Why is there superior economic performance under Democratic Presidents?

James Hamilton directs our attention to a useful new paper on this topic by Alan Blinder and Mark Watson (pdf).  Blinder and Watson conclude:

Democrats would no doubt like to attribute the large D-R growth gap to better macroeconomic policies, but the data do not support such a claim….It seems we must look instead to several variables that are mostly “good luck.” Specifically, Democratic presidents have experienced, on average, better oil shocks than Republicans, a better legacy of (utilization-adjusted) productivity shocks, and more optimistic consumer expectations (as measured by the Michigan ICE).

Perhaps one could attribute some of the “confidence gap” to policy differences, though the authors point out “…direct measures showing increasing optimism after Democrats are elected are hard to find.”  In any case this paper is a useful corrective to some common claims about superior economic performance under Democratic Presidents.  Invoking the partisan composition of Congress also does not seem to explain the observed patterns.

Since we are sometimes told that macroeconomic problems dwarf micro in importance (not a division of categories I would support, but you hear this often), well…draw your own conclusions.

Blinder and Watson also debunk a myth you commonly hear from conservatives:

In sum, with the exception of the Greenbook forecasts for the early part of the first Reagan administration, forecasts suggest little reason to believe that Democrats inherited more favorable initial conditions (in terms of likely future growth) from Republicans than Republicans did from Democrats.

This is interesting too:

There is, however, a slight tendency for both the nominal and real Federal funds rate to trend upward during Democratic presidencies and downward during Republican presidencies, suggesting that the Fed normally tightens under Democrats and eases under Republicans. Of course, such an empirical finding does not imply that the Fed is “playing politics” to favor Republicans. Rather, it is just what you would expect if the economy grows faster (with rising inflation) under Democrats and slower (with falling inflation) under Republicans—as it does.

In the UK, economic performance is overall better under the conservatives, although the difference is not statistically significant.


This post has only been up for 10 minutes on a Saturday morning but I seriously expected 100+ posts of a flame war.

With or without unremarked deletions? Because who would know how many posts have actually been made in such a purely hypothetical situation.

I think he's trying to start a riot, so he probably won't delete that much.

The readership here is full of anarchists, reactionaries and socialists, not Republicans and Democrats. Wrong topic for a flame war here because nobody likes either mainstream party much.

Nobody outside of this blog likes either mainstream party much either.

How would you classify someone who sometimes votes Dem, sometimes Rep? An anarchist, reactionary, or socialist?

"Nobody outside of this blog likes either mainstream party much either."

Maybe, but partisan flame wars are common on most blogs where politics is the topic.

"How would you classify someone who sometimes votes Dem, sometimes Rep? An anarchist, reactionary, or socialist?"

Such a person is a neopaleoconsocialist, of course.

cute, but contrarian is shorter and covers more commenters here

I disagree.

how contrary of you. I did not expect you to agree.

I've always thought that presidential elections were counter-cyclical. Republicans essentially get elected as a reward for good times, and as affirmation of market success. Democrats get elected when, wait a minute, things aren't going so well. That would tend to give both parties "regression to the mean," with the Republicans suffering the bad kind, while Democrats enjoy the good kind.

A reasonable, non-partisan, logical explanation. Thank you--that was definitely not what I was expecting in these comments.

It is sample size and most Democrats are elected towards the end of recession/depression when it appears the market economics is the problem (1932, 1960, 1992 for example and the Carter/Reagan was the reverse.) Also it does take years for such things to go into effect.

Of course one of the historical oddities is that Carter (airline, phone (with choice of judge), railraod and beer) had a very strong econimc deregulation record and should have put Volcker in charge in 1977.

The Koch Brothers. In the FTC Building. With a time machine.


My worry is lags. Most of these top level policies are slow to act. How do they account for that? Could one maybe consider the transitions? Say, R-R-D versus D-D-D etc.?

The fact that lags are likely to wash out across R's and D's means you shouldn't have to worry much about controlling for lags.

I didn't understand what you meant. Could you elaborate?

There is what is normally considered a significant signal in the data. Maybe it is spurious (likely it is spurious) but if it isn’t, what is a reasonable hypothesis which may explain it? A hypothesis which explains that the correlation is due to lags from the policies of one administration to the next or from one to several later administrations has a lot of problems:

1) Some of the policy effects would likely take place during the current administration (100% of the lag is not likely to fall squarely into the next administration.)
2) Of the 11 changes in administrations since WW2, 7 have been a party reversal, 4 not, so any “legacy effect” is diluted by at least 64% in terms of administration changes.
3) Of the 17 presidential terms since WW2, 9 were a continuation of the previous party, 8 not, so any legacy effect to the immediate next term is basically a wash.
4) Some legacy effects should spread over multiple later terms, diluting the effects across both parties further.

The point is not to explain why a signal should not exist in the data. The study says that a signal does exist (or appears to). The point now is to try to explain why, not why not.

For the reasons given above, it is unlikely that policies with lagging effects would be the reason one party’s presidency has a much greater correlation with economic growth than the other’s. If there is a non-spurious reason for the correlation, it is much more likely due to something where long lags are not an issue.

I guess I don't worry about lags because I don't count on this being super predictive or super important. History is just one damn thing after another.

The paper and the blog post address that argument and find that it doesn't match the data.

Really? I did look (skim). I didn't see anything that looked at "economic confidence" vs "party preference." When they say "forecasts suggest little reason to believe that Democrats inherited more favorable initial conditions" that is actually different than my suggestion. That seems to look past confidence to potential growth or some such.

In England, is economic performance better, or is the difference not statistically significant?

Are you really that confused? It is better, but that may just be random.

It is random, but it is better.

I can think of ways that makes sense. I don't think you're using any of them.

No I'm not actually confused -- just being pissy.

The point is that he undermined his own statement by first speaking historically, and then statistically. Why say both?

"Not statistically significant" isn't the same thing as "random". "Not statistically significant" just means "not necessarily not random".

Don't get me wrong, I think the phrasing you're criticizing tries to include too much information in too little space and as a result is just more vague than either stating "is better" or "has no significant difference." (A p-value and effect size would do much better in terms of specificity but I'm told people get confused when you include things like in writings for the general public.)

So, how does Thatcher rate in terms of an oil bonus? Because her time of being Prime Minister (followed by Major) pretty much parallels the ramping up of North Sea oil production, which peaked in 1999 - pretty much the time Labor took over government for another decade.

In other words, if the 'difference is not statistically significant,' then the difference between increasing oil and gas production compared to decreasing production would seem to indicate that the conservatives were simply better at being in the right place at the right time in terms of economic growth.

Except that during her era oil prices were way down, compared to the earlier Labor Party era. The oil had been there all along; Labor Party had a chance to benefit from high oil prices and blew it. The opening of the North Sea was due in part to Thatcher's policies, and she should get credit for them.

Maybe Democratic presidents cause more confidence, not through policy, but because they are more confident leaders. If confidence numbers don't change much right after the election, it's likely because people often know who is likely to win months before the election. (Another hypothesis is that elections increase confidence and that confidence tends to decrease after a Republican win, but one would think this study would have tested for that.)

Note that Reagan was the one Republican president of the last 50 years with the confidence of a Democrat, possibly because he was more like a Democrat at heart even after switching parties.

Personally, Reagan's party affiliations over his life pale in comparison to his ability to act the role of president better than anyone in my lifetime.

In contrast, USNA graduate Carter's lack of such ability was the most glaring failure of his term. The fact that one followed the other adds to this perspective about the importance of being able to play a role, as compared to actually understanding the sorts of things taught at one of the world's premier military academies to those expected to play a strategic role in America's future.

"Ability to act the role of president" is exactly what I mean by "confident leaders" here. Notably bad actors were: Nixon, Carter, Ford, Bush and Bush Jr.. Actually Bush Jr. was often good with the display of cocky overconfidence required, but then he'd undermine that overconfidence with those deer-in-the-headlight moments.

Does executive confidence lead to consumer confidence? Who knows -- but the idea doesn't seem entirely crazy.

The a American President is head of state. Acting the role is part of the job description.

Neat paper (I love anything with sentiment in a starring role) and I will read it carefully. Maybe the authors are more careful in the body of the paper, but the labels of "luck" factors (even in quotes) in the conclusion seems a bit of stretch given the topic. For example, they argue: "It seems we must look instead to several variables that are mostly “good luck.” Specifically, Democratic presidents have experienced, on average, better oil shocks than Republicans, ..." Well maybe but oil price shocks do not come from a random number generator. Without getting anyone riled up, it seems possible that the foreign policies of Presidents Bush (both) versus President Obama may have fostered different oil price shocks, and that difference might generalize to other presidencies. It is simply too odd for all the "luck" to attach itself to one party, especially if initial conditions are similar. Likewise, confidence is rarely an out of the blue "luck" factor - it is tied back to economic outcomes that policies may affect. In any case, interesting paper with lots of empirical nuggets and possibly big takeaways.

Disregarding "luck", the sample size is small. Though statistically significant, I would like to see some analysis of correlation, predictability, and influence of outlier points. Before searching for causality, I'd like to be more convinced that there is really something robust going on.

yes, welcome to macro ...

The statistical significance is drive by effect sizes, not on the likelihood of clustering of particular events. That oil shocks are not the product of a random number generator suggests that using the t test is inappropriate to gauge the probability of chance occurrence.

There are also inverse oil shocks, such as the decline in prices under Reagan and in some respect as we are undergoing now with increased shale gas (fracking) and shale oil. these are hardly foreign policy events and have very long time lags between investment and production. Had energy gone to trend, I suspect the economy would be much worse off now. Attribution for the break from trend is a separate question that seems unrelated to foreign policy, at least directly.
One oil shock hit under Carter, but Reagan bore the brunt. Carter began deregulation of domestic oil but Reagan extended the deregulation to old oil. The anti-shock breaking OPEC for the time was more or less bipartisan and irrespective of foreign policy.
Clinton, if memory serves, avoided an oil shock by using the strategic reserve that had been build in previous administrations.

Again, small sample, and likely a statistical illusion. Conventional statistics exaggerate the remoteness of chance and timing.

"That oil shocks are not the product of a random number generator suggests that using the t test is inappropriate to gauge the probability of chance occurrence." I am pretty sure that is not entirely correct. As long as the shocks follow a stable distribution (of course, you need to pull out trends or unit roots, which I believe they do), one can draw inference from t-statistics. Yes, small sample sizes or more precisely the fact that we only get one run at history (we can't see what the same world with would have been with a Democrat Bush or Republican Clinton) makes it hard to draw firm conclusions. I would not want to underemphasize chance (or pure noise) but sometimes it is a cop out too.

I still consider the Iranian oil shock to be The Big One. That hinged more on Iranian "administration change" than American. Unless you go back to 1953 and assign an American influence. If so, your blowback is 25 years later, and not nearly in the same US presidential term.

All in all, I see confirmation for my idea that each election is a response to current (or immediately previous) conditions, and that reversion to the mean is likely to take things in the opposite direction.

It may be odd for luck to attach itself to one particular party, but an unspoken assumption you're making is that the parties and their worldviews are roughly stable over time. Nixon favored price controls/expanding the welfare state, ended the Vietnam War, and began normal relations with China, but he still suffered a an oil shock.

True enough, but the entire world view may not affect the economy ... only a few pieces may be relevant and those might be constant. The shifting world view actually makes it even more surprising that there are significant differences.

Maybe luck is the wrong word. Aren't you economists always controlling factors? How about if they had said "Controlling for oll price ....."

Their word not mine but oil prices are sometimes considered exogenous to policy, like weather events or natural disasters. So yes it's controlled for but where do the movements come from? Who can claim responsibility or must take the blame? Relatedly, I spend a lot of time trying to parse and forecast confidence and I would not call it a "good luck" variable, it's a filter of a lot of real factors.

My favourite way to interpret this while confirming my existing priors is,

Isn't this mostly saying that we don't really know how to improve the economy at a macro level? If it were in the power of any given actor, why wouldn't you pump up the economy and guarantee not only a smooth re-election but a favourable view in the hindsight of history? If our standard measurement of "good job, mr president" is largely unaffected by their policies or political bent, then I would definitely prefer to optimize towards those seeking to minimize human suffering; this is definitely increasing my desire to experiment with more social programs.

Maybe the lesson to draw here is that given that almost by definition anyone elected to power in the US is going to be pretty conservative/pro status-quo by nature everyone tries to rock the boat as little as possible. Or maybe absent major exogenous (regulatory, market) shocks, the mass of people from which the economy is made of mostly keeps humming along.

"My favourite way to interpret this while confirming my existing priors is, ..."

Now there's a quote! We should all understand that this is what we do.

I think that's automatically added to every new comment on MR these days.

In the past 100 years I can only think of 2 Republican presidents who implemented pro-growth policies: Coolidge & Reagan. On other hand Hoover, Nixon and the Bushes actively pursued anti-capitalistic policies. Seeing that presidents of the same party can be in such disagreement on economic policy what is the value of drawing partisan conclusions?

What I find most illustrative is that the 1990s are generally recognized as a very prosperous decade. It is also the ten year period (1991 - 2000) of the lowest increase in government spending since WWII began. Note that if sequester level budgeting persists the 2010s might be a repeat of 1990s spending patterns. If so great things may be in store for the economy in the last half of the decade, but only if a Democrat is president and the purse strings are controlled by the opposition party.

Out of curiosity, what is your explanatory narrative here, though?

Is that variable important under "government crowding out private investment" or some other mechanism?

I subscribe to the crowding out theory. To be more specific, I believe that economic growth depends on positive economic activity and you get more of that happening as the government's role is reduced.

Crowding out of what? The 2000s had easy money and high unemployment. Business had the opportunity to both expand and hire. They did in fact, following the consumers who were leveraging up and building out single family homes on a mass scale. What the heck did GWB crowd out?

When it is far more productive to hire connected fools to get government money vs. hiring innovative game changers at high risk, the economy suffers.

I am no fan of foolish programs, but I really don't see government spending overpowering the consumer market in recent US history (if ever). Why is Apple #1? Did they do it by chiseling tax dollars?

Housing? Energy? Health care? A few off the top of my head.

I don't really follow how those categories really "crowd out." They are really diverse. In housing we do subsidize the 30 year mortgage, and without it we'd be much more a nation of apartment dwellers. Are you saying you'd choose that sort of economic efficiency? (I might too, but I see it as a political nonstarter, and I'm not really sure how more apartment living would actually lead to more economic activity. Quite the opposite, it seems a way for each individual to live more efficiently, with lower spending.)

Lower spending on housing would leave more room for spending on stuff other than housing. It also frees people to move to other areas in search of jobs more easily. It is hard to sell a house and leave a declining region because everyone else wants to do so at the same time. To the extent that people value owning their own homes, let them do so, but having more people choose to own rather than rent is hardly a public good.

Obviously you can't buy more "stuff," Mike. It doesn't fit in your apartment. So what then, more travel? Possibly, but I'm skeptical that "experience" spending could really match the "stuff" spending that goes with a housing boom.

@Mike Owning a house can give one more of a sense of belonging to the community, which can have beneficial effects. As a concrete example, I once found I wasn't invited to my neighborhood homeowner's meetings because... I wasn't a homeowner, just a renter. But my landlord, who lived on the other side of town, attended the meetings.

So how did that work out? The purpose of encouraging house ownership was to make people better off. Didn't work. Could it be that price fixing in mortgage markets distorted the pricing and in fact made people poorer? Or that simply making borrowing cheaper took competitive pressure off the various parts that make up a house price. For example, how much of the subsidized interest rate driven price increase is simply paying for the very high land development costs imposed by state and municipal regulation? In other words a tax increase.

We saw here in a post that government sets the price for medical care. Removing the price signals from a marketplace creates very perverse situations.

Crowding out is your term. And yes, there are a bunch of other influences, but with housing, the GSE's controlled 40% of the mortgage market and were as bankrupt at the end of 2008 as Lehman. Prices are set by the marginal buyer, and when it is government, who has the resources to change it? Try pricing Treasuries differently than the Fed does when it buys 85 billion or so of them.

Note that I said above "I might too, but ... Quite the opposite, it seems a way for each individual to live more efficiently, with lower spending."

I'm completely fine with eliminating subsidies for home ownership. I'm just skeptical that such a move would produce growth. More likely it would produce efficiency in the other sense.

"The purpose of encouraging house ownership was to make people better off."

When I grew up, the primary objective of a homeowner was paying off the debt so you actually owned your home. I have done everything I could to pay off my homes and I found the whole idea of cash out refis that all my friends were doing to be total idiocy, and that was in the 90s.

I thought the advise from realtors in the 80s to borrow lots more to buy a bigger house to be sheer idiocy. Then in the Bush years, it was constantly about getting cash out of houses that were insanely large and expense to get rich without working and saving.

From what I could tell, the Bush policies had nothing to do with home ownership, but all about getting into the pump and dump real estate speculation ponzi scheme to get rich the only way you can if you aren't connected enough to create a startup that IPOs and makes you rich.

No one in the past two-three decades have been promoting home ownership because no one was saying "you must pay off the mortgage as soon as you can so you can use the mortgage payments to pay your kids tuition and then for your retirement."

They were saying "become a real estate speculator by leveraging up as high as the lenders will allow you because stock speculation is 50% down, but real estate speculation can be zero down."

But hey, what do I know - I'm a kid who grew up in the 50s and 60s when homeownership meant you owned it free and clear, not when homeownership means you owe all your income to the bank or you have no place to live.

Just to add, that is bipartisan. The military spending buildup during the Bush years did the same thing as the green energy 'investments' more recently.

In some way the original post is a very simplistic. What about Congress? Clinton had a vigorous Republican Congress intent on stopping him. So did Reagan. Bush and Obama arguably did their worst when they had the Congress to go along. What about Kennedy cutting taxes vs. Bush 1 raising taxes? Or Bush 2 with Medicare D and the CPSC, vs Clinton welfare reform, or Johnson? Or more interesting, Carter deregulation and appointing Volker to the Fed? What about the time lag for policies to have effects? Something like deregulation or Johnson's welfare implementations will have some immediate effects, but the full effect will come maybe two terms later. How much of the Democrat gains are akin to Nixon going to China, ie Clinton welfare reform and Carter deregulation?

Would less military spending have made Apple's $1.70/hr assembly work in China look less attractive? I don't see it.

I'm not sure what point you are making. Apple is somewhat of an outlier in the US economy; growing and innovating, facing competition from the far east head on.

My point is simple. If you want to make money easily, low risk, get yourself in line for a government subsidy of some sort. That is what Wall Street is doing, someone has to handle the trillions that between the Treasury and Fed are sloshing around. Get to Washington, work as a military contractor. When Obama showed up, be connected and say 'electric car' or 'wind energy' or 'solar' lots, get money. There is always lots of this type of nonsense going on, and generally is a cost rather than a benefit to the economy. The more of it the worst; see US Housing Policy. Or even worse, Detroit, the beneficiaries of decades of accommodating government policy on all levels.

Both parties do this. Ford bailed out Chrysler iirc, Obama bailed out General Motors. They both have their client groups and industries. Some are bipartisan, such as agriculture. Some subsidies are great, some incentives or leaning towards a direction may be fine. But I suggest that larger the influence in an industry the federal government has, the more it will be characterized by the rent seeking mentality than by innovation. The great stagnation may be the system working as designed.

An interesting case study. The Canadian dairy industry is highly regulated by a supply management system. The most valuable asset a farmer will have is milk quotas. (Quebec is a different story). The industry has been characterized by amalgamation and homogenization of the product lines over the last years. A stagnant industry with little growth. Recently a trade deal with the EU was signed in a preliminary stage, and the Canadian dairy industry will probably be steam rollered by an influx of european products that are innovative and geared to the consumer, rather than geared to the producer. The Canadian producers will probably lose market share as they are stuck in a regulatory structure that prevents them from responding to market pressure. The marketing board was the market, it no longer will be.

Quebec is a different story because the regulators in that province, a small piece of the market with a regulator with an attitude, encouraged the development and continuation of local specialties. I suspect the European regulators are similar, the products and specialties precede the creation of the EU.

My point is that Washington policies and incentives are counter productive in most cases, and harm economic growth. Too slow, too far from the ground to react, too many opposing interests to influence policy. And by definition anyone willing to work in such a place is going to be a net loss to the economy. The less of it the better.

I don't deny that governments perturb markets far more than they should, but I just see the big trending factors as globalization (esp. as off-shoring of production) and automation. I think the US government has been the tail on that dog.

It's much less likely that a policy with much lag time would show up as signal for one party than that something without a lag would show as signal. Policies which introduce long and variable lags are much more likely to wash out as noise. Hence my hypothesis: perceived confidence of the president by the public influencing consumer confidence, has no lag issue.

Oh we agree that people are just voting current confidence.

Hasn't the drive to off shore manufacturing been to get away from out of control costs imposed by the various levels of government? Intel says it costs another billion dollars to build a plant in the US compared to elsewhere. Is that not showing the dog is in fact the diverse agencies of government?

It's interesting, derick. People suggest again and again that "the drive to off shore manufacturing [has] been to get away from out of control costs imposed by the various levels of government."

But this $1.70/hr number stares us in the face.

No amount of "government subtraction" in the US will lower manufacturing wages to $1.70/hr. Nor would we really want it to.

JFK never got his tax cuts. First there was concern about the debt and deficits. Then with the economy recovering, the question was "why cut taxes and cause too much growth and thus inflation from higher wages?". It was LBJ who pushed it through, something he could have easily done if not for RFK reinforcing the distrust of LBJ and keeping him sidelined.

And the tax cuts only reduced the tax rates to unimaginably high tax rates for any conservative today, tax rates they would consider total confiscation and thus sure to kill all economic growth and all job creation. Those rates were concurrent with high rates of job creation and GDP growth. LBJ then hiked taxes a number of times to pay for SS and Medicare, and then hiked taxes to pay for the cost of war, with a 10% war surcharge added to your tax bill after you calculated your tax bill. The war tax made Vietnam extremely unpopular, and Nixon was promising to end the war and the war tax, something that trumped fear of "losing Vietnam to commies with the rest of the world tumbling like dominoes". Another reason for the war tax was to cool the economy because wages were rising due to too many jobs to fill and too few workers. That situation changed in just a few years as the boomers finished college along with their high school and trade school peers.

I remember the alarms over the ups and downs in the economy in the late 50s and 60s as a public school student, and those were given as much coverage as the cold war and duck and cover and sputnik. Unemployment increasing 0.2% was like a crisis requiring government action, while it falling raised the specter of hyperinflation. And then, the government regulation of banking was killing savers because the best they could earn was 4.5-5% at an S&L, and it was just too hard to borrow to buy a house because you needed 20% down unless you were a veteran, and new car loans required you have a good job for a year.

The economy performance in the 60s was great compared to the economies since 1980, and that was a time of war, cold and hot, but the things economists today say are highly destructive to the economy were all present and would today be called draconian if proposed today.

And with all the draconian central planning from Washington and the crushing taxes, in the 60s the idea was that eliminating virtually all poverty and moving the last few million into the middle class was possible.

Since 2000, the virtuous deregulation of banking and slashing of tax rates has failed to deliver growth and has significantly hollowed out the middle class. Meanwhile, the elimination of Regulation Q to let savers earn more than 5% has led to savers getting 0%.

I wonder how economic performance correlates to control of Congress? To the extent that growth in the '90s had anything to do with government, was it because of Clinton or Gringrich? I suspect it was both.

Bingo! Look at Congress. Not sure which way it will go overall, but Gingrich was the last one to balance the budget.

My view is that people get conservative and elect Republicans when thing aren't going well, then elect Democrats when they feel a bit more free-spending.

Small sample size and the power of the incumbent in any economic situation to get reelected really overpowers any of the other variables though.

That would seem to be the largest question. Presidents are only so powerful.

Most important variables: Change in, and absolute values of, ((Federal Government - Interest - Social Security - Medicare) / GDP). Inversely correlated with GDP growth. Clinton has a great score, great economic growth. Reagan good score, good economic growth. GWB lousy score, and ...

I don't know if they should treat oil shocks as exogenous. It may be Republican belligerence in the Middle East that causes higher oil prices during their administrations. It was Nixon's decision to help the Israelis in the Yom Kippur War that resulted in the OPEC embargo. On the other hand, one could easily argue any Democratic president would have done the same given our relationship with Israel. A better example is probably our wars with Iraq and their effect on oil prices.

Sample size

I'm interested in how they handle lagged variables. Presidents seldom have programs go into effect immediately, although Fed policy can be immediate, so looking at how both are lagged, or not, is relevant. So far, I am a little skeptical of papers that talk about "growth rates" when you could inherit a bad economic situation, and the only way to go is up if you believe in reversion to the mean.

Also one must question whether the underlying measures measure what the authors think they do. That GDP growth was astronomically high during the "Great Depression" is hardly news, but it should give one pause about using GDP as a proxy for economic well being.

I flipped a coin 100x and it came up heads 53x, which I think is pretty conclusive proof that heads will always outperform tails.

If the results aren't statistically significant the authors (and the post) should have simply stated that and moved on, so it's reasonable to assume some level of significance without reading the paper.

Of course it's likely blind luck, but it's interesting to ponder the possibility it's not and why.

Tables 2 and 3 pg 45-46 in PDF: http://www.princeton.edu/~mwatson/papers/Presidents_Blinder_Watson_Nov2013.pdf show some of the differences. p-values in the last column less than 0.10 would be conventional for 'statistically significant' differences. There are several, but it's not across the board.

Lags, control of Congress, 'Nixon to China' effects (GOP Prez enact welfare reform?)

Doesn't sound so useful.

"Democrats inherited more favorable initial conditions (in terms of likely future growth) from Republicans than Republicans did from Democrats. "

Obama did inherit the success of the TARP program , allowing him to take credit for the huge bull market & economic boom when in actuality Bush, Paulson, Bernanke deserve the credit while according to an economic consensus, the Obama stimulus was a dud. Had Bush been given two extra terms we would probably be at dow 20,000 by now. Clinton inherited the success of Reaganomics which undid the hyperinflation of the Carter administration. Republican policies are more conducive from growth; in fact, even Bill Clinton was forced to adopt some such as welfare reform and NAFTA

you cut off a key part of that quote "forecasts suggest little reason to believe that ..." regardless of your examples they don't find a pattern as such.

I seem to very dense. Unlike so many people I can't seem to grasp how Clinton inherited the success of Reagonomics but Bush I didn't.

And the economic success of TARP? Is that supposed to be parody?

TARP did most of the work stopping a possible melt down, and ended up being 90% paid back.


You take your dog to the vet to kill the tapeworm, not to make the tapeworm happy and healthy. That the tapeworm proves willing and able to cough up for a big share of the vet bill at some later date should be irrelevant. And if the dog remains sick for many years but both you and the vet are bragging about how good the tapeworm is looking, well, I know you aren't going to want to hear it, but trust me on this - your vet is insane and it's best if you don't own a dog.

I wouldn't expect large differences in performance, but I would expect large differences in what They Say about the Performances.

Bah, this doesn't accord with my (current) mood affiliation.

The whole thing sounds like (Republican) excuse making and sour grapes to me.

As an example of Democrats vs. Republicans, there seems to be no doubt that Clinton was head and shoulders above Bush Junior in economic policy (except for that one big mistake he made listening to Republicans, economists and bankers where he supported repeal of Glass Steagall).

Clinton got the economy on an upswing and left it in a recession, rode a huge bubble, and watched it burst.
I don't think his policies were bad, but he hit a sweet spot in time going in and coming out.

Bush got the economy on an upswing and left it in a recession, rode a huge bubble, and watched it burst. I don’t think his policies were bad, but he hit a sweet spot in time going in and coming out.

Exactly this. Clinton inherited a huge Republican deficit and left a surplus. GWB left a deficit to the tune of $5-6 trillion, and Obama has been little better than GWB in that regard.

Clinton benefited from a tech boom, and GWB benefited (and later suffered) from the financial/real estate boom.

I think you'll find that economic performance correlates more closely with Congressional party. In particular, when congress is controlled by the same party, economic performance deteriorates, regardless of the party. When there is split-party control, economic performance improves. Presidential party had lifetime to do with it, independent of Congress.

nope, no significant difference and you don't even have the sign right, see Table 5 http://www.princeton.edu/~mwatson/papers/Presidents_Blinder_Watson_Nov2013.pdf

Why is there superior economic performance under Republican Congresses?

And why do we keep hearing about the Democratic presidents and never the Republican Congresses?

maybe cause it's not true?

Table 5
Average GDP Growth by Congressional Control

Democrats control both (n=168)
3.47 (0.47)

Divided Congress (n=40)
2.70 (0.87)

Republicans control both (n=48)
3.36 (0.53)

standard errors in parenthesis. note, none of the differences are statistically significant at conventional levels ... Congressional control does not matter on its own.

from the paper: "Table 5 displays average GDP growth rates when the Democratic Party controls both houses of Congress, when control of the two houses is split (regardless of which party controls which house), and when the Republican Party controls both houses. We see that the average growth rate is highest when Democrats control Congress (3.47%), but the difference with Republican control (3.36%) is trivial. Apparently, it’s the president, not Congress, who matters."

Sigh, here we go again.


Note the age of this post. It is so old Google has to call out an old emulator for the page layout. But the argument still hasn't changed.

And the post references an even older post!

Question for the more mathematically trained: how many more presidents would it take for people for it to basically become highly problematic to ignore this correlation?

C'mon people, do you really think there's a difference between the two parties in the USA? Bush II increased federal spending more than Clinton, while Reagan was a populist who increased federal debt. Carnival of ideologies (A mural by Mexican painter J.E. Orozco in the Palacio del Gobierno, Guadalajara, Jalisco, Mexico)

The only constant is what Todd Ramsey mentioned upstream, which is that increasing government, which mostly does nothing useful, will eat into your GDP growth.

I'm surprised you would pay any attention to something so silly, Tyler.

If someone wanted to do a serious analysis, they would have to weigh control of the legislature and judiciary, to say nothing of downticket. And even then we have to adjust for the ever-shifting status quo.

Anyways economic policy has its churn -- for a long time the Dems were the free-trade party.

They aren't now? Standard Dem's passed NAFTA against the opposition of the far right (recall fears of the "giant sucking sound"), and now Dems are starting to favor free trade in labor/immigration. Republicans more and more oppose free trade in labor, and I still get emails from my Tea Party uncle saying that we need to repeal NAFTA and join Michele Bachman in stopping Obama's plot to "enslave" us through the Asian free trade deal.

The donor class of both parties is pro free trade, the actual voters are more skeptical but who cares what they think?

Why must you cynics be correct? Can't you see we have an oligarchy to maintain?

The crucial part of the paper, to any practitioner of frequentist statistics, is "The p-value for quarters-in-recession is also 0.01, so that the lopsided realization of recessions is similarly unlikely under the assumption that party and economic performance are independent."

In other words, randomly shuffling the parties in a Monte Carlo simulation suggests that it is unlikely that the Dems are better by chance.

This seems to contradict the final conclusion that Dems do better than GOP because of 'good luck' (which explains only 46-62% of the 1.80% Dem advantage anyway). The frequentist stats already established that such good like is unlikely at the 99.9% level. You then have to explain why Dems were so darned lucky!

Liberals claim correlation proves causation.

Republicans simply ignore all data. And science. And thinking in general. Much cleaner.

Let's keep electing Democrats then and test it out.

Maybe it's not Democrats exactly, but rather that in general Democrats tend to follow one approach and Republicans another. Not to blogwhore, but here's how I interpret the data: http://angrybearblog.com/2013/11/previewing-blinder-and-watson-2015.html

Make me LAUGH while you're at it. Everything in the last 60 years WRONG with America has been the democrats doing and it all involves money, LOT'S of it. We've only been trying to dig ourselves out from underneath the democrats little HOUSING BUBBLE for SIX long years now and that is highly doubtful it will EVER happen in obamaland. Also the WOES in our healthcare system are ALL due to LBJs trashing and his addition of medicare and medicaid and now on to our FATAL WOUND obamacare. Who's paying for all this democrats I'd like to know GREECE?

I wonder if this is really the right question.

Presidents certainly have a big effect on Foreign policy, but on Domestic policy? Obama has been President for almost 5 years but only really had a liberal domestic policy for the first 2 years when Democrats ruled the senate. Since 2010 I think Republicans have been setting policy more than anyone else through their control of the house of representatives.

Everyone looks at the President as the main variable since he's a single super-identifiable figure. But on issues of domestic policy and the economy I suspect he's more a figurehead and the real driver is congress.

But what kind of deficits?

They note that Dems and Reeps have similar levels of deficit spending -- but presumably, those deficits under Dems are driven relatively more by growth in spending, while under Reeps, by shrinking revenue due to tax cuts.

Growth in government spending is a part of overall economic growth. Thus, by definition, in the short term, all other things being equal, higher deficit spending = more economic growth.

my initial reaction was one of surprise that a scientific or economic study would attribute something to luck.
if it is something that is occurring in multiple outcomes how could it be luck.

might it be as simple as how money flows. money belongs and comes from the government. corporations cannot "make" money. they can profit which means they get a dollar if someone decides give a dollar. if the government spends then there is more money in the economy to be had. the more money and movement the better the economy. so my non researched guess is democratic spending is a good place to start from as a hypothesis.

Way cool! Some extremely valid points! I appreciate you writing this write-up and also the rest of the website is extremely good.

Comments for this post are closed