Elections have consequences. And the consequences of last week’s presidential election are already revealing themselves .
In an interview with ABC last Wednesday, House Speaker John Boehner stated that the Affordable Care Act (“Obamacare”) was the “law of the land.” A dramatic shift from a man who oversaw more than 30 separate legislative attempts to repeal the law in its entirety in less than two years.
It appears however, that the Speaker is not the only one whose stance on repeal has been shaken.
According to the findings of a Kaiser poll conducted immediately after the election, support for the full repeal of the ACA (regardless of if it were accompanied by an express Republican alternative) is lower now than it has been at any point since January of 2011. To be sure, the decline in support for repeal should not be confused for a wide embrace of the ACA. But it is preliminary evidence that Obama’s re-election has eroded confidence in repeal as a viable option.
With Obama’s re-election behind us and support for full repeal in decline, the only question now is how much of an effort Republicans will make to cripple and impede the ACA’s implementation.
Acknowledging a failure to reach Hispanic voters as crucial to Romney’s drubbing Tuesday night, George Will proposed the Republican party take swift action to increase the visibility of Florida Senator Marco Rubio. No doubt lifting up the a young Cuban success story would accomplish some good. Not only would it show a willingness to reorder the party in light of failure, it would possibly prevent the image of the Republican party as a uniquely old, white, and southern franchise from firmly setting in the public mind.
This advice however, fundamentally misunderstands what is principle to the Republican party’s failure to reach Hispanic voters. The problem the party confronts is not the conspicuousness of its homogeneity. Nor, in my view, is it as simple as the party’s callous approach to immigration.
Rather, it is the Republican party’s systemic opposition to investing in those things that Hispanics understand will help their families realize generational success. It is the reflexive unwillingness to see value in communal investment. It is the consistent failure to see human capital as something, above all else, that need be cultivated.
The Republican march toward irrelevance has nothing to do with the party’s makeup being too old, too white, or too Southern. It does however, have everything to do with the increasingly narrow lens through which they view the world and the increasingly rigid ideology that dictates their policy portfolio.
The problem facing the Republican party is not a cosmetic one. It is a substantive one. Re-branding and repackaging will not be sufficient. So there is no reason to think simply changing the salesman will be either.
If you have read any of my polling updates you likely have noticed that I often reference partisan instincts and behaviors. I do this for two reasons. The first is that I think these are terms that are adequately descriptive and more easily digestible for people who are unfamiliar with much of political science (avoid jargon when you can). Second, it reflects the quasi-behavioralist (jargon!) approach that serves as the bedrock of my polling analysis.
The nice thing about this approach is that it defers to well established and long observed trends, but remains sensitive (without being over responsive) to the manifestation of new ones. As such, data that hints at new trends are not wholly ignored nor do they supplant already established work. Put another way, new trends and data are threaded into the existing latticework of reputable observation — building upon what we already know and putting new information within its proper context, as to produce a more fully formed and accurate work.
As with every election, this current one has provided a deluge of new information. Much of it useless. Some of it (perhaps) useful. But for curiosity’s sake, I though it might be interesting to see how a purely behaviorialist approach to predicting the outcome of this election based on data from December of last year stacks up with a prediction based on aggregated polling data from a few days out from the election.
As of late this afternoon, the Real Clear Politics poll-of-polls (POP) gives Obama 0.4-point advantage, the Talking Points Memo POP gives Obama 1.1-point advantage, and the Pollster POP gives Obama a 0.7-point advantage. Averaged, these three major POPs predict Obama to win the popular vote by a margin of 0.7-points.
Aggregated and weighted polling data is orders of magnitude more accurate than any one poll, and the polling conducted immediately before an election is (as a rule) highly predictive. As such, this average provides both a baseline and a steep point with which to test the value of the behavioralist approach.
Now at the core of the behavioralist approach is the idea that large groups with shared characteristics are rather predictable despite individuals being unpredictable. Within the electoral frame work, partisans are the large groups of interest. And luckily, decades of research studying the voting behaviors of partisans has done a considerable amount to reveal predictable habits, trends, and behaviors. In turn, predicting the outcome of this election through such lens (only) requires information about the partisan makeup of the electorate.
Such data was provided in December of 2011. In short, this data revealed both candidates to wield support levels of about 47%. This means that 94% of the electorate was composed of decided-partisans. Moreover, this data revealed the partisan leanings of the 6% who were declared undecideds – 40% were democrats, 25% were Republicans, and the remaining 35% were Independent/Unsure. This means that 4% of the electorate is composed of undecided-partisans.
Now if we apply to this data earlier observations which provides that strong partisans (decided-partisans) can be relied upon to vote for the candidate that affiliates with their partisan leanings 85-95% of the time, that weak partisans (in this case undecided-partisans) do so around 80% of the time, and undecided-non partisans are equally likely to vote for either candidate, it becomes possible to make a rudimentary prediction about the outcome of the election.
And that rigid behavioralist prediction suggests that Obama would win the two-party vote 50.3% to 49.7%. Or, Obama would win by 0.6-points. That is only one-tenth of one percentage off from where an average of the three major poll aggregators has placed the race 2 days out from the election. Not only is this “not bad,” it is downright good.
If nothing else, this is evidence that a flexible and appropriately applied behavioral approach can serve as a sound foundation to political and polling analysis. But I also think this is evidence that elections far less unpredictable when one makes a concerted effort to survey the terrain of each election in full No one could have predicted the 47% video or Obama’s dull first debate performance. What we could predict however, is that these things would not do much to persuade voters. Considering all of this, it comes as no surprise that what we see in the polls today is not markedly different from what a thorough exploration of the 2012 voting universe revealed nearly a year ago.
I have spent a lot of time pushing against most of the polling based narratives that have dominated the coverage of the presidential election over the past few months. Recently I took a break from dismantling these shaky narratives. In large part, I did this because reading so much analysis grounded in the systemic misreading of polling data and willful dismissal of precedent had become exhausting. And honestly, after staking out that Romney had stopped gaining support upon his base having reached critical-mass and a lack of evidence that Obama was losing his Electoral College advantage, I thought I could remain idle until just before vote day.
I guess I was too optimistic.
Desperate for a hook since the death of the Romney momentum narrative, some outlets have cast the line that Obama’s “historic” struggle among white voters might well cost him the election. The precipitating data having been an ABC/Washington Post national tracking poll which showed Obama losing the white vote by as much as 21-points.
No doubt this is disconcerting. Even if the ABC/Washington Post poll’s findings diverge a bit from the consensus as to the magnitude of Obama’s marginal disadvantage among white voters, losing the white vote by say 17-19 points presents a rather large hurdle. That is of course, unless the primary driver of this marginal disadvantage is the wide-spread rejection of Obama by white voters in deep red states.
So I think to better discern whether Obama’s standing among white voters has eroded to the point that it might cost him the election we need to do two things:
1) Look at the state of race among white voters in battle ground states.
2) Compare Obama’s stands among white voters with past Democratic presidential candidates.
The following tables show the Democratic presidential candidate vote margins among white voters and the final vote margin in each of 9 battle ground states. I included 2000 data for Gore where data was available. All 2000, 2004, and 2008 data comes from exit polling and final reported election results. The 2012 election has not happened so there is no final vote data and the 2012 white voter margin is taken from recent polling in each state.
In every battle ground state but North Carolina, the president is performing far better among white voters than he is nationally. He is performing just as well if not better in Virginia and Ohio than he did in 2008 (I am slightly skeptical of the Nevada and Florida numbers because other polling has shown his marginal disadvantage in the teens in those states). Finally, in those states where he is currently performing worse than he did among white voters in 2008, he is far outperforming both Kerry and Gore in 2004 and 2000 respectively.
The short take is this: Yes, it is the case that Obama is struggling among white voters nationally. No, this is not evidence that Obama is at significantly greater risk of losing the election than he was 4 days, 3 weeks, or 2 months ago. Obama’s struggle among white voters nationally is clearly driven by movement in deep red states, and the president is performing sufficiently well among white voters in enough swing states to conclude that his Electoral College advantage is far from put at risk by a waning support among white voters.
I thought I would take a little time today and work my way through Mitt Romney’s “full” 8 page jobs plan provided on his website. At the end of the “full” plan is an appendix that provides information on the assumptions and growth estimates that underlie Romney’s plan. Within that appendix are five sections. The one that I thought was most interesting was the one that contends fiscal consolidation — i.e. austerity — can increase short-term economic growth if accompanied by a modest tax cut. More specifically, it purports that austerity in the immediate term could feasibly increase GDP growth by 2 percent. Two additional percentage points would represent a full two-fold increase in current growth rates.
I found this striking for three reasons:
1) As MIT’s Simon Johnson explains, there is significant evidence that austerity policies are only capable of increasing growth in remarkably rare and narrow cases (of which the United States is analogous to none).
2) The body of work establishing that austerity policies (commonly referred to as contractionary policy) cause economic growth to slow — contract, and risk undermining economic recoveries is extensive and well established.
3) The Intentional Monterrey Fund, released a report (after having promoted European nations implement austerity policies in response to the global recession) suggesting that Europe’s recession has been worse than first forecast because of the austerity policies pursued in many European nations.
Now I could just write Romney’s economic plan off, but I think it would be a mistake to simply brush it aside without exploring the underlying research. As such, I decided to read the research Romney et al cite as evidence that all history, prior research, and current observation are wrong when it comes to relationship between austerity and economic growth. For those who are interested, here is the link to the study.
At this point, there are a lot of ways I could go about dismantling this study — the fact that the author is a Romney policy adviser, the fact that the study was published in the middle of this election cycle, the fact that the level of spending and taxes the paper analyzes are not the current respective rates. For the sake of time however, I think I will opt to address the study’s multiplier assumptions and the use of charts.
The paper Romney cites assumes that a 1% of GDP reduction in federal spending accompanied by (what amounts to) a 0.005% of GDP reduction in the capitol gains tax will have a profound positive effect on economic growth. This is only possible if the positive economic multiplier of cutting a single dollar in capitol gains taxes were 100’s of times greater than the economic drag of cutting a dollar of government spending. The following chart series reflect the theoretical assumptions of the work underlying the Romney economic plan.
To show how totally detached from reality this assumption is, I offer the following chart which depicts the historical relationship between reductions in the capital gains tax rate and per-capita GDP (in the U.S.). As you can see, the friendliest reading of this is to conclude that their is no statistically significant relationship between reducing capitol gains taxes and GDP growth (economic out put). A less friendly reading would conclude that cutting the capitol gains tax actually depresses growth.
To pile on further, here is the economic multiplier effect of a variety of different measures. The single boxed item reflects the per-dollar multiplier of a 5% reduction (from 20% to 15%) in the capitol gains tax rate (the rate assumed in the paper Romney cited was 2%). The group of items at the bottom reflect a variety of different forms of government spending. As you can see, the multiplier effect of a cut in the capitol gains tax (2.5 times greater than the one analyzed in the paper cited by Romney) is not even close to being 100’s times greater than that of different forms of government spending. In fact, it is a small fraction.
This leaves us with this. A) Mitt Romney’s entire economic plan pivots on the idea that austerity increases GDP growth — which all evidence reveals to be an objectively false notion. B) Mitt Romney’s entire economic plan pivots on the idea that the economic multiplier of cutting taxes is literally hundreds of time greater than what third-party analyses conclude. C) Some cut and paste combination of the aforementioned.
If you dig into what Mitt Romney has revealed about his proposed plan to spur economic growth, you find that it is founded upon a series of theoretical assumptions. Even worse, they are theoretical assumptions that have been proven flawed by historical analysis. Or perhaps I am missing something — some hidden multiplier or mechanism woven into his plan that I cannot read, see, feel, or find. Alas, I sense that I will never find the magic ingredient that will make these policies work this time, despite their systemic failures in the past.
Starting about three weeks ago, I began writing a series of posts that amount to a grinding rebuttal to the kind of day-to-day covering of the polls which cause people to substantiate false causal links and lend undue importance to relatively unimportant events.
I would recap all of my observations, analysis, and predictions here, but I won’t. At this point, doing so would establish a dangerous precedent (almost compelling me to start every post with 600 word overview of how we got to where we are). So, if you have not read my previous posts and want to catch up, you can follow these links (the last of the links has a short summary): Part-1,Part-2,Part-3,Part-4,Part-5,Part-6,Part-7
With that said, here we go:
Early on in this process — just as the public polling became highly volatile , I had argued (in essence) that the movement in the polls we had been seeing was precipitated by two separate events. One each effecting the two separate partisan voting blocs. The first was the Republican “coming home” — the consolidation of Republican support that pushed Romney up, and the second was the deflating effect Obama’s poor debate performance had on Democrats. It is this second of the two events I would like to draw attention to now.
After the first debate, there had been much talk about Romney’s commanding win having earned him a “second look” from swing and undecided voters. Moreover, many had concluded that it was this, that had induced the dramatic swing in the polls. I do not doubt that Romney earned himself many second looks. But the idea that the change in the polls had been induced by an en masse change in voter preferences is laughable. First, there are not enough genuinely persuadable voters in the electorate this cycle to reasonably conclude that movement among this group caused the change in the polls. Second, this understanding of what happened totally ignored the underlying data which showed a sharp dip in the percentage of self-identified Democrats in likely voter samples.
In effect, this meant that the movement we had seen did not reflect any sort of lasting change in the race. Rather, it reflected a temporary mood-swing among partisans.Which, at the time, was good for me (as someone attempting to analyze the movement we are seeing). Had the change in the polls been driven by movement among previously undecided voters without partisan-leanings, predicting future movement in the polls would have been much more difficult.
Anyway, because Obama’s drop in the polls had been driven by a mild erosion in Democratic enthusiasm, the question we were left with was not if, but when, Democrats would re-engage (thus lifting Obama in polls). As I wrote on Monday, it appeared that the “when” was this past weekend. As such, I predicted that we would see a steady rise in Obama support throughout the week as Democratic enthusiasm grew and Democrats gradually re-entered the likely voter pool. This is what we saw:
(CHART NOTES: I took this before the Gravis poll came out last night. As such, the horse race should be 47/46 for Obama. The debate lines show when samples were primarily post-debate.)
At every turn, I have weighed the arguments and evidence that some single event had fundamentally changed the nature of this election– that the 47% video really had sunk Romney’s candidacy or that Romney’s surge would continue to grow and carry into the waning moments of this election. And at every turn, I found these arguments and evidence wanting. This is not to say that the final two weeks of this race are immune to the possibility of some tectonic shift. It is to say however, that it is highly unlikely.
The fundamentals of this race favored Obama in January as much as they favored him in July, and nothing I have seen in the past few months (let alone the past few weeks) has provided cause to believe that this has changed. No doubt this race is tight (and likely remain a statistical tie for the foreseeable future). But just because it is tight does not mean that predicting the winner amounts to a coin flip.
The Economic Policy Institute has run the numbers on both Obama’s jobs plan and Romney’s jobs plan. In their analysis, the EPI compared how many jobs Obama’s plan would create in 2013 and 2014, and how many jobs Romney’s plan would create under two scenarios — one revenue neutral and one deficit financed — in the same time frame. To make the comparison, the EPI used the jobs legislation Obama introduced in 2011 (American Jobs Act), and the white paper Romney has put on his website. Here is what the EPI found:
Based on the EPI analysis, Obama’s preferred combination of spending increases and revenue increases would add an additional $5 billion to the debt and create 1.3 million additional jobs in the fiscal years 2013 and 2014.
If deficit-financed, Romney’s preferred mix of discretionary budget cuts, military spending increases, and tax cuts would add $690 billion to the debt and result in 641,000 job losses over the same time frame.
If revenue-neutral, Romney’s preferred mix of discretionary budget cuts, military spending increases, and tax cuts would add $135 billion to the debt and result in 1.3 million jobs losses.
This has nothing to do with ideology. This has nothing to do with which person is more likeable. This is about which candidate has a more plausible plan to help improve America’s economic outlook. The numbers speak for themselves.
A new swing state poll released by Gallup/USA Today found Obama and Romney are effectively tied among women in swings states. Considering the rather commanding lead Obama has held among women through much of this election, this findings marks a rather dramatic change. So dramatic in fact, that I thought it would be worth exploring it a bit further.
The “swing state poll” conducted interviews of likely voters across 12 swing states — NV,NM,CO,IA,WI,MI,OH,PA,FL,NC,VA, and NH — and sampled people in these states as if they were a single population. To better understand whether it is actually the case that Obama has lost his advantage among women voters in these states, I thought I would take a look at some of the recent polling in these 12 states. Further, I made a conscious effort to pick polls that tend to produce results that are more favorable for Romney.
Here is the two-way race between Obama and Romney in these twelve states among women:
Th ere is no way anyone could weight these kind of results and produce a reliable outcome where Obama and Romney split the women’s vote 49-48.
It is distinctly possible that Obama has shed a ton of support among women in recent weeks. It is highly unlikely however, based on the findings in each individual state sampled by USA Today/Gallup, that Romney and Obama are effectively tied among women. I think we can safely assume that the USAToday/Gallup findings are an outlier.
Over the past few weeks I have written a series of posts in an effort to push against the overly emotional analysis of excitable journalists and pundits, and provide willing readers with (what I would consider) a more level headed reading of what polling movement we have seen over the past few weeks. Here is a brief summary of my conclusions:
1) Most of the gains Romney made in national polls were precipitated by a Republican consolidation phase (“coming-home”), as evidenced by the large surge in Romney support prior to the debate.
2) The first presidential debate did have a real (“identifiable”) impact on the race. However, that impact was measurably acute and did not fundamentally change the trajectory of the race.
3) There was no evidence that Romney had persuaded pliable voters and that his base of support had reached critical mass.
4) Obama’s drop in support reflected a drop in enthusiasm among Democratic voters, and did not reflect meaningful movement away from the president.
5) Given (3) and (4), we could conclude that Romney’s support would not continue to grow (as many hyperbolic and hackish pundits predicted) but begin to wane slightly. Further, we could expect to see Obama’s support return rather quickly as Democrats re-engaged.
That in mind, the following is the RCP poll of polls as of late this morning. And much as I predicted, over the past few days we have seen a gradual erosion of Romney’s support and a rather fast uptick in Obama’s Support.
Trying to diagnose the nature of this election based on day-to-day events is a losing game, and it is why all the people who predicted that the 47% video had left Romney dead in the water, and or, that Romney’s first debate performance had sunk Obama were all wrong.
In the coming week (save for some truly extraordinary event in tomorrow’s debate), we should expect Romney’s support to continue to decrease as Republican enthusiasm continues to settle, and Obama’s support to continue to rise as Democrats continue to rally. I do not doubt for one second, that the narrative in the coming week (so long as Obama does not fall flat on his face) will be about how the second debate resurrected Obama’s foundering campaign. This narrative will be a false one.
Almost all of the movement we have seen in the past few weeks has been driven by predictable partisan behavior, and emotionally laden changes in enthusiasm. Put another way, this race is the same as it ever was. I hope Republicans have returned their seats to the full upright and locked position, because they are in for a rough landing.
In last night’s Vice Presidential debate, Paul Ryan discussed raising the Social Security retirement age as paramount if we are to keep Social Security solvent. Ignoring the broader fact that Social Security cannot become insolvent, this talk about raising the retirement age needs to stop.* For all the notional sense it makes, there is absolutely no evidence that raising the retirement age will, in any way, address the larger issue that the Social Security program’s liabilities are in excess of its receipts.
In a report from the CBO scoring the effectiveness of more than 30 different proposed policies, the CBO found that raising the Social Security retirement age from 67 (what the retirement age is currently slated to increase to in 2022) to 70 would do nothing to extend the life span of the Social Security trust fund within a 75-year window.
This is what the CBO said:
“This option … would not significantly extend the trust fund exhaustion date. After this option was fully phased in, scheduled lifetime benefits for people born in the 1980s and 2000s would be reduced by about 15 percent relative to current law. Payable benefits would decline by smaller percentages.”
So here’s the rub, raising the Social Security retirement amounts to a regressive across the board benefits cut — imposing a larger cut as a percentage of total income to low-income individuals than to high income individuals, and will NOT measurably extend the life span of the Social Security trust fund within the next (now) 73 years.
The Social Security trust fund risks “emptying-out” because Social Security’s receipts have not kept up with its outlays. An issue almost wholly driven by an impossibly low cap on the Payroll Tax that has not been adjusted to address the inequality in income gains over the past 30 years.
Currently the Payroll Tax is not applied to non-wage income — capitol gains, carried interest, rent, etc. — and only applied to regular wage earnings under $107,000. To give you a sense of how irresponsibly low the Payroll Tax cap is: In 1983, 91 percent of all earnings in the U.S. were subject to the Pay Roll Tax. In 2009, only 83 percent of earnings were (and no doubt that number is even lower today).
If we were to simply raise the Payroll Tax cap to capture 90 percent of earnings (less than the 91 percent that it did in 1983), not only would it extend the life span of the Social Security trust fund by more than a decade, it would allow for all scheduled increases in benefits. If scheduled benefit increases were capped, it would extend the trust fund’s life span even further. If we were to eliminate the cap on the Payroll Tax, the Social Security trust fund would be fully funded within view of the next century and allow for all planned benefit increases.
Paul Ryan’s plan to increase the retirement age for future retirees might make all the notional sense in the world, but it would do nothing to sure up Social Security’s structural faults. Further, it would impart tremendous harm on low-income seniors now and would do nothing to prevent the infliction of further harm in the future. Paul Ryan has fooled Capitol Hill watchers for years by wrapping his arguments in wonkish jargon, but he can’t fool me. He is nothing but a Ayn Randian ideologue in a shabby disguise, and his policies are nothing but a vehicle.
* As long as Social Security remains on the books it cannot become insolvent. The trust fund however, can. But once the trust fund runs out it simply pays out what it takes in on a yearly basis. So the whole Social Security will become insolvent and vanish is highly misleading.