With little time remaining until the midterm elections, campaigning is intensifying and the outcome for control of Congress remains uncertain. Voter enthusiasm and turnout will be big factors in the elections, where Republicans have demonstrated a leg up in the party’s primaries.

On October 20, Brookings expert Michael McDonald answered your questions about what the polls and early voting are telling us about the upcoming midterm elections, in a live web chat moderated by POLITICO Assistant Editor Seung Min Kim. McDonald, with Seth McKee, is author of "Revenge of the Moderates," in today's POLITICO.

The transcript of this chat follows:

12:30 Seung Min Kim: Good afternoon, everyone! We have just under two weeks until the Nov. 2 midterm elections, and the Brookings Institution's Michael McDonald is here to answer your questions. Thanks and welcome, Michael.

12:30 [Comment From Dale Dean (Arlington): ] I was wondering from the historical record how closely early results mirror the actual results. Are there systemic distortions in early voting that are the same over many elections or do they differ with each election?

12:30 Michael McDonald: Early voting does not necessarily correspond with Election Day voting. Several data sources suggest the following: Overall, prior to 2008, more Republicans tended to vote early. In 2008, it was Democrats who voted early. We have to see 2010 will be a continuation of 2008 or a reversion to previous elections.

12:30 Michael McDonald: Another important factor is the number of early votes. For high early voting states like Oregon and Washington, essentially ALL votes will be cast early. In other states that require an excuse to vote absentee, the early voting electorate will be much smaller, and have a partisan character more similar to pre-2008.

12:31 [Comment From Katy Steinmetz: ] Are black voters going to turn out for Obama like they did in 2008? Why or why not? How big of a difference do you think this will make?

12:31 Michael McDonald: Since we started surveying, pollsters have found that midterm electorates -- compared to presidential electorates -- tend to be older, wealthier, better educated, and composed of fewer minorities. Sometimes Democrats can overcome this hurdle, as they did in 2006, of course. It would be highly unusual for African-Americans to vote at the same rate as they did in 2008. In some key races, in states with large minority populations, lowered levels of minority voting could be a critical determinant to the outcome.

12:32 [Comment From tim: ] Do the polls accurately reflect the relative turnout of Democrats, GOP and Independents?

12:33 Michael McDonald: Pollsters try as best they can. They try to forecast who is likely to vote by various methods that are not consistent across polling firms. So, this is as much as art as a science. There are a number of factors that may further affect the partisan composition of polls, such as if people are interviewed by live interviewers or automatically or whether or not cell phones are interviewed.

12:34 [Comment From Katy Steinmetz: ] When Republican pundits like Karl Rove predict gains of 60 or so seats in the House, does that help or hurt them (in terms of making Republicans complacent and driving Democrats to the polls)?

12:36 Michael McDonald: One of the big questions in this election is the relative effects of enthusiasm versus voter mobilization. Republicans are hoping the enthusiasm gap will help them to victory, while Democrats are banking on their organization to GOTV. So far as I can tell, neither side has a distinct edge yet.

12:37 [Comment From Casey (DC): ] I have a question about the margin of error. Let's say candidate A has been consistently polling a point above candidate B, with a 3% margin of error. Is the fact that A has beaten B in all recent polls statistically significant, even with a margin of error? That is, wouldn't it be misleading to claim that A and B are tied (due to the margin of error) since A has been beating B consistently in the same poll, even by just a point? If they're truly tied, wouldn't we see A beating B half the time and B beating A the other half??

12:41 Michael McDonald: To quickly review, the MoE is determined by the number of respondents to a survey, and it does not linearly decline as the number of respondents increases [it declines by a factor of 1/sqrt(# of respondents)]. Suppose you have two polls with 1,000 persons each, then. You may treat them as two polls of 2,000. So, the MoE would decline, but it may not decline as much as you might think. Further, as I describe above, different pollsters use different techniques to create likely voter screens (and many other survey issues), so the polls themselves are not entirely comparable.

12:42 Michael McDonald: As a general rule, I like averaging polls and looking at trends among the same pollster. If all the polls are moving in the same direction, I tend to believe that a trend is real and not just statistical noise.

12:43 Michael McDonald: Finally (I know a long answer!): never trust a single poll. Unfortunately, the media tend to report their poll, or a surprising poll, and disregard others.

12:43 [Comment From Jazziette Devereaux (AZ): ] Do you think that early voting can prevent voters from learning facts about candidates that are presented in the feverish last two weeks of the election?

12:44 Michael McDonald: My favorite example is a John Edwards voter who was upset in 2008 that he had cast his vote before he dropped out of the race.

12:46 Michael McDonald: Early voting has certainly changed campaign dynamics. No longer can an opponent release the October surprise the last week. Their opponent gets a chance to respond. And it makes elections more expensive since campaigns need to be active throughout the entire election period. So, there are pluses and minuses.

12:46 [Comment From Mark, Greenbelt: ] Is it your feeling that early voting favors one party over another generally, or is it all case-by-case?

12:48 Michael McDonald: Prior to 2008, more Republicans voted early. In 2010, more Democrats voted early. So, far more Democrats are voting early in 2008, so it may be that 2008 was a watershed election for early voting. Still, in a state-by-state basis, Republicans tend to do better among early voters in states that require an excuse to vote an absentee ballot (early voting rates are much lower, too!).

12:48 [Comment From Rosemarie (NH): ] How do you think negative campaigning impacts turnout?

12:50 Michael McDonald: It used to be that people thought negative campaigning decreased turnout, but since then, numerous studies have shown it increases turnout. People are apt to be interested in slowing down and watching the accident on the side of the road. The media certainly enjoy covering the most negative campaigns, too.

12:50 [Comment From Malcolm, DC: ] Do you have any stats about early voting so far, and can you draw any conclusions?

12:50 Michael McDonald: They are here. So far, over 2 million people have already voted!

12:52 [Comment From Borys Ortega: ] How do you see the Obama support base (liberals, young people, etc) in terms of enthusiasm?

12:52 Seung Min Kim: And in addition to that, it seems like the White House and Democrats are doing a lot more outreach to young voters, with the MTV/BET town halls and the large rallies at universities. Do you think that will have any effect, considering young people have a low turnout rate for midterm elections?

12:53 Michael McDonald: Since we began surveying, polls consistently show that young people, minorities, the poor and uneducated tend to vote at lower rates -- perhaps the most ironic thing about this election is that the people most affected by the economic downturn are the least likely to vote.

12:55 Michael McDonald: The Democrats need to counter the Republican enthusiasm by expanding the electorate. Their strategy is to do voter mobilization targeted at the low propensity midterm voters, like the youth. We will again have to see how effective the Democrat's mobilization will be compared to the Republican's enthusiasm.

12:55 [Comment From Rosemarie (NH): ] Has there been any correlation between the level and campaign spending (especially on advertising) and the results?

12:57 Michael McDonald: A funny statistic is that the more an incumbent spends, the worse they do. This is because they are spending to counter a threat from a viable challenger. This is why this is one of the most difficult questions to answer -- surprisingly. We do not know the marginal effect of another dollar spent because the other campaign is also spending money.

12:57 [Comment From Sally: ] There was a flap this week about Univision airing ads that seek to depress Hispanic voter turnout. How common is that practice?

12:59 Michael McDonald: Voter suppression targeted at minorities has a long and ignoble history in American politics. Generally, I think everyone should vote since democracy works best when its citizens are engaged. This particular episode may ultimately backfire since it may rile up Nevada Latinos in a campaign that has had many racial overtones.

1:00 [Comment From Drew C.: ] What's your evaluation of early vote-by-mail, vs. in-person voting? Are both being done well?

1:00 Michael McDonald: In 2008, approximately 500,000 mail ballots were rejected. These were people who thought they voted by their vote did not count.

1:02 Michael McDonald: Why does this happen? People do not follow the procedures properly -- the return the ballot in the wrong envelope, they do not sign the envelope, etc. I do like California's method of allowing voters to drop their ballots off on election day at their polling places. This allows poll workers to check that the voter followed procedures.

1:03 Michael McDonald: An advantage of in-person early voting is that these problems do not occur, and their is a chance for a voter and election administrators to fix any problems, such as a first time voter forgetting to bring mandatory ID.

1:03 [Comment From Nick, DC: ] Along the lines of what Sally was asking about, we hear a lot about voter suppression, and we also hear a lot about alleged voter fraud. Are either of them really very common? And are voting machines more subject to tampering than the old paper ballots?

1:05 Michael McDonald: Vote fraud -- someone actually intentionally casting an illegal vote -- is extremely rare. When it happens, it tend to happen among mail ballots. Although there are potentially security flaws with electronic machines, there is little evidence of tampering (of course, that may be because there is no way to check!).

1:06 [Comment From Peter G.: ] If you could make one voting reform nationwide to make the system work better, what would it be?

1:08 Michael McDonald: Universal voter registration. There is plenty of evidence that our system of requiring voters to register themselves does not work well. Just about every other advanced democracy registers their own voters. In states with Election Day registration, turnout is much higher (5 to 7 percentage points). So, not only would we increase turnout, but we would get third party organizations like the now-defunct ACORN our of the business of registering voters.

1:09 [Comment From Ben Griffiths: ] You said incumbents fare worse when they spend more. is the same true of challengers? I'm thinking this year of Sharron Angle's $14 million in Nevada. Is it even possible to spend that much in the time left?

1:10 Michael McDonald: The spending in Nevada is tremendous. Despite that likely about half the voters will have already voted by Election Day -- Nevada is a high turnout state -- I think the campaigns will continue spending to the end since the election appears to be going down to the wire.

1:11 Michael McDonald: As for your first question, there is a point where a challenger spends enough money to become viable, which triggers a response in spending from an incumbent.

1:11 [Comment From Rosemarie (NH): ] Is overall turnout higher in states that allow early voting?

1:13 Michael McDonald: I testified to the U.S. Senate that I believe the answer is yes, though the turnout effects are a modest one to two points in presidential elections. There are studies that find big turnout increases in non-presidential elections. Indeed, the very first usage of all-mail ballot elections was in local jurisdictions that needed to meet threshold turnout rates to pass local bond measures.

1:13 [Comment From Nancy: ] Which party gets the early bragging rights?

1:14 Michael McDonald: So far, Democrats have jack rabbited out of the starting line in most states where we have a clue of which party's registrants are voting early. Nevada is an interesting departure, where Democrats have a lead, but it is not as great as 2008.

1:14 [Comment From Carson P.: ] One of your Brookings colleagues - Bill Galston - has proposed the idea of mandatory voting, like they do in Australia. Could that work here? Is it a good idea?

1:15 Michael McDonald: Good luck trying to convince Americans that they will be fined if they do not vote. I do not think this is practical for the U.S., though it obviously increases turnout.

1:15 [Comment From Don: ] What are the prospects for Lisa Murkowski come election day? Do you think she has a realistic shot at beating Joe Miler?

1:16 Michael McDonald: The polls are close. I think it is anyone's game in Alaska. In fact, I wrote an op-ed with my co-author Seth McKee, which was published at Politico today.

1:16 [Comment From Greg Dworkin: ] Thanks for all your hard work on this! How 'institutionalized' do you see the early vote by the parties? are they incorporating early voting as part of GOTV or are they behind in realizing so many people vote early these days?

1:19 Michael McDonald: As I document with another co-author -- Tom Schaller -- the Democrats created a strong early voting GOTV organization in 2008, and Republicans only belatedly tried to mobilize their voters to vote early. We will have to see how well Democrats will roll over this organization to 2010. Eventually, I believe the Republicans will have to build as strong as an organization. Early voting allows a party to mobilize over a longer period of time.

1:19 [Comment From Mary H. Hager, PhD: ] Please clarify polling methodology. Who is reached; who is not. The role of technology (email, telephonic, etc.) in defining the subpopulation for polling data.

1:20 Michael McDonald: That is quite a tall order for a chat :) We discuss many of these issues on Pollster -- which now has a home in the politics section of Huffington Post (I also blog at Pollster).

1:21 [Comment From Don (Ossning, NY): ] Does Christine O'Donnell have a chance in Delaware?

1:21 Michael McDonald: No.

1:21 [Comment From Geoffrey V.: ] Over the years, I've gotten the sense that campaigns are moving faster, that there are more undecided voters and that many voters don't make up their minds until the last minute. Is that supported by the data?

1:23 Michael McDonald: Well, given the tremendous increase of early voting from 20% in 2004 to 30% in 2008, it appears that many voters are making up their minds sooner, not later. Still, in a midterm election, the rule has generally been that people tend to hold their ballots longer because they do not have as much information about the candidates. It appears that this election may break that previous pattern.

1:23 [Comment From Joan: ] Do you think compromise will come back to Congress after the midterms?

1:24 Michael McDonald: No. Historically, we still have a ways to go before we reach the highest levels of polarization in our politics observed in the late 19th century.

1:24 [Comment From Al Amundson, ND: ] It seems sometimes that pollsters are "surprised" by wins. Polling is so scientific these days, and there's so much money behind it -- how often does a real surprise actually occur?

1:25 Michael McDonald: Surprises more often occur in primary elections, where the electorate is difficult to predict and information is fluid. I do not expect we will be greatly surprised by the 2010 election outcomes.

1:25 [Comment From Rosemarie (NH): ] Do you think that even with early voting, people just want to get it over with, go in to vote and make up their minds while they read the ballot?

1:27 Michael McDonald: Want the campaigns to stop bugging you? Vote early if you can. Election officials track who has a mail ballot in hand and who has voted, and they share this information with the campaigns.

1:27 [Comment From Bert C.: ] How is Sharron Angle still holding on in Nevada even after her numerous public gaffes?

1:27 Michael McDonald: The economic crisis has hit Nevada VERY hard (and I don't often write in caps!).

1:28 [Comment From Peggy: ] What role do you think the Tea Party will play in future elections? Is this a one-off movement or something more serious in American politics?

1:30 Michael McDonald: Shameless plug: see my Politico op-ed. A conservative/populist movement is nothing new to American politics. At least in the short run, I expect the tea party to continue to be influential, especially if Republicans take the House -- I do not expect they will take the Senate as of today. Victories will further embolden the activists.

1:31 Michael McDonald: Thanks to everyone for your questions. Sorry I could not answer them all!

1:31 Seung Min Kim: And that's it for today. Thanks for all the great questions as we count down the days until Election Day. And thanks to Michael for his insightful answers!