Investment vs. the Warfare-Welfare State

In Launching the Innovation Renaissance and my Atlantic article The Innovation Nation vs. the Warfare-Welfare State I showed that the Warfare-Welfare state has crowded out federal investment in research in development.

In a short report titled Collision Course: Why Democrats Must Back Entitlement Reform, Jessica Perez, Gabe Horwitz, and David Kendall cut the data in a slightly different way but come to the same conclusion:


Entitlements are squeezing out public investments. In 1962, spending on investments was two and a half times that of entitlements. But today, as a result of this Great Inversion, entitlement spending is three times that of investments. And this trend will only accelerate in time as the Baby Boomers retire and their benefits grow faster than inflation and wages.

…The fact that entitlement spending is crushing investments is bad news for U.S. growth.

Hat tip: Arnold Kling.


Once full employment is attained, increased investment must come out of some component of consumption, not necessarily from that portion of consumption financed by "entitlements." That realization liles behind most Democrats favoring a combination of entitlement reform and tax reform as a way of addressing the long term budget trajectory of the US government.

Which Democrats would those be?

"consumption financed by “entitlements.”"

Another good example of doublespeak.

Imaginary ones.

One would be the President who sought the "grand bargain" Perez, Horwitz, and Kendall advocate only to find House Republican leadership incapable of closing the deal.

Which President would that be and what were the details of his plan?

Clever juxtaposition of images.

On the warfare side, the president has not been good for cutting defense spending. I expect that Romney would be worse. Could Obama spend less than the appropriated amount on defense, therefore avoiding the debt ceiling, etc.? Or is he legally obliged to spend the whole defense budget? I don't think that he is champing at the bit to do so, but I was wondering this question while looking at the article. Obviously a president cannot unilaterally cut social security or medicare spending or increase R&D expenditures.

It is my understanding that not spending appropriated money - impoundment - is unconstitutional. That seems like a good thing, as giving the president a de facto line item veto is quite a centralization of power.

Not unconstitutional, just a violation of statutory law. This is a remnant of the post-Watergate legislative muscle flexing vs. the executive branch.

unless they are willing to impeach the president or his officials for it, its meaningless.

Didn't the Supreme Court strike down Nixon's attempts to not spent Congressional appropriations?

It is my understanding that Romney has accused Obama of turning his back on U.S. security (or some such nonsense) and that he wants to restore this lost security by increasing defense spending. So liberalarts suspicion is confirmed by Romney's explicitly stated policy preference.

Also, in a post entitled "Investment vs. the *Warfare*-Welfare State" (my emphasis), why is there no mention of defense costs? A related line on the chart would have been germane and informative...

"entitlement spending" is a non-specific category, right?


Another point: this is a case of crocodile tears. Let's say Congress and POTUS were able to rein in entitlement and defense spending. Wouldn't the next round of grousing be of the "government is bad at building/maintaining things, leave those things to private enterprise" ilk?

“government is bad at building/maintaining things, leave those things to private enterprise” ilk?

No. I think that it's not a coincidence that as the government got out of the legit public good business the anti-government meme began to stick.

Thus, the liberals really are for "more government" as an aggregate without much analysis as to whether the resources are completely wasted, because in their mind "more" is by definition not a waste.

That's your mood affiliation talking. Not all liberals are for bigger government.

I'm in favor of smaller but smarter government. My preferred model would cut some entitlements, lop the defense spending in half or more, but spend more on infrastructure upgrades/maintenance. Government R&D I'm not so interested in.

I think that model would result in a net contraction in "government size" (if size = expenditures); I could be wrong.

I wish I had some mood affiliation.

What kinds of infrastructure upgrades/maintenance?

Why not "government R&D"?

Why would you cut some entitlements?

And why would you cut defense in half or more?

Infrastructure upgrades/maintance projects: shoring up the national grid, water infrastructure, roads/bridges, rail improvements (not new buildouts). I'd ensure that central services (e.g. the CDC) are fully funded.

I'm not exactly clear what falls under the aegis of R&D. Military R&D? Science R&D? Medical R&D?

In terms of entitlement cuts I'd crack down on abusers (e.g. welfare and disability). I'd increase the retirement age levels. I'd crack down on employee pension abuse and also redo pension terms. That certain government employees can "retire" after 20 years of service and start collecting pensions while working a new private sector job makes no sense.

I'd cut Congress's perks. A drop in the proverbial bucket, but a signal to the plebes that they'll live like the rest of us (sort of).

I'd cut defense for obvious reasons. We're grossly overinvested. Further, we as a nation suffer from "when you're wielding a hammer, everything looks like a nail" syndrome. We're broke. We need to fix things at home before we should even consider white knighting the world over.

Means-test Social Security.

Only if I can means test pensions.

Social Security is already sort of means-tested. Depending on other income, as much as 85% of SS can be taxable, and the exemption thresholds are not indexed, so there is some future, gradual means-testing already built into the existing structure.

How about some means-testing for public sector pensions while we're at it?

Social Security is already means tested. If you make over a certain amount (not sure the figure), some of it is taxed back.

Maybe the book is better than the essay but there are just a lot of mixed ideas in The Atlantic piece that confuse things. I don't think that it's fair to include Social Security and Medicare as Federal spending that is crowding out investment. Both have separate trust funds and certainly Social Security is pretty solvent for a number of years to come and the fix to improve this is pretty straight forward. I will grant you that Medicare is a separate issue but still one that can be solved. The statement that the AFCA ".. was almost entirely about welfare and redistribution, about dividing the pie..." is incorrect. There is a good part of it that will focus on cost control. Confirmation of Dan Berwick would have helped move things along in this regard but the recalcitrant Republican Senators don't seem to want to explore any new possibilities here.

I am puzzled by your statement on reducing cancer mortality by 10%. With the exception of blood cancers (including lymphomas), chemotherapy has not been terribly helpful. We spend lots of money on treating solid tumors to prolong life by maybe six months at great cost (best example here being colorectal cancer); solid tumors are best treated by early diagnosis and surgery. You state "... The net gain would be especially large if we could reduce cancer mortality with new drugs, which are typically cheap to make once discovered. A reduction in cancer mortality of this size does not seem beyond reach. Medical research spending is far more valuable on the margin than medical care spending yet because we lack an innovation vision, we endlessly debate how to divide the pie while we overlook potentially huge improvements in human welfare." I spent my entire working career in biomedical research and drug regulatory affairs and I think this is just plain wrong. The National Cancer Institute is the highest funded of all the NIH institutes and has been ever since Nixon announced the war on cancer. I don't think that this effort has suffered for a lack of funding or vision. It has suffered because the problem is very difficult. Even if we get some medicines developed, they are likely to focus on small patient populations where the genetic basis of the cancer can be attacked. Sure the pills/infused medicines might be "cheap" to make but the cost recovery and amortization over a small patient population means that they won't be cheap (just look at how much some of the biotech drugs for arthritis, Crohn's disease, etc cost relative to a cardiovascular drug such as Lipitor; it's a factor of 10 different).

I agree that there is a lot of regulatory red tape that is in need of fixing. Certainly the Clinton, Bush-2, and current administrations have had programmatic assessments to do just this (I think Cass Sunstein has not been given enough credit in this regard). Yet a lot of what plagues us are the myriad of state and local regulations as well as NIMBY that can cause more problems (wind farm off of Martha's Vineyard is a good example here).

If Governor Romney is as supportive of the Paul Ryan budget as he appears to be, things will only get worse. His statement on increasing the defense share of a decreasing pie is particularly troublesome (why do we spend as much on defense as the next 17 countries in the world combined?). Retreating from being the world's policeman would free up enough money to deal with what you argue is an investment problem. Zeroing out Social Security and Medicare will not.

Social Security and Medicare are transfer payments (they aren't forced savings...the money has long since been spent). The transferee has a mix of consumption and savings; the transferor has a mix of consumption and savings. To know if SS and Medicare "crowd out" investment, you would have to know the relative mix of consumption and spending where no transfer payments would have been made and where the transfer payments have been made (the current circumstance). It is a very difficult problem to think through (because it involves a counterfactual) and has nothing to do with any of the things you are mention.

Yes, I fully understand that point you make, but it's still incorrect to make a blanket assumption that they are crowding out public investment. The latter is a function of income (e.g., tax revenues) and spending (e.g., choices such as more defense spending vs. government funded R&D of which there is some in the DOD budget and more than Alex presumes in his article). If we insist on being a low tax country, we need to recognize that this will also be a low spending country as well. Certainly we are seeing the "fruits" of this approach right now as many public universities are being decimated by decreasing state spending and the pipe dream that putting courses on the Internet is going to raise lots of money (witness what happened with the failed attempt to oust the UVa president).

Norman, I suggest you go to the free Finance class taught at Yale by Schiller (you can log on here: where he talks of SS and medicare so that you can understand the economics of SS, medicare and other social insurance better. If these are your religious beliefs, fine, I can't change them; but if you think saying SS isn't forced savings, if you think "money has long since been spent" when you are in a program that is continuously replenished with new workers, if you think there is no benefit from an annuity program with diverse members...then there is no hope, other than you learn a bit more.

I strongly agree - looking at aggregates makes little sense when we have a demographic bulge the size of the baby boom. Also, the article is not enlightening when it muddles together spending on infrastructure such as airports (would AT include roads?) with research aimed at "reduc[ing] cancer mortality by 10%." For the former, we MAY need public plus private investment to keep up with growth in the absolute size of the traveling public (and how do you measure that, really, in our Skyping age?), whereas for the latter, it is far from clear there should be much of a link between population size and public plus private research spending. And how is it meaningful to lump defense spending together with Social Security spending, and contrast those totals with publicly supported medical research? The case for (or against) generous public support of each of these areas could not be more different. There are various theories that would address them all (most informed strongly by public choice theory) much more persuasively than some sort of lack of "innovation nation" spirit! (BTW, The Emperor of All Maladies by S. Mukherjee is a very readable discussion of the challenges of cancer treatment progress, for the non-specialist.) I also question the statement that the health care reform was all mainly about redistribution - the insurance market breakdown should not be so casually overlooked by economists, say what you will about the need to logroll wrt many groups to get it done. And finally, why does the article set up its claim that we are not an "innovation nation" with an arresting presentation relating to federal government spending, only to point out that the extent to which a society throws federal money at all these various challenges may not be particularly relevant?

" Both have separate trust funds and certainly Social Security is pretty solvent for a number of years to come"

You know they don't and you know we know they dont, so why are you writing this?

If right-wingers are going to continue to deny "anthropomorphic climate change" despite hard data, the expertise of legions of scientists, and your own lying eyes, liberals will continue to deny your claims that SS is insolvent.

Besides which, SS won't be much of a problem once the Carousel program is instituted for baby boomers.

Republican Utopia: The culture of a economically successful society that has solved its retiree "problem":

"If right-wingers are going to continue to deny “anthropomorphic climate change” despite hard data, the expertise of legions of scientists, and your own lying eyes, liberals will continue to deny your claims that SS is insolvent."

Makes no sense on a couple levels, but that's fine. We'll just hold pat and see who blinks first.

Don't miss the point that you are currently joking about killing people because they are taking money away from the government. Those jokes will become slightly more serious. Then we'll propose quality of life year committees and have prominent liberal economists semi-jokingly explain why we really do need "death panels." I'm just kidding. We are already past that point.

Easy. Just make seniors wait 6 months to 3 years for 'elective' medical care. How do you think Canada spends half per capita on health care.

People in the US often have to wait months for elective medical care too. You don't just show up in the operating room and say "I'm ready, so slice and dice!" Your name goes on a schedule, often with triage applied depending on the severity of your need, and you wait until it's your turn. If there's a difference between the US and Canada in this respect it's a difference in degree, not in kind.

I know, "reality based community, but only when it's convenient"

But you're just justifying right-wing lying here.

Not really, I'm mostly just messing around. There's no excuse for either. Numbers don't lie. Liars lie.

Wow, that was a pretty amazing non sequitur.

Anyway, if the Trust Fund were invested in an income producing asset, like a national toll road system, the government could rely on it to help balance its books.

But it's not. Every single bond that you want to count as an asset for the government is simultaneously a liability for the government in an equal amount. The government cannot "cash in" a half-billion dollars in its trust fund instead of taxing/borrowing a half-billion, because in order to redeem those bonds the government would need to raise a half-billion first, by taxing/borrowing it.

The other reason Democrats should do it is because this is undermining respect for government and fueling the fundamental and perceptual deligitimization of government.

Do it for yourselves, don't do it for me.

Despite flat investment, capital stock per worker rises with aging, all else equal. Diminishing returns to investment imply a greater marginal rate of substitution towards consumption.

The main point not addressed here is that the growth in entitlement spending *is* growth in healthcare costs. Currently per unit private healthcare costs are higher than public healthcare costs, and that increasing the government share of healthcare costs will decrease the average unit costs of healthcare.

Yet corporations are sitting on piles of cash. Isn't there plenty of money for private investment at the moment?

This is the best rebuttal, even as I think that the thesis of crowding is worth considering.

Many american businesses have grown accustomed to growing without as much investing. They do it through wage-cutting, off-shoring, out-sourcing, LBO activity and so on - it is hard to invest when the finance sector will gut you on a quarterly basis if you miss their(!) targets. It has been this way for the last 30 years.

It's really a societal trend, and I have seen it first hand in my business.

I recall reading that after the 30's recession tax rates were set very, very high, partly to push idle capital into motion. In essence, the government adopted a use it or lose it approach to capital. I guess that would now be considered communism or something.

You're right bt, that would be communism, government control of the means of production.

We will force you to invest your money then tax the hell out of any returns.

Should work just fine.

No worries, we can just keep our profits offshore.

Paying taxes is for the little people.

Mother of God . . . another chart that lumps all "entitlement spending" together. The problem here, is Medicare. As Seth points out, Medicare spending is driving driving "entitlement spending." And Medicare spending is being driven by growing healthcare costs. Social Security, on the other hand, has grown some, and will grow a little more, but will flatten once the population growth rate flattens out. These two budgetary items need to be separated, not combined. Its kinda cute when Breitbart or Drudge do this, but I would think that Tyler and Arnold would at least be smarter and more honest than this.

Again...this may never stick, but why is something only a problem when it DESTROYS the government budget?

Why is not THE FIRST dollar of social security crowding out a dollar of other investment?

We'd have a better idea of this if we separated them out.

Separated what out?

You have a level of government spending available. The liberal side that "we need tax increases to provide consumers with the level of government services they desire" would make more sense if we were not ALREADY at the budgetary breaking point, which we are because this is the only reason we are having the discussion.. Liberals assume that we'll just increase the historical spending as a percentage of GDP.

To say that we need tax increases is just saying that we need to increase the size of government. But the government consumers are living a fantasy. We are already borrowing the dollars for the growth in entitlements. The growth in spending has never been paid for by sustainable revenues. A restoration of historical norms would be composed almost entirely of spending reductions.

"A restoration of historical norms would be composed almost entirely of spending reductions."

Tax revenue as a percentage of GDP is at a historical low. Spending is at historical high. Both need to be addressed.

Tax revenue as a percentage of GDP is at a historical low.

No it isn't. What kind of "historical low" are you speaking of? Federal tax revenue as a percentage of GDP was a fraction of what it is today pre second world war. Federal tax revenue was around 2-3% for most of the first two decades of the 20th century. If we're talking post-world war two than 1950 was a low point at 14.82% compared to 15.26% last year. If you are talking state and federal than it is higher today than the 50s and 60s. See:

@The Original D:
"Tax revenue as a percentage of GDP is at a historical low. "

If by historical low you mean well above historical averages.

What do Tyler and Arnold have to do with a chart by Jessica, Gabe, and Dave, mentioned in a post by Alex?

You're right, it was my mistake, shame on me. Not Tyler and Arnold, only Alex.

What is your point? Yes, a big part is Medicare. So? However you label it, the point is that investment spending is being driven out of the budget.

He's defining the problem as when the government budget becomes illiquid (it's already way insolvent if you consider off-budget liabilities).

That's not really what Alex is talking about. If you assume government has a fixed budget, give or take, then every dollar of one thing crowds out the other thing. That's a simplification, but if SS and the Meds were good investments we would be awash in ROI and wouldn't be having the discussion about what to get rid of in the national fire sale.

It would be interesting to know how many of the respondents on this thread are Medicare or SS eligible. I am so I guess I can speak from some experience on this. I fully understand the arguments that folks made to my original post above but the fact remains that there are some relatively straight forward things that can and should be done. Means testing for both programs can be increased. Now I'm in a position that I don't need to take SS until I turn 70 and that's the plan for both my wife and I. Medicare is already means tested and adjusted on a yearly basis based on AGI from one's most recent tax return. Part A is free but Part B costs money. We were fortunate to be "high earners" when we entered Medicare this year and our premium is $220/month per person for the remainder of this year (based on 2010 tax filings). Now this will go down next January when the 2011 tax return is considered (since I retired in 2010 our joint AGI has dropped). Now on top of this I'm still paying for my employer policy which becomes our Medigap policy. I'm old enough (obviously) to remember why LBJ pushed to get Medicare enacted back in the 1960s and believe that it was a good thing.

Regarding SS, there is no reason why current withholding should be capped so as to advantage high income individuals (and I benefited from this cap through a considerable portion of my working life. I don't agree with proposals to raise the retirement age. This may be fine for university professors, lawyers, and others who are in the privileged 5% of this country. It isn't fine for the person working in construction, a factory, a warehouse, or some lessor occupation. These individuals will have very little saved for retirement through either an employer plan or IRA and are much more reliant on SS payments than I am (or I venture to say most readers of this blog will be). Now we can say that this should not be the job of the government and that the system ought to be privatized. I don't think this is the correct approach.

Remember not everybody can be John Galt.

Means-testing social security will inevitably include reducing payments to those who have saved for their retirement.

The long-term effects of that strategy should be obvious.


So what do we do about these folks: We find that a substantial fraction of persons die with
virtually no financial assets—46.1 percent with less than $10,000—and many of these households
also have no housing wealth and rely almost entirely on Social Security benefits for support. In addition
this group is disproportionately in poor health =

working one or two more years is like magic to the retirement equation. If people in factories are living 10-15 years longer than they used to, surely they are more fit than previous generations to work an extra couple years..

any other solution is very ugly. Americans understand this and it's already happening.

Brian, What is this number that people are suddenly living 10-15 years longer???? SS changed in the mid-80's--what was the death rate for 65+ then, and what is it now? Is it 10-15 years?

If you look at the US life expectancy table, 1947 cohort (65 this year) life expectancy was 67; someone born today it's 78. Maybe that's the number that Brian is referring to. I'm not sure that there is a break down by occupation though we do know from OSHA statistics about work place injuries and hazards. I doubt that someone engaged in a repetitive stress jobs like a slaughterhouse is the same as someone working on a automobile assembly line. Of course it also is dependent on whether they are able to keep their job for their entire working life. I don't think that one can make this argument. I still believe that someone who is a university professor can work longer than a factory worker particularly if he/she gets tenure! :-)

But, in 1983, when SS was last adjusted, the life expectancy tables did not change by this amount. Should I not have believed the Maestro that SS would be sound forever with the 83 changes.

Bottom line: I don't by the longevity argument; SS has changed since '47, it was adjusted in '83, and I truly am interested in the facts that would show that since '83 longevity increased by 10 15 years.

I don't believe it unless I see the facts.

Just because it was changed in 1983 doesn't mean it changed enough. The retirement age is still 65 for practically everyone on Social Security. It only changes to 67 for people born after 1959. People born before 1955 get it full benefits at 66. This being 2012, those people won't qualify for 9 more years. It would be 8 years under the old plan.

I am referring to increases in life expectancy over the 77 years since Social Security was enacted.

The 1977/1983 amendments eliminated the need for ad-hoc wage and price adjustments- an important element of long-term stability.

Longevity was addressed for the first time, with an ad-hoc 2-year increase in the normal retirement age (phased-in).

The ad-hoc adjustment was far-sighted, inasmuch as common sense at the time seemed to indicate that trends toward earlier retirement would continue.

Since then, life expectancy has continued to increase, but no further adjustments have been made to retirement age. Without ongoing "indexation" in respect of increases in longevity, the system is not long-term viable.

The fact that the career arc in some lines of work may not be able to adjust to increasing longevity does not change any of these facts.

I think the examples of these types of jobs are way overdone, but, to the extent it's real, such as in the traditional trades, the fact is recognized and addressed explicitly by providing pensions as an element of compensation that are paid for during the career arc.

If working a couple more years is taken off the table, the problem looks intractable. If it's on the table, the problem, for the most part, dissolves into insignificance.

Cultural attitudes toward working an extra couple years are, in my opinion, an excellent signal of how well a society will manage this issue.

Documentation of my 'breezy' 10-15 year number. According to this, 15 looks closer to the mark:

More facts. The above numbers are "life expectancy at birth",so they reflect, e.g., improvements in child mortality that aren't relevant.

Here are some recent data showing a 0.9 year increase in "life expectancy at age 65" between 2000 and 2006. These trends have been remarkably durable for several decades. An epidemic or medical breakthrough could throw a monkey wrench into these trends, but it seems prudent to me to plan around a central case that assumes these trends continue.

Oops...forgot link:

"Remember not everybody can be John Galt."

And everybody can be a net cost even less so.

@Bill - I can't reply directly to you above, but you are correct that the 1983 birth cohort does not show that big a difference. It's only about 4 years. I think you and I are on the same wavelength here.

That's all well and fine, except that some on the left believe that the fruits of the Warfare State can also be defended as investment. "Integrated circuits, hybrid seed, internet… the list is endless."

Is it so hard to believe the great stagnation has made such investments less worthwhile? More necessary perhaps, and more required long term, but less effective in the near term in any case. You could probably do better plotting globalization against investment since increased globalization, while increasing foreign development, diminishes the advantage of domestic research, as the low lying fruit of labor arbitrage and lowering costs is so much more profitable than the hard work, risk, and expense of research that everyone leaves it to others to do, but that no one will do until those low risk foreign investments are no more worth making. The good news is this depression may be an indication we are approaching that point and that more domestic research will be both more necessary, likely, and profitable.

Note also the largest investment in research is carried out by the DOE, and who wants to eliminate that.

Does Tyler have a self-discrediting article category?

Any time I hear the words "entitlement reform" a little devil appears on my shoulder and says these people probably haven't broken out the entitlements into Medicare vs Social Security and haven't look at most of the problem being Medicare and then most of that due to health care spending.

And they do. They have a great line:

"—in ten years, Social Security alone will occupy almost one-fourth of our budget"

If you note their graph Social Security has occupied almost one-fourth of the budget for the past 10 years, too.

I think most pople who follow this at all get that Social Security, though big an absolute dollars, is manageable with minor adjustments, while Medicare, on its current course, is not close to sustainable.

The largest growing segment of social security spending is disability payments... which ... acts a lot like medicare.

Comments for this post are closed