of great sites out there that are all the polls. it's the demographics of what people sometimes call the internals but winning elections if you're a campaign manager is about maximizing shared performance and getting polls right is about getting sharing performance right so i found myself when a number of polls came out and i say never pay attention to a single poll. of course i do. you look at the internals and i wanted to know how many dem

i mean we're looking at our abc poll that shows him having 73% of noncollege educated white men. she has 57% of college educated white women which is great but not anywhere near as big as his. when she starts talking about trump's relationships to beauty college educated women saying, oh, do we really want that in the white house? maybe i better get more energized and out there. >> and, jon, this morning you're hearing about donald trump complaining about the mike. saying it was maybe fixed and not liking lester holt. how hard will they push this going into future debates getting guarantees from the mod raters. >> when you have won the game you don't complain about the refs and they were complaining and as you mentioned about the mike. everything else. clearly that's a sign they think they did not do well. rudy giuliani as you pointed out in your interview with mike pence suggested trump might not do the next interviews unless he is guaranteed better treatment from the moderators. i think it's very much an open question. >> cokie, going back to you -- sorry, did you want to respo

and especially you hear it a lot, in the upper midwest and places like that, where particularly noncollege educated white working class people have been left behind, is that people are coming in and taking our jobs. the fact is that, as has often been pointed out, that the benefits of immigration are very widely disperse but -- and globalization in a larger sense, but that there are people out there who see themselves as real victims of this, and so it is donald trump but it is also something i think that it speaks to something larger and, again, it is a real sense of people feeling like they're being left behind. >> okay. david, let's have you add perspective here. i'm struck by the fact that when you see the stalemate that exists over immigration and that's existed as karen says, she would have thought it was not a question of if but when we would do something about immigration reform. when you look back we have in presidents reagan and the first president bush and president clinton, presidents that all signed bipartisan immigration bills. they were bills that had things in them that non