Outside the Box


Outside the Box was retired on April 25, 2018, to make way for the new and improved premium research service, Over My Shoulder.

If you’re interested in joining John Mauldin, Patrick Watson, and the thousands of Over My Shoulder subscribers as they analyse important research several times a week, please click here to find out how you can subscribe for less than $10 per month.

Do Larger Federal Budget Deficits Stimulate Spending? Depends on Where the Funding Comes From

November 30, 2016

In the true spirit of stepping outside the box, today’s OTB is a counterintuitive argument against the concept that fiscal deficits and/or infrastructure spending constitute effective economic stimulus. It comes from Paul Kasriel, who was one of my favorite reads when he was at Northern Trust, and I am glad he continues to write in “retirement.” He always has a way of looking at things from different angles than everybody else does.

Paul is a self-confessed reformed Keynesian. He likens his own longtime tendency to revert to Keynesian macro analysis to the damnable difficulty of fixing a faulty golf grip: “If you start out playing golf with an incorrect grip, you will have a tendency to revert to it on the golf course even after hours of practicing at the driving range with a correct grip.” But as he struggled with his unfortunate tendency over the years, Paul was forever reminded of a question a fellow student asked him when Paul delivered his very first homily on the wonders of Keynesianism to an undergraduate political science class: “Where does the government get the funds to pay for the increased spending or tax cuts?”

As Paul himself notes, the post-election US stock market rally has been due in part to the expectation that the Trump administration will enact stimulative fiscal policies, which in turn will jumpstart growth. Paul begs to differ. He tells us that after some years out in the real world, he realized that tracing through where government gets the funds to finance tax cuts and increased spending is the most important issue in assessing the potential effects of stimulative fiscal policy. And, to cut to the chase, his conclusion was and remains that “Tax-rate cuts and increased government spending do not have a significant positive cyclical effect on economic growth and employment unless the government receives the funding for such out of “thin air.”

“Thin air.” You know, that stuff they bottle at the Federal Reserve, slap a fancy label on, and sell by the boatload. Or that emerges – POOF! – from banks as they create credit.

Paul engages us in a thought experiment to make his case, and I think I’ll just step aside and let him lead the thinking. He got me thinking, that’s for sure.

I click on Bloomberg (the website, not the terminal, which I don’t have) and see a sea of green. As I write, the stock market is once again at new highs. OPEC announces that it will somehow or other figure out how to cut the production of their oil, which will raise the price, which will immediately mean that more wells are drilled all over the world, especially in the US, and so production will rise sooner rather than later back to the level where it is now – and, globally, we are already outproducing demand. Interest rates are rising and bonds are falling off. I keep hearing talk about inflation coming back. The euro is close to a seasonal low and the yen is (finally) once again rising. The pound is down some 40% from its peak. US GDP was revised up to over 3% last quarter, making for growth of roughly 2% year-over-year. Gold is down over 10% from its peak just prior to the election. The Fed’s Beige Book just came out, and it was generally positive.

This just isn’t making for a gloom and doom mindset, but gloom and doom is pretty much what you hear if you pay attention to the mainstream media. Evidently, US markets are not as concerned about the political scene as the news media are. Or maybe they have different priorities.

Sidebar: I wonder how many individuals and businesses are going to try to rearrange their end-of-year income and push as much as they can into 2017 to take advantage of what everyone believes will be lower tax rates. It is going to be interesting to see what corporate earnings look like for the fourth quarter versus the first quarter. And especially what pro forma income looks like compared to tax form income. Let the games begin.

You have a great week.  I’m already working on this weekend’s letter, as I want to write about Italy and I really want to get it to you before the actual election there on Sunday. So until this weekend,

Your the more things change analyst,

John Mauldin, Editor
Outside the Box

Get John Mauldin's Over My Shoulder

"Must See" Research Directly from John Mauldin to You

Be the best-informed person in the room
with your very own risk-free trial of Over My Shoulder.
Join John Mauldin's private readers’ circle, today.

Do Larger Federal Budget Deficits Stimulate Spending? Depends on Where the Funding Comes From

By Paul Kasriel
November 21, 2016

The U.S. equity markets have rallied in the wake of Donald Trump’s presidential election victory. Various explanations have been given for the stock market rally. President-elect Trump’s pledge to scale back business regulations are favorable for various industries, especially financial services and pharmaceuticals. Likewise, President-elect Trump’s vow to increase military spending is an undeniable plus for defense contractors. But another explanation given for the post-election stock market rally is that U.S. economic growth will be stimulated by the almost certain business and personal tax-rate cuts that will occur in the next year, along with the somewhat less certain increase in infrastructure spending. It is this conventional –wisdom notion that tax-rate cuts and/or increased federal government spending stimulate domestic spending on goods and services that I want to discuss in this commentary.

Although I have been a recovering Keynesian for decades, I got hooked on the Keynesian proposition that tax-rate cuts and increased government spending could stimulate domestic spending after having taken my first macroeconomics course way back in 19 and 65. I was so intoxicated with Keynesianism that I made a presentation about it in a political science class. I dazzled my classmates with explanations of the marginal propensity to consume and Keynesian multipliers. My conclusion was that economies need not endure recessions if only policymakers would pursue Keynesian prescriptions with regard to tax rates and government spending. Reading the body language of my classmates, I believed that I had just enlisted a new cadre of Keynesians. That is, until one older student sitting in the back of the class raised his hand and asked the simple question: Where does the government get the funds to pay for the increased spending or tax cuts? I had to call on all of my obfuscational talents to keep my classmates and me in the Keynesian camp.

When I graduated from college with a degree in economics, I still was a Keynesian, perhaps a bit more sophisticated one, but not much. At graduate school, I became less enchanted with Keynesianism. But Keynesianism is similar to an incorrect golf grip. If you start out playing golf with an incorrect grip, you will have a tendency to revert to it on the golf course even after hours of practicing at the driving range with a correct grip. Bad habits die hard. So, even though I had drifted away from Keynesianism, it was easy and “comfortable” to slip back into a Keynesian framework when performing macroeconomic analysis. Yet, I continued to be haunted by that question my fellow student asked me: Where does the government get the funds to pay for the increased spending or tax cuts?

I guess I am a slow learner, but after a number of years in the “real world”, away from the pressure of academic group-think, I realized that tracing through the implications of where the government gets the funds to finance tax-rate cuts and increased spending is the most important issue in assessing the stimulative effect of changes in fiscal policy. And my conclusion is that tax-rate cuts and increased government spending do not have a significant positive cyclical effect on economic growth and employment unless the government receives the funding for such out of “thin air”.

Let’s engage in some thought experiments, beginning with a net increase in federal government spending, say on infrastructure projects. Let’s assume that these projects are funded by an increase in government bonds purchased by households. Let’s further assume that the households increase their saving in order to purchase these new government bonds. When households save more, they cut back on their current spending on goods and services, transferring this spending power to another entity, in this case the federal government. So, the federal government increases its spending on infrastructure, resulting in increased hiring, equipment purchases and profits in the infrastructure sector of the economy. But with households cutting back on their current spending on goods and services, that is, increasing their saving, spending and hiring in the non-infrastructure sectors of the economy decline. There is no net increase in spending on domestically-produced goods and services in the economy as a result of the bond-financed increase in infrastructure spending. Rather, there is only a redistribution in total spending toward the infrastructure sector and away from other sectors.

What if a pension fund purchases the new bonds issued to finance the increase in government infrastructure spending? Where does the pension fund get the money to purchase the new bonds? One way might be from increased pension contributions. But an increase in pension contributions implies an increase in saving by the pension beneficiary. The pension fund is just an intermediary between the borrower, the government, and the ultimate saver, households or businesses saving for the benefit of households. Again, there is no net increase in spending on domestically-produced goods and services in the economy.

What if households or pension funds sell other assets to nonbank entities to fund their purchases of new government bonds? Ultimately, some nonbank entity needs to increase its saving to purchase the assets sold by households and pension funds. Again, there is no net increase in spending on domestically- produced goods and services in the economy.

What if foreign entities purchase the new government bonds? Where do these foreign entities get the U.S. dollars to pay for the new U.S. government bonds? By running a larger trade surplus with the U.S. That is, foreign entities export more to the U.S. and/or import less from the U.S., thereby acquiring more U.S. dollars with which to purchase the new U.S. government bonds. Hiring and profits increase in the U.S. infrastructure sector, decrease in the U.S. export or import-competing sectors.

Now, let’s assume that the new government bonds issued to fund new government infrastructure spending are purchased by the depository institution system (commercial banks, S&Ls and credit unions) and the Federal Reserve. In this case, the funds to purchase the new government bonds are created, figuratively, out of “thin air”. This implies that no other entity need cut back on its current spending on goods and services while the government increases it spending in the infrastructure sector. All else the same, if an increase in government infrastructure spending is funded by a net increase in thin-air credit, then there will be a net increase in spending on domestically-produced goods and services and a net increase in domestic employment. We cannot conclude that an increase in government infrastructure spending funded from sources other than thin-air credit will unambiguously result in a net increase in spending on domestically-produced goods and services and a net increase in employment.

President-elect Trump’s economic advisers have suggested that an increase in infrastructure spending could be funded largely by private entities through some kind of public-private plan. This still would not result in net increase in U.S. spending on domestically-produced goods and services and net increase in employment unless there were a net increase in thin-air credit. The private entities providing the bulk of financing of the increased infrastructure spending would have to get the funds either from some entities increasing their saving, that is, by cutting back on their current spending, or by selling other existing assets from their portfolios. As explained above, under these circumstances, there would be no net increase in spending on domestically-produced goods and services.

Now, it is conceivable that an increase in infrastructure spending, while not resulting in an immediate net increase in spending on domestically-produced goods and services, could result in the economy’s future potential rate of growth in the production of goods and services. To the degree that increased infrastructure increases the productivity of labor, for example, speeds up the delivery of goods and services, then that increase in infrastructure spending could allow for faster growth in the future production of goods and services.

Another key element in President-elect Trump’s proposed policies to raise U.S GDP growth is to cut tax rates on households and businesses. To the degree that tax-rate cuts result in a redistribution of a given amount of spending away from pure consumption to the accumulation of physical capital (machinery, et. al.), human capital (education) or an increased supply of labor, tax-rate cuts might result in an increase in the future potential rate of growth in GDP, but not the immediate rate of growth unless the tax-rate cuts are financed by a net increase in thin-air credit.

At least starting with the federal personal income tax-rate cut of 1964, all personal income tax-rate cuts have been followed with cumulative net widenings in the federal budget deficits. So, for the sake of argument, let’s assume that the likely forthcoming personal and business tax-rate cuts result in a wider federal budget deficit. Suppose that households in the aggregate use their extra after-tax income to purchase the new bonds the federal government sells to finance the larger budget deficits resulting from the tax-rate cuts. The upshot is that there is no net increase in spending on domestically-produced goods and services nor is there any net increase in employment emanating from the tax-rate cuts.

My conclusion from the thought experiments discussed above is that increases in federal government spending and/or cuts in tax rates have no meaningful positive cyclical effect on GDP growth unless the resulting wider budget deficits are financed by a net increase in thin-air credit, that is a net increase in the sum of credit created by the depository institution system and credit created by the Fed.

Let’s look at some actual data relating changes in the federal deficit/surplus to growth in nominal GDP. I have calculated the annual calendar- year federal deficits/surpluses, and then calculated these deficits/surpluses as a percent of annual average nominal GDP. The red bars in Chart 1 are the year-to- year percentage-point changes in the annual budget deficits/surpluses as a percent of annual-average nominal GDP. The blue line in Chart 1 is the year-to-year percent change in average annual nominal GDP. According to mainstream Keynesian theory, a widening in the budget deficit relative to GDP is a “stimulative” fiscal policy and should be associated with faster nominal GDP growth. A widening in the budget deficit relative to GDP would be represented by the red bars in Chart 1 decreasing in magnitude, that is, becoming less positive or more negative in value. According to mainstream Keynesian theory, this should be associated with faster nominal GDP growth, that is, with the blue line in Chart 1 moving up. Thus, according to mainstream Keynesian theory, there should be a negative correlation between changes in the relative budget deficit/surplus and growth in nominal GDP. The annual data points in Chart 1 start in 1982 and conclude in 2007. This time span includes the Reagan administration’s “stimulative” fiscal policies of tax-rate cuts and faster-growth federal spending, the George H. W. Bush and Clinton administrations’ “restrictive” fiscal policies of tax-rate increases and slower-growth federal spending and the George H. Bush administration’s “stimulative” fiscal policies of tax-rate cuts and faster- growth federal spending.

In the top left-hand corner of Chart 1 is a little box with “r=0.36” within it. This is the correlation coefficient between changes in fiscal policy and growth in nominal GDP. If the two series are perfectly correlated, the absolute value of the correlation coefficient, “r”, would be equal to 1.00. Both series would move in perfect tandem. As mentioned above, according to mainstream Keynesian theory, there should be a negative correlation between changes in fiscal policy and growth in nominal GDP. That is, as the red bars decrease in magnitude, the blue line should rise in value. But the sign of the correlation coefficient in Chart 1 is, in fact, positive, not negative as Keynesians hypothesize. Look, for example, at 1984, when nominal GDP growth (the blue line) spiked up, but fiscal policy got “tighter”, that is the relative budget deficit in 1984 got smaller compared to 1983. During the Clinton administration, budget deficits relative to nominal GDP shrank every calendar year from 1993 through 1997, turning into progressively higher surpluses relative to nominal GDP starting in calendar year 1998 through 2000. Yet from 1993 through 2000, year-to-year growth in nominal GDP was relatively steady holding in a range of 4.9% to 6.5%. Turning to the George H. Bush administration years, there was a sharp “easing” in fiscal policy in calendar year 2002, with little response in nominal GDP growth. As fiscal policy “tightened” in subsequent years, nominal GDP growth picked up – exactly opposite from what mainstream Keynesian theory would predict.

Of course, there are macroeconomic policies that might be changing and having an effect on the cyclical behavior of the economy other than fiscal policy. The most important of these other macroeconomic policies is monetary policy, specifically the behavior of thin-air credit. In Chart 2, I have added an additional series to those in Chart 1 – the year-to-year growth in the annual average sum of depository institution credit and the monetary base (reserves at the Fed plus currency in circulation). “Kasrielian” theory hypothesizes that there should be a positive correlation between changes in thin-air credit and changes in nominal GDP. With three variables in chart, Haver Analytics will not calculate the cross correlations among all the variables. But E-Views will. And the correlation between annual growth in thin-air credit and nominal GDP from 1982 through 2007 is a positive 0.53. Not only is this correlation coefficient 1-1/2 times larger than that between changes in fiscal policy and nominal GDP growth, more importantly, this correlation has the theoretically correct sign in front of it. By adding growth in thin-air credit to the chart, we can see that the strength in nominal GDP growth in President Reagan’s first term was more likely due to the Fed, knowingly or unknowingly, allowing thin-air credit to grow rapidly. Similarly, the reason nominal GDP growth recovered from the George H. W. Bush presidential years and was relatively steady was not because tax rates were increased in 1993 and federal spending growth slowed, but rather because growth in thin-air credit recovered in 1994 and held relatively steady through 1999.

In sum, there may be rational reasons why the U.S. equity markets rallied in the wake of Donald Trump’s presidential election victory. But an expectation of faster U.S. economic growth due to a more “stimulative” fiscal policy is not one of them unless the larger budget deficits are financed with thin-air credit. Fed Chairwoman Yellen, whether you know it or not, you are in the driver’s (hot?) seat.

Paul L. Kasriel - 
Founder, Econtrarian, LLC - 
Senior Economic and Investment Advisor Legacy Private Trust Co. of Neenah, WI
The Econtrarian: 

“For most of human history, it has made good adaptive sense to be fearful and emphasize the negative; any mistake could be fatal”, Joost Swarte

Discuss This


We welcome your comments. Please comply with our Community Rules.



Dec. 1, 2016, 10:31 a.m.

Mr. Kasriel is raising an important question which needs further examination.  It is heartening to note that he distinguishes between prompt effects of Q.E. type fiscal stimulus and longer term effects on economic growth of directed stimulus of wealth-creating/cost-reducing capital goods and infrastructure.  I concur that government ‘stimulus’ spending funded by taxation or bond issuance results in no net economic activity at first glance.  From a national perspective however, it may be that the stimulus expenditures consist of a greater proportion of domestic goods and services than would be the case if the taxed or saved money had been spent on a more typical consumer consumption profile with a larger import or foreign travel component. That is, infrastructure spending funded by taxation or government bond issuance (to domestic buyers/savers) will likely reduce the ‘negative’ components in the factors making up GDP, resulting in a higher GDP.
While I agree that use ‘created’ money to fund fiscal stimulus will generate additional economic activity, I believe there is an important caveat.  The stimulus comes at the cost of higher prices for goods and services than would otherwise prevail.  All other sectors of the economy bear this obscured cost.  There simply is no ‘free lunch’.
Fiscal stimulus distorts price signals in the economy and results in malinvestment.
I also concur with commenter S. Walton’s point about budget deficits.  The debt of the federal government according to the Debt to the Penny page of the Treasury Department’s website shows that there has not been an actual surplus in decades.  One cannot ignore expenditures by declaring them ‘off budget’.  They show up on the balance sheet as increased debt.


Dec. 1, 2016, 9:37 a.m.

Mr. Mauldin, thanks for bringing such inciteful material to your readers.
I think this analysis could benefit from another lens being brought to bear.

One is the lens of time: When the Fed buys “from thin air”, it is borrowing from the future. When foreign investors buy bonds they are spending trade surplus from the past. In the latter case, I would suggest there is a net increase in GDP.

Another useful lens is velocity of money. When the working class investor buys bonds, they are reducing consumption, and the velocity of money. When an investor is simply transferring their investment vehicle to bonds, they aren’t changing their consumption, but the bonds lead to an increased velocity of money (by paying workers).

The article brings up another point about increasing productivity. I believe that infrastructure investment is literally about"keeping the lights on”, not productivity. It reduces the probability of adverse events like a bridge collapse, brown-out, or water supply interruption. For a humorous coverage of this I suggest John Oliver’s episode “Infrastructure”.

BTW, I am always glad to hear the story of “recovering Keynesians” like myself. I stopped drinking the Kool-aid” when I realized that politicians would never actually raise taxes and cut spending at the top of the business cycle.
Thanks again,
—Toby Sarver

Anthony Mann

Dec. 1, 2016, 8:44 a.m.

Duh!  Tanstaafl! My Keynesian education is 10 years younger than the author’s, but I do recall that Keynes was clear that he expected governments to maintain a low debt burden and to use additional tax revenues during boom years to pay off debt and to prepare a slush fund to allow spending during a bust. Basically a countercyclical counterweight.
Just because our political masters of both parties want to ‘have our cake and eat it too’ by running debt to the max doesn’t make it a good idea and doesn’t make it a Keynesian concept!

John Harris

Dec. 1, 2016, 6:28 a.m.

Please consider the possibility that even if tax rates go down, both individual and corporate tax income can go up when individuals roll over capital gains and corporatioms bring “home” profits stored off shore. These can be very large numbers, but I don’t know how to estimate them from available data. Please try.



Dec. 1, 2016, 12:03 a.m.

National Production Monthly Stimulus Checks
Don Soards

Jobs give workers paychecks.  We use those paychecks to buy goods and services.  Automation makes those products cheaper.  Our money buys more, and our standard of living goes up.  Automation is constantly increasing, so our standard of living should keep going up, right?  Not necessarily.

What would happen if all the jobs were automated?  Then we wouldn’t have money to buy products, causing automated retailers to call automated suppliers and tell them to stop production because there are no customers.  When production stops, our standard of living drops.

This balancing act between the production of humans and machines versus human paychecks is the “Production Parabola.”  Note that production drops when there are fewer than 50% employed.  This drop is caused by a lack of demand in which too few people make enough money to buy all the suppliers’ goods and services.  The buyer’s-market side has too few dollars chasing too many goods.

The upward side (seller’s market) of the parabola is characterized by a focus on increasing production so we can have enough goods and services to satisfy our needs.  The seller’s-market side can be generally thought of as “too many dollars chasing too few goods.”  On the seller’s-market side, increases in automation increase our national production and much of that increase is reflected in our standard of living.

Which side of the production parabola are we on now?  The June 2016 labor pool adjusted for recently unemployed, low-wage part-timers, and discouraged workers stands at 56.6%.  From a standpoint of national economic decision making, we and other industrialized civilizations (Europe, Japan, Russia, etc.) entered the buyer’s market many years ago.  Here are some qualitative differences.

Seller’s-market side               Buyer’s-market side
Positive interest rates             Negative interest rates
Businesses grow sales               Businesses cut costs
Full time jobs                   Part-time jobs, discouraged workers
Good job benefits                 Fewer benefits
Widespread hiring                 Widespread layoffs
Want cable TV                     Millions have “Cut the Cord”
Inflation                       Deflation
Individual families having their own home Multi-generational families in one                         home
Rising salaries                   Jobless recovery
Less crime                       More crime
Robust middle class                 Shrinking middle class
Debts paid off                   Huge credit card debt

One characteristic of a buyer’s market is that seller’s-market strategies don’t work very well.  Attempts to stimulate job growth by lowering interest to low or even negative rates have not worked because no one wants to loan to new businesses that lack customers.  Additionally, potential customers are not keen to get a low interest loan if they fear that they can’t pay it back.  If we were in the seller’s market, lowering interest rates down to today’s rock-bottom values would cause extreme inflation because we would be encouraging money to enter a business environment that already had too many dollars chasing too few goods.
Strategies designed to increase business and job expansion like lowering taxes, free college, eliminating regulations, etc. will all have muted results because they do not address the fundamental buyer’s-market problem that consumers are underfunded. 
We need to face the fact that we are now on the buyer’s-market side and our biggest problem is funding the consumer.  We can take heart.  The buyer’s-market side offers us one tool that the seller’s-market side doesn’t.  If we simply printed money and passed it out to American adults while we were on the seller’s-market side, we would cause unacceptable inflation because we would now have way too many dollars chasing too few goods.  However, on the buyers-market side, we can print money and pass it out to American adults because we have too few dollars chasing too many goods.  We can print an amount such that “too few” becomes “just right.”
How much monthly money should we send to American adults to fund our economy?  One criterion would be to replace the money removed from our middle class economy by capital-funded automation.  This is about $6000 per year per adult American citizen.  This would result in a monthly stimulus of $500 per U.S adult citizen.
I recommend that we start with a stimulus of $100 (pre-tax) per month and observe the effect on the economy.  If we see a little inflation, we can hold at the $100 level.  If we don’t see inflation, then we can raise the monthly stimulus to $200 per month.  We keep raising until we see enough inflation that small interest rate raises are required.  At this point, we stop the stimulus increases.
We need to allow the Fed to print money (off the balance sheet) and distribute it to U.S. adult consumers.  This will help all family budgets and create some new jobs.
See https://sites.google.com/site/monthlystimuluschecks/  for frequently asked questions.

Charles DuBois

Nov. 30, 2016, 11 p.m.

This article is so poor I don’t know where to start.  Misleading info about mechanics of spending and mechanics of thin air money.  Just for starters, OF COURSE, higher deficits tend to be ocrrelated with economic weakness.  A weaker economy causes lower revenues and higher spending.  So, for example,  deficits jump up during recessions.  But the recession caused the higher deficit - not the other way around. He needs to correlate deficits with growth, holding other things equal. Then he will get an entirely different answer.


Nov. 30, 2016, 6:49 p.m.

I get angry when I here this outrageous lie about budget surpluses during the 1990s. Here is a simple rule - if any entity, including the US Government, has more debt at the end of the year than the beginning, it certainly didn’t earn a surplus. Here is the 12 years from 1990 through 2002.
FY Ending, Billions
9/30/2002, 6,228 Bn
9/30/2001, 5,807 Bn
9/30/2000, 5,674 Bn
9/30/1999, 5,656 Bn
9/30/1998, 5,526 Bn
9/30/1997, 5,413 Bn
9/30/1996, 5,225 Bn
9/29/1995, 4,974 Bn
9/30/1994, 4,693 Bn
9/30/1993, 4,411 Bn
9/30/1992, 4,065 Bn
9/30/1991, 3,665 Bn
9/28/1990, 3,233 Bn.
I don’t see any surplus here, do you?