The Savings from Lower Administrative Costs in a Medicare-for-All System

 

A.  Introduction

One of the most important issues facing the US is our high cost of health care.  We have a terribly inefficient system, with the highest costs in the world (reaching 18% of GDP, which is 50% more than in the second most expensive country and close to double the average of the OECD countries), yet with only mediocre results compared to other countries.  It is a market-based system, with competing health care providers (doctors, hospitals, and so on) and competing private health insurance companies.  However, the extremely wide variation in prices for the same treatments and procedures (often varying by a factor of ten or more) is a clear sign that this market is not working as it should.  And those skilled at exploiting these inefficiencies are able to profit handsomely, with CEOs and other senior staff of the major private insurance companies paid well.  Indeed, total compensation packages have occasionally even topped $100 million.

Despite so much spending, the US is still far from providing affordable access to health care for our entire population.  While the situation improved substantially following the introduction of Obamacare (with the share of the US population without any form of health insurance falling by about 40% after Obamacare went into effect), the Trump administration is doing all it can to reverse these gains.

Faced with these issues, a number of analysts and politicians (Senator Bernie Sanders as just the most prominent) have proposed that the US move to what is termed a “single-payer” system, such as what they have in Canada, France, and a number of other countries.  In a single-payer system, doctors, hospitals, and healthcare service providers remain as they are now, as independent and typically private agents serving their patients.  The only difference is that there is only one insurer, run as a government agency.  This is what the US has in the popular Medicare system, but Medicare is restricted only to those aged 65 and above.  Hence in the US context, a single-payer system for all is often referred to as “Medicare-for-All”.

A key question is whether a Medicare-for-All system would reduce the high cost of healthcare in the US.  Those opposed to any such government managed programs have argued that costs would rise.  And they have issued reports with headline findings that can only be interpreted as being deliberately misleading.  For example, in late July, Charles Blahous (a former Bush administration official) issued an analysis through the Mercatus Center of George Mason University (a center that has received major funding from the Koch Brothers) that concluded government spending would rise by $32.6 trillion over ten years under a Medicare-for-All system.  This has received a good deal of press coverage, and is being used (as I write this) in a number of ads being televised by Republican candidates in the 2018 midterm elections.

But while worded carefully, this claim is misleading in the extreme.  First of all, that such high amounts will be spent on health care should not be a surprise, when added up over ten years.  Total US health care spending is expected to reach $3.7 trillion this year, would rise to $5.7 trillion by 2026 if nothing is done, and would total $45.0 trillion over the ten-year period of 2017 to 2026 (using National Health Expenditure data and forecasts, which will be discussed in detail below).  The portion of this covered by various forms of personal health insurance (both private and public, such as Medicare, but excluding the military and the VA) is expected to reach $2.7 trillion this year, $4.2 trillion by 2026, and would sum to $33.1 trillion over the ten years 2017 to 2026.

So high amounts will be spent on health care, unless measures are taken to improve efficiency and reduce costs.  In per capita terms, the US population will be spending in 2018 an average of $8,190 per person through the various forms of personal health insurance our system currently employs.  This is, without question, a lot.  It will be an estimated 17.9% of the median wage this year.  But if we had the far lower administrative costs that Medicare has been able to achieve for the health insurance it manages directly, instead of the significantly higher administrative costs incurred under a variety of mostly private health insurance plans (discussed below), the average per capita cost would be just $7,480 per person in 2018.  There would also be other savings (such as what health care providers will enjoy from a simplified system, which we will also discuss below), but the savings from those sources, while certainly significant, are harder to estimate.  The $7,480 figure simply reflects savings from lower administrative costs on the part of the insurers if we were able to achieve what Medicare already does.

Thus the correct question is whether we should prefer sending a check for $8,190 per person to Aetna, Cigna, United Healthcare, and the other insurers (and including what is paid through taxes for Medicare and other publicly managed insurance), or a check for $7,480 just to Medicare under a Medicare-for-All system.  The doctors we see would be the same, and the treatments and procedures would also be the same as what we have now.  The savings here is purely from more efficient administration of our health insurance.  That the check in one case goes just to the government, and in the other to a mix of private and public insurers, should not be, in itself, of consequence.  But the Blahous argument, in saying that we cannot afford the $32.6 trillion he forecasts for healthcare spending over ten years, is that for some reason a larger check (of $8,190) to our current mix of insurers is fine while we cannot afford to send instead a smaller check (of $7,480) if that check goes to a government entity.  This is silly.

For the nation as a whole, the savings from the greater efficiency of a Medicare-for-All system is substantial.  As we will see, it would add up to $204 billion in 2016, had this system been in place that year, growing to $365 billion by 2026.  For the ten year period from 2017 to 2026, the savings would sum to $2.9 trillion.  This is not a small sum.

This main point is that we should look at the data, and not presume certain outcomes based on ideology or political beliefs.  We will thus start in this blog post with an examination of what administrative costs actually are, for Medicare and for private insurance.  We will see that the cost for administering Medicare, for the portion of Medicare managed directly by government, is far less than what is spent to administer other health insurance, including in particular private health insurance.  There are many reasons for this, where the most important is the relative simplicity and scale of the Medicare system.  An annex to this blog will discuss in detail what these various factors are for the different health insurance systems that could be folded into a Medicare-for-All system.  We will also discuss in that annex why Medicare is able to achieve its far lower administrative costs, and address some of the arguments that have mistakenly asserted that this is not the case, despite the evidence.

Taking the administrative costs that Medicare has been able to achieve as a base, we will then calculate what the savings would add up to, per year for the US as a whole, under a Medicare-for-All system.  The basic result is depicted in the chart at the top of this post, and as noted above, the savings from greater administrative efficiency would rise from $204 billion in 2016 (had the system been in place then) to $365 billion in 2026.

These savings are substantial.  But there are also other savings, which are, however, more difficult to estimate.  The penultimate section of this post will discuss several.  They include savings that will be possible in the administrative and clerical costs at doctor’s offices and at hospitals and other healthcare facilities.  Doctors, hospitals, and other facilities must hire specialist staff to deal with the complex and fragmented system of insurance in the US, and the costs from this are substantial.  There will also be savings on the part of employers, who must now manage and oversee the contracts they have with private insurers.

A final, concluding, section will summarize the key issues and discuss briefly why such an obvious and large saving in costs has not been politically possible in the US (at least so far).  The short answer:  Vested interests profit substantially under the current fragmented system, and it should not be a surprise that they do not want to see it replaced.  With extra spending in the hundreds of billions of dollars each year, there is a lot to be gained by those skilled at operating in this fragmented system.

B.  The Cost of Administering Current Health Insurance Plans

It is often difficult to estimate what costs and savings might be under some major reform, as we do not yet know what will happen.  But this is not the case for estimating administrative costs for health insurance.  We already have excellent data on what those costs actually are for a variety of different health insurance providers, including Medicare.

The primary sources of the data are the National Health Expenditure Accounts (NHE), produced annually by the Centers for Medicare and Medicaid Services, and the Annual Report of the Medicare Trustees.  The current NHE (released in February 2018) provides detailed historical figures on health expenditures (broken down in numerous ways) through to 2016, plus forecasts for many of the series to 2026.  And the Annual Report of the Medicare Trustees (with the most recent released in June 2018), provides detailed financial accounts, including of government administrative costs, for the different components of Medicare and the supporting trust funds (with past as well as forecast expenditures and revenues).

Table 19 of the historical tables in the most recent NHE provides a detailed break down of health care expenditures in 2016 by payer (mostly various insurance programs, both public and private).  The expenditures shown include what is spent on administration by government entities (separately for state and federal, although I have aggregated the two in the table below), and for what they term the “net cost of health insurance”.  The net cost of private health insurance includes all elements of the difference between what the private insurer receives in premium payments, and what the insurer pays out for health services provided by doctors, hospitals, and so on.  Thus it includes such items as profits earned by the insurer.  For simplicity, I will use “administrative costs” to include all these elements, including profits, even though this is a broad use of the term.

Table 19 of the NHE shows Medicare expenditures for all components of Medicare on just one line.  While it shows separately the administrative costs incurred by government in the administration of Medicare (with all of it federal, as states are not involved), and the administrative costs (as defined above) incurred by private insurers for the Medicare programs that they manage, the NHE does not show separately which of those costs (government and private) are linked to which Medicare programs.

For those figures one must turn to the Medicare Trustees Annual Report.  Medicare Parts A and B are managed directly by Medicare officials, and provide payments for services by hospitals (Part A) and doctors (Part B).  Medicare Part C (also now called Medicare Advantage) is managed by private insurers on behalf of Medicare, and cover services that would otherwise be covered by Medicare directly in Medicare Parts A and B.  And the relatively recent Medicare Part D (for prescription drugs) is also managed by private insurers, either as a stand-alone cover or folded into Medicare Advantage plans.

Any such combination of numbers from two separate sources will often lead to somewhat different estimates for those figures that can be compared directly with each other.  There might be slight differences in definitions, or in concepts such as whether expenses are recorded as incurred or as paid, or something else.  But the figures which could be compared here were close.  In particular, the figure for total Medicare expenditures in calendar year 2016 was $678.8 billion in the Trustees report and $672.1 billion in the NHE, a difference of just 1%.  Of greater relative importance, the Trustees report has a figure for government administration (for all Medicare programs combined) of $9.3 billion, while the NHE has a figure of $10.5 billion.  However, while the difference between these two figures may appear to be large, what matters is not so much the difference between these two, but rather the difference (as a share of total costs) between either of these and the much higher cost share for privately managed insurance (as we will see below).  We will in any case run scenarios in Section C below with each of the two different estimates for government administrative costs in Medicare, and see that the overall effect of choosing one rather than the other is not large.

Based on these sources, the costs paid in 2016 under most of the major health insurance programs in the US were:

Current Expenditures for Health Care and for Administrative Costs 

   2016 data ($ billions)

Gross Cost

Gov’t Admin

Private Admin

Total Admin

Total   as %

Private Health Insurance

$1,123.4

$129.6

$129.6

11.5%

Medicare:

$678.8

    Gov’t Administered

$390.7

$9.3

$9.3

2.4%

    Privately Administered

$288.1

$36.3

$36.3

12.6%

Medicaid

$565.6

$24.2

$36.1

$60.3

10.7%

CHIP (Children’s Health Insurance Program)

$16.9

$1.5

$1.4

$2.9

17.3%

Worker’s Compensation

$50.7

$2.3

$16.4

$18.8

37.0%

Total: 

$2,435.3

$37.4

$219.8

$257.2

10.6%

* Medicare Gov’t Admin –   NHE estimate

$390.7

$10.5

$10.5

2.7%

Sources:  Medicare expenditures, other than private administrative costs, are from the 2018 Medicare Trustees Annual Report.  All other figures are from the NHE accounts, Table 19 (historical), released in February 2018.

 

The table leaves out the health care programs of the Department of Defense and the Veterans’ Administration (as they operate under special conditions, with many of the services provided directly), as well as a number of smaller government and other programs (such as for Native Americans, or worksite or school-based health programs).  Those programs have been set aside here due to their special nature.  But while significant, the $2,435.3 billion of expenditures in the programs listed in the table account for 89% of the total spent in the US in 2016 on all health care services to individuals covered through either some form of health insurance or third-party payer.  While some portion of the remaining 11% could perhaps be folded into a Medicare-for-All system (thus leading to even higher savings), we will focus in this post on the 89%.

The table shows that the administrative cost ratios vary over a wide range, from just 2.4% for the health insurance Medicare administers directly (using the Medicare Trustees figures, or 2.7% based on the NHE figures), up to 37% for the administration of the health portions of Workers’ Compensation.  The administrative cost for direct private health insurance is 11.5% on average, while the administrative cost for the privately managed portions of Medicare (Medicare Part C and Part D) is a similar, but somewhat higher, 12.6%.

This wide variation in administrative cost ratios provides clues on what is going on.  These will be discussed in the Annex to this post for those interested.  Briefly, the programs (other than government-administered Medicare) are complex, fragmented, have to make case by case assessments of whether the claim is eligible (as for Workers’ Compensation plans) or whether the individual meets the enrollment requirements (as for Medicaid and CHIP – the Children’s Health Insurance Program), and do not benefit from the scale economies that Medicare enjoys.

But while such explanations are of interest in understanding why Medicare can be provided at such a lower cost than private and other insurance, the key finding, in the end, is that it is.  The data are clear.  The next section will use this to calculate what overall savings would be at the national level if we were to move to a system with the cost efficiencies of Medicare.

C.  National Savings in Administrative Costs from a Medicare-for-All System

Medicare (for the portion managed directly by government) costs far less to administer than our current health insurance system with its complex and fragmented mix of plans (most of which are privately managed).  Only 2.4% of the cost of the portion of Medicare managed directly by government was needed for administration of the program in 2016, while the costs to administer the other identified health insurance programs range between 10.7% (for Medicaid) and 11.5% (for private health insurance) to 37% (for workers’ compensation plans).  With $2.4 trillion spent on these health insurance plans (in 2016), the savings from a more efficient approach to administration will be significant.

An estimate of what the nation-wide savings would be can then be calculated based on figures in the NHE forecasts of health expenditures (by health insurance program) for the 2017 to 2026 period (Table 17 of the forecasts), coupled with the Medicare system forecasts provided in the Medicare Trustees Annual Report.  Applying the share of administrative costs in the portion of Medicare managed directly by government (2.4% in 2016, but then using the year by year forecasts of the Medicare trustees for the full forecast period), rather than what the administrative cost ratios would have been for the other programs that would be folded into a Medicare-for-All system (private health insurance, Medicaid, CHIP, and Workers’ Compensation), using their 2016 cost ratios, yields the savings shown in the chart above.

Had a Medicare-for-All system been in effect in 2016, we would have saved $204 billion in administration, with this growing over time (with the overall growth in health expenditures over time) to an estimated $365 billion by 2026.  The savings over ten years (2017 to 2026) would be $2.9 trillion, and would by itself bring down the cost of health care (for the programs covered) from a ten year total of $33.1 trillion forecast now, to $30.2 trillion with the reform.  There would be other savings as well (discussed in the next section below), but they are more difficult to quantify.  However, a very rough estimate is that they could be double the magnitude of the savings from the more efficient administration of health insurance alone.  See the next section below for a discussion.

The calculations here required a mix of data from the NHE and from the Medicare Trustees report, and as I noted above, the estimates of the cost of government administration in these two sources were not quite the same.  The Medicare Trustees report gave a figure for government administrative costs of the overall Medicare system of $9.3 billion in 2016 (and then year by year forecasts going forward to 2026), while the NHE estimate was $10.5 billion in 2016.  As shown in the last line of the table above, the $10.5 billion figure would lead to an administrative cost share of 2.7%, compared to the 2.4% figure if the cost was at the NHE figure of $10.5 billion.  The savings in moving to a Medicare-for-All system would then not be as large.

But the impact of this would be small.  One can calculate what the cost savings would be assuming government administration would cost 2.7% rather than the 2.4% figure in the Medicare Trustees report (with also its forecasts going forward), using the same process as above.  The total national savings would have been $199 billion in 2016 rather than $204 billion, growing to savings of $345 billion in 2026 rather than $365 billion.  The ten-year total savings would be $2.7 trillion rather than $2.9 trillion.  The savings under either estimate would be large.

D.  Other Efficiency Savings in a Medicare-for-All System 

The $2.9 trillion (or $2.7 trillion) figure for savings over ten years from moving to a Medicare-for-All system comes solely from the lower administrative costs that we know can be achieved in a Medicare type system – we know because we know what Medicare in fact costs.  But there are other savings as well that will be gained by moving to this simpler system, and this section will discuss several of them.  How much would be saved is more difficult to estimate, so we have kept these savings separate.  But some rough figures are possible.

But before going to these other sources of efficiency gains, we should mention one possible source of lower costs which has often been discussed by others, but which I would not include here.  It has often been asserted that Medicare pays doctors, hospitals, and other health service providers, less than what other insurance plans pay.  But first, it is not clear whether this is in fact true.  It might be, but I have not seen reliable data to back it up.  The problem is that most of what is paid to doctors, hospitals, and others by private health insurance plans is now at network negotiated rates, and these rates are kept as trade secrets.  It is not in the interest of the doctors and other health care providers to reveal them (as it would undermine their bargaining power with other insurers), nor in the interest of the insurance companies to reveal them (as other insurers would gain a competitive advantage in their negotiations with the providers).  Indeed, secrecy clauses are common in the negotiated agreements.

In the absence of such publicly available data, one is limited to citing either anecdotal cases, or statements by various health care providers who have a vested interest in trying to persuade Medicare to pay them more.  Neither will be reliable.

But second, and aside from this difficulty in knowing what the truth really is, the focus in this blog post is solely on the gains that could be achieved by moving to a more efficient system.  If doctors and hospitals are indeed paid less under Medicare, costs would go down, but this would be in the nature of what economists call a transfer payment, not an efficiency gain.  Efficiency gains come from being able to do more with less (e.g. administer more at a lower cost).  Transfers are a payment from one party to another, with no net gain – the gain to one party is offset by a loss of the same amount to the other.

Excluding such transfers (if they in fact exist), what are other efficiency gains that one would obtain with a Medicare-for-All system (other than the gains from lower administrative costs for the health insurance itself, which we estimated above)?  There are several:

a)  Doctors offices now need to employ specialists in handling billing, who are able to handle the numerous (and often changing) health insurance plans their patients are enrolled in.  These specialists are critical, and good ones are paid well, as they are needed if the doctors want to be paid in full for the services they provided.  Based on personal experience, I am often amazed that the staff good at this are indeed able to stay on top of the numerous health insurance plans they must deal with (I find it difficult enough to stay on top of just my own).  While essential to ensuring the doctors can survive financially, such staff are a significant cost.  While one will still need to ensure proper billing under any Medicare-for-All system, it would be far simpler.

b)  Similarly, hospitals and other medical facilities need to employ such specialist staff to handle billing.  The same issues arise.  They must contend with numerous health insurance plans, each with its own set of requirements, and ensure the bills they file with the insurers will compensate the facilities properly (and from their perspective most advantageously) for the services provided.  This is not easy to do under the present highly complex system, and would be far simpler under Medicare-for-All.

c)  There are also costs that must be borne by employers in managing the primarily employer-based health insurance system used in the US.  The employer must work out which health insurance provider would work best for them, negotiate a complex but critical and expensive contract, and then oversee the insurer to ensure they are providing services in accordance with that contract.  Firms must often hire specialist (and expensive) consultants to advise them on how best to do this.  With the cost of healthcare so high in the US, these health insurance contracts are costly.  It is important to get them right.  But all this necessary oversight is also a major cost for the firm.

How much might then be saved from such sources by moving to a more efficient Medicare-for-All system?  This is not so easy to estimate, but one study looked at the costs in the US from such expenses and compared them to similarly measured expenses in Canada, which has a single-payer system.  As noted above, a Medicare-for-All system is a single-payer system, and thus (along with the other similarities between the US and Canadian economies, such as the similar levels of income) the difference between what the costs are in the US and the costs in Canada for the same services can provide an estimate of how much might be saved by moving to a single-payer, Medicare-for-All system.

The study was prepared by Steffie Woolhander (lead author – Harvard Medical School), along with Terry Campbell, and David Himmelstein, and was published in the New England Journal of Medicine, August 2003.  They drew from a variety of sources to arrive at their estimates, and some had to be approximate.  The data is also from 1999 – almost 20 years ago.  Things may have changed, but with the upward trend in costs over time in the US, the cost shares now are likely even worse.  The authors presented the basic figures in per capita terms (and all in US dollars), and I have scaled them up to what they would be in 2016 (assuming the shares are unchanged) in accordance with the overall growth in US personal health care spending (from the NHE accounts).

The results are:

Admin costs 1999/2016

Per capita in $

Per capita in $

Per capita     in $

Total in $ billion

US –    1999

Canada – 1999

US excess – 1999

US excess – 2016

Insurance overheads

$259

$47

$212

$156.9

Doctors, hospitals, other

$743

$252

$491

$363.3

    Doctors only

$324

$107

$217

$160.6

    Hospitals & other facilities

$419

$145

$274

$202.8

Employers’ admin costs

$57

$8

$49

$36.3

Total:

$1,059

$307

$752

$556.5

Total excluding Insurance overheads

$399.6

Source:  Calculated from Woolhander, Campbell, and Himmelstein, “Costs of Health Care Administration in the United States and Canada”, New England Journal of Medicine, 349: 768-775, August 21, 2003.

Note:  “Insurance overheads” exclude health insurer profits as well as certain expenses (such as for advertising and marketing).

 

The first three columns show the estimated spending in per capita terms (and in US dollars) for each category of costs, for the US, for Canada, and then for the difference between the two.  US spending is always higher.  Thus, for example, for the line labeled “doctors”, the authors estimate that doctor’s offices have to spend an average of $324 per every US resident for expenses related to billing and other dealings with health insurance companies in 1999.  The cost in Canada with its single-payer system, in contrast, is on average just $107 per resident (in US dollar terms).  The difference is $217 per person, in 1999.  Grossing this up to the US population, and rescaled to total health care expenditures in the US in 2016 relative to 1999, the excess cost in the US in 2016 is an estimated $160.6 billion.  This is what would be saved in the US in 2016 if doctor’s offices were able to manage their health insurance billings with the same efficiency as they can in Canada.

The other lines show the estimated savings from other sources.  The top line is for insurance overheads.  The estimate here is that the US would have been able to save $156.9 billion in 2016 if health insurance administration were as efficient as what is found in Canada with its single-payer system.  While on the surface this appears to be less than the $204 billion savings estimated (for 2016) if the US moved to a Medicare-for-All system, they are in fact consistent.  The estimate in Woolhander, et. al., of the excessive cost of health insurance administration excludes what is paid out in insurance company profits and certain other expenses (such as advertising and marketing).  As discussed in the Annex below, insurance company profits can add one-third to administrative costs, so a $150 billion cost would become $200 billion when one uses the same definitions for what is encompassed.  The two estimates are in fact surprisingly consistent, even though very different approaches were used for the estimation of each.

Overall, the US would have saved about $400 billion (excluding the savings from lower expenses at the insurance companies) had a single-payer system been in effect in 2016, according to these estimates.  That is double the estimated $204 billion in savings from lower administration costs at the health insurers alone, estimated in the section above.  These additional cost savings from moving to a Medicare-for-All system are clearly significant, but are often ignored in the debate on how much would be saved from efficiency gains in a Medicare-for-All system.  They are (I would acknowledge) rough estimates.  They cannot be estimated with the same precision as one can for the savings from the more efficient administration of health insurance alone under a Medicare-for-All system.  But neither should they be forgotten.

E.  Summary and Conclusion

Medicare is a well-managed and popular program.  It is a single-payer system, but currently restricted to those aged 65 and above.  And administrative costs, on that portion of Medicare managed directly by government, are only 2.4%.  This 2.4% is far below the 11.5% administrative cost share for regular private health insurance, or 12.6% for that portion of Medicare that is managed through private health insurance companies.

And even with such low costs, Medicare is a popular program, where numerous surveys have found Medicare to be more highly rated (including in terms of user experiences with the program) than private health insurance plans (see, for example, here, here, here, and here).

Creating a Medicare-like system to cover also those Americans below the age of 65 would not be difficult.  We already have the model of Medicare itself to see what could be done and how such a system can be managed.  And we also have the examples of other countries, such as Canada, that show that such systems are not only feasible but can work well.  It is also not, as conservative critics often assert, a government “takeover” of healthcare (a criticism also often used in attacks on Obamacare):  Under a single-payer system, the providers of health care services (doctors, hospitals, and so on) remain as they are now, as private or non-profit entities, competing with each other in the services they offer.

Nor would an extension of health insurance under a Medicare-like system to those below age 65 lead to issues for the current Medicare system.  This has now become an attack line being asserted in numerous Republican political campaigns this fall, including in a signed piece by President Trump published on October 10 by USA Today.  This was in essence a campaign ad (but published for free), which fact checkers immediately saw contained numerous false statements.  As Glenn Kessler noted in the Washington Post, “almost every sentence contained a misleading statement or a falsehood”.

There is no reason why extending a Medicare-like system to those below age 65 should somehow harm Medicare.  The cost for the health insurance for those below age 65 would be paid for by sending the checks we currently must send to private insurers (such as Aetna or United Healthcare), instead to the new single-payer insurer.  As noted above, with such an entity copying the Medicare management system and achieving its low administrative costs, we would have been able to reduce the average per person cost of healthcare in 2018 from the $8,190 we are paying now, to $7,480 instead, a savings of $710 for each of us.  That $7,480 would still need to be paid in, but it is far better to send in $7,480 to the single-payer (for the same health care services as we now receive) than to send in $8,190 to the mix of insurers we now have.

Furthermore, these savings are solely from the more efficient administration of health insurance that we see can be done in Medicare.  There will also be very substantial savings from other sources in a Medicare-for-All system, including in what doctors and hospitals must now spend to deal with our currently highly fragmented and complex health insurance system, and savings by employers in what they must spend to manage their employer-based private health insurance plans.  The magnitude of such additional savings are more difficult to estimate, but they might be on the order of double the size of savings from the more efficient administration of the health insurance itself.  That is, total savings in 2016 might have been on the order of $600 billion, or three times the $200 billion in savings from more efficient administration of health insurance alone.

And such savings (or rather the lack of it under our current complex and fragmented system) can account for a significant share of the far higher cost of health care in the US than elsewhere.  As noted before, health care costs about 18% of GDP in the US, or 50% more than in the second most expensive country where it is just 12%.  Had the US been able to save $600 billion in health care expenditures in 2016 by moving to a Medicare-for-All system, US healthcare spending would have been reduced from 18% of GDP to below 15% (more precisely, from 17.9% in 2016 to 14.7%).  This, by itself, would have gotten us over halfway to what other countries spend.  More should be done, to be sure, but such a reform would be a major step.

So why has it not been done?  While the lower costs under a Medicare-for-All system would be attractive to most of us, one needs also to recognize that those higher costs are a windfall to those who are skilled at operating within our complex and fragmented system.  That is, there are vested interests who benefit under the current system, and the dollar amounts involved are massive.  Private health insurers, and their key staff (CEOs and others), profit handsomely under this system, and it should not be surprising that they lobby aggressively to keep it.  Under a Medicare-for-All system, there would be no need (or a greatly reduced need, if some niches remain) for such private health insurance.

This is not to deny that there will be issues in any such transition.  Just the paperwork involved to ensure everyone is enrolled properly will be a massive undertaking (although for all those currently enrolled in some health insurance plan, mostly via employer-based plans, the paperwork could presumably be transferred automatically to the new program).  Nor can one guarantee that while on average health care consumers will save, that each and every one will.  But the same is true in any tax reform, where even if taxes on average are being cut, there are some who end up paying more.

One should also acknowledge that many doctors and hospitals are concerned that in a Medicare-for-All system they will have little choice but to agree to the Medicare-approved rates for their services.  However, it is not clear this is much different from the current system for the doctors, where they must either agree to accept the in-network rates negotiated with the private health insurers, or expect few patients.  And surveys of doctors on their support for a Medicare-for-All system show a turnaround from earlier opposition to strong support.  A survey published in August 2017 found 56% of physicians in support (and 41% opposed), a flip from the results of a similar survey in 2008 (when only 42% were in support, and 58% opposed).  A key reason appears to be the costs and difficulties (discussed above) doctors face in dealing with the multiple, fragmented, insurance plans they must contend with now.  Even the American Medical Association, a staunch opponent of Medicare when it was approved in the 1960s, and an opponent ever since, may now be changing its views.

Finally, 70% of Americans now support a Medicare-for-All system, according to a recent Reuters survey.  It is time for such a system.

 

 


Annex:  The Causes of the Wide Variation in Administrative Cost Shares

a.  The Wide Range of Administrative Cost Shares

Administrative cost shares vary enormously across different health insurance programs, from just 2.4% for government-managed Medicare to 37% for health insurance provided through Workers’ Compensation plans.  The figures are shown above in the top table in the post.  Some might say that this cannot be – that they are all providing health insurance so why should the differ by so much.  But they can and they do, and this annex will discuss why.

Take the case of Workers’ Compensation first.  Workers’ Compensation insurance was established by states in the US starting in 1902 (Maryland was the first).  Most states passed laws between 1910 and 1920 requiring businesses to arrange for such insurance, and by 1920 all but five states (all in the South) had such coverage (and by 1948 all states had it).  And in most (but not all) states, health care benefits are provided through the purchase of privately managed insurance.

But these programs are expensive to administer.  Each individual claim must be scrutinized to determine that it was in fact due to a covered workplace injury.  This leads to the extremely high (37%) administrative cost share.  If the injury is indeed covered, the workers’ compensation insurance arranged by the business will pay for the associated health care costs.  But if it is not, the injury will now normally be covered by the individual’s regular health care insurance.  The treatment is still needed, and is provided.  The issue is only who pays for it.

Hence the time and effort spent to ascertain whether the injury was in fact due to a covered workplace injury is a pure social cost, and would not be needed (at least for the health care treatments) in a Medicare-for-All system.  The injuries would still be treated, but funds would not need to be spent to see whether the costs can be shifted from one insurer to a different one.  And when each individual claim must be assessed (with many then rejected), the administrative costs for Workers’ Compensation plans can be a high share of what is in the end paid for healthcare treatments.

When workers’ compensation programs were first set up, in the early 20th century, individual health insurance was not common.  Such health insurance (set up through employers) only began to be widespread during World War II, when the Roosevelt administration approved favorable tax treatment of such insurance by businesses (who were trying to attract workers, but were subject to general wage and price controls).  But workers’ compensation programs continue to exist, despite their high administrative costs.  And from the point of view of the private insurer providing the workers’ compensation cover, spending such money to assess liability for some injury makes sense, as (from the private perspective of the individual insurer) they would gain if the health treatment costs can be shifted to a different insurer.  But such expenditures do not make sense from the perspective of society as a whole.  They are just a cost.  And under a Medicare-for-All system the injury would simply be treated, with no need to ascertain if one insurer or a different one was responsible for making the payment.  Overall costs would be less, with the same health care treatments provided.

There are similar socially wasteful expenditures in other health insurance programs, which drive up their administrative costs.  CHIP (Children’s Health Insurance Program) has a relatively high administrative cost share (17.3% in 2016) in part because it is relatively small ($16.9 billion in expenditures in 2016, which can be compared to the $678.8 billion for Medicare), so it does not enjoy the economies of scale of other programs, but also because eligibility for the program must be assessed for each individual participating.  While rules vary by state, children and teens are generally eligible for CHIP coverage up to age 18, for families whose incomes are below some limit, but who are not receiving Medicaid (or in coordination with Medicaid in certain cases).  The CHIP insurance for the children and teens is then either free or low-cost, depending on family income.

Confirming that children to be enrolled under CHIP meet the eligibility requirements is costly.  Hence it is not surprising that this (along with the lack of the economies of scale that larger programs can take advantage of) leads to the relatively high share for administrative costs.  But this eligibility question would not be an issue that would need to be individually assessed in a Medicare-for-All system.  It is a socially wasteful expenditure that is required only because the program needs to confirm those enrolled meet the specific eligibility requirements of this narrow program.  And a Medicare-for-All system would of course enjoy huge economy of scale advantages.

Medicaid also has to bear the cost of assessing whether eligibility requirements have been met, and certain states are indeed now making those eligibility requirements even more burdensome and complex (in the apparent hope of reducing enrollment).  Most recently, the Trump administration in early 2018 issued new rules allowing states to impose work requirements on those enrolled in Medicaid, and several states have now started to impose such restrictions.  But such requirements are themselves costly to assess.  While enrollment in Medicaid may then fall (leading to the health care costs of those individuals being shifted on to someone else), administrative costs as a share of what is spent will rise.  But from the point of view of society as a whole, shifting the cost of health treatment for those individuals who would otherwise be enrolled in Medicare on to someone else does not save on the overall cost of health care.  And indeed, if it shifts such treatment from doctor’s offices to treatment in emergency rooms, the cost will go up, and probably by a lot.

This would no longer be an issue in a Medicare-for-All system.  There would be no need to waste funds on assessing whether the individual meets the eligibility requirements of some specific health insurance program or another.

Despite such special costs. the overall costs of administration for Medicaid were 10.7% in 2016.  This is a bit below the cost for regular private insurance of 11.5%, and probably reflects the significant economies of scale Medicaid is able to benefit from.  And while a significant share of the Medicaid administrative costs are incurred by private insurers contracted to manage the Medicaid programs in many of the states ($36 billion of the $60 billion total for administration according to the NHE figures), government itself takes on a significant share of the administration.  And the overall administrative cost combined is still less than what private health insurance requires (as a share).

b)  The Cost of Administering Private Health Insurance

Which brings us to the question of why private health insurance costs so much to administer, at 11.5% of the total paid for such insurance.  Medicare, when administered directly by government, has a cost of just 2.4%.  Why does private insurance cost so much more?

First, a note on terminology.  Up to this point, as we have discussed various government health insurance plans (such as Medicaid or CHIP), we have not had to distinguish the total cost of the health insurance plans (the total of what is paid out in benefits to health care providers, plus what is paid for administration) from the total paid for the insurance cover.  We need to be more precise for private insurance cover.  One has the total paid in any period (a year in these figures) in insurance premia by the subscribers, and the total in what is paid by the private insurer in each such period to cover benefits.  The NHE has estimates for each of these, and then calculates the difference between the two as the “net cost of health insurance”.  We have referred to this as a broad concept of administrative costs, as it includes any profits earned by the insurers as part of their current operations.  But private insurers have an additional source of earnings, and that is from revenues on invested capital.  Premia are paid upfront and benefits paid out later (in overall probabilistic terms), and an important source of income to insurers comes from what they earn on those funds as they are invested in various asset markets, such as stocks and bonds, real estate, commodities, and so on.

For private insurance we should therefore be clear that what we have so far referred to as the “total cost” of the health insurance is synonymous with the total premia paid (which some sources refer to as “underwriting revenue”).  Subtracting the total paid to health care providers under the insurance policies from the total paid in premia will then lead to the broad concept of administrative costs, including profits earned from the current period insurance operations.  On top of this, private insurers will generate earnings from investments on their accumulated capital (obtained, in part, from premia being paid in before benefits are paid out).  For the figures here we are excluding these latter earnings.  Such earnings will be on top of those obtained from their current insurance operations.

Why then, do private insurers incur administrative costs (as defined here) of 11.5% when government-administered Medicare has a far lower cost of just 2.4%?

There are a number of reasons.  First, private health insurance is a tremendously fragmented system, where health plans are mostly organized at the individual firm level.  This is costly, and the cost share varies systematically by firm size.  Administrative costs (including insurer profits) will typically range between 5 and 15% of the total paid for the insurance in firms with greater than 50 employees, between 15 and 25% in firms with fewer than 50 employees, and (in the period before the Obamacare market exchanges were set up) between 25 and 40% of the total for individuals seeking health insurance (see, for example, this report from the Commonwealth Fund).

These high and rising costs (in inverse direction to firm size) arise as there are significant fixed costs in setting up any such system at some firm, which leads to a high cost-share when there are fewer workers to spread it over.  Commissions paid to insurance brokers also play a role, as the use of brokers is typical and especially significant for the small-group market.  The Commonwealth Fund report cites figures indicating these commissions can account for 4 to 11% of the total in premia paid for insurance in such markets.  And in those cases where the insurers themselves take on the risk (as opposed to simply managing the claims process while the firm itself pays the claims – this is called “self-insurance”, and is typical in large firms with 1,000 employees or more, as it ends up cheaper for such firms), the insurers must then invest significant resources in assessing the risk of the pool of workers covered in order to price the policy appropriately.  The costs the insurance company will need to pay out will depend not only on the local cost of health care services (which can vary tremendously across different parts of the country), but also by the industry of the firm (as the health risks of the typical workers employed will vary by industry) and specifics of the firm being covered (such as the average age of the workers employed, the male/female ratio, and other such factors).

There are also high fixed costs of the insurers themselves under their business model.  They typically offer dozens of insurance plans, each with different features on what is covered and by how much.  And most of the plans are built around networks of care providers (doctors, hospitals, and so on) with whom they have individually negotiated “in-network” prices for subscribers of the particular health insurance plan.  These in-network prices can still vary tremendously (even by a factor of ten or more, for those I have been able to check with my own insurer, and all for the same metropolitan area), and are set through some negotiation process.  The price eventually agreed to reflects some balance in negotiating strength.  If you are a hospital chain that dominates in some metro area, you will be able to negotiate a price close to what you wish to charge as the insurer has to include hospital services.  Similarly, if you are an insurance company that dominates in some metro area, then the hospitals have to agree to charge something close to what you are willing to pay, as otherwise they will not have many patients.  And individual doctors operating in private practices will generally have very little negotiating power.

But such negotiations (for each and every health care provider, and then for each possible service) are expensive to carry out, regardless of the outcome.  And while some argue that such negotiations hold down the cost of health care, it is not at all clear that such is the case.  The US, after all and as noted before, has by far the most expensive health care services in the world (close to double the average in OECD countries, as a share of GDP), and yet achieves only mediocre results.  Furthermore, the actual volume of health care services provided in the US (as measured, for example, by doctor consultations per capita per year, or hospital beds per 1,000 of population, and so on) has the US at close to the bottom among OECD countries.  The problem is not excessive health care services utilized, but rather their high cost in the US.  Negotiated in-network pricing has not helped, and quite possibly (due to the resulting fragmentation into non-competing markets) has hurt.

This complex and fragmented system does lead, however, to high rewards to those who are good at operating in it.  Hence CEOs (and other senior staff) of insurance companies skilled at this are rewarded handsomely, with such CEOs typically receiving compensation of more than $10 million a year, and in some cases far more.  Indeed, as recounted in an earlier blog post, the CEO of UnitedHealth Care personally received total compensation of more than $1.3 billion over his 15-year tenure of 1991 to 2006 (even after the SEC forced him to forfeit stock options worth a further $620 million due to illegalities in how they were priced).  Such salaries are reflected in the administrative costs of the health insurance plans offered, and account for a substantial share of it.

Finally, this complex and fragmented market has also led to high profits for the private health insurance companies.  If this were due to the exceptional efficiency of certain of the health insurance firms as compared to others, all in a competitive market, then such high profits of such firms might be explained.  But there is no indication that health insurance markets operate anywhere close to what economists would call “perfect competition”.  The extremely wide variation in prices for the same health care services (often by a factor of ten or more) is a clear sign of markets that are nowhere close to perfectly competitive.

And the amount paid to cover such profits is high.  For example, an examination of health insurance markets in New York State found (in data for 2006) that profits from underwriting (i.e. excluding profits from capital invested) accounted for 4.9% of total underwriting revenue (the total premia paid) before taxes, or roughly one-third of the total 14.9% in administrative costs (including underwriting profits).  After taxes, it would be roughly one-quarter of the total.  Applying that ratio to the 11.5% administrative cost share found in the NHE accounts for the nation as a whole in 2016, the charge to cover profits would be close to 3% points.  That, by itself, would be greater than the 2.4% cost share for government-managed Medicare.

c)  The Cost of Administering the Portion of Medicare Managed Directly by Government

Why, then, does the portion of Medicare (Parts A and B) managed directly by government cost so little?  It is fundamentally because Medicare does not bear many of the costs discussed above for the other insurance plans, and can spread the costs that remain over a far larger enrollment base.  Specifically:

1)  Medicare enjoys huge economy of scale advantages:  The portion of Medicare managed directly by government is huge, at $390.7 billion spent in 2016 ($381.4 billion of which went to health care providers, and only $9.3 billion to administration).  And this is for a single plan.  Private health insurers instead each manage dozens of plans covering millions of firms (at rates which vary firm to firm, depending on the risk pool).

2)  Medicare does not have to make a determination for each individual claim as to whether it will be covered (as Workers’ Compensation plans must), nor whether the individual is eligible (other than whether they are of age 65 or more, and have paid the relevant premia and taxes).  That is, Medicare does not need to contend with the complex (and now being made increasingly complex) eligibility requirements for participants in Medicaid, CHIP, and other such programs.

3)  Medicare has one set of compensation rates, which doctors and hospitals accept or not.  The compensation rates vary by region and other such factors, but they are not individually negotiated each year with each of the possible providers.

4)  And Medicare does not have the costs private insurers need to pay to retain the CEOs and other senior staff who are skilled at operating within the fragmented US healthcare system, nor do they pay large amounts for marketing and such.  Nor does Medicare pay profits, and profits, as noted above, are high for private health insurers in the US.

It is this “business model” of Medicare which keeps its costs down.  It is a relatively simple model (relative to that of private insurers – no health care payment system is simple in an absolute sense), and enjoys great economies of scale.  Thus Medicare can keep its costs down, and needs to spend on administration only a fraction of what private health insurers spend.

d)  The Conservative Critics of Medicare Costs

There are critics who contend that Medicare costs are not in fact low.  These critics have issued analyses through such groups as the Heritage Foundation (conservative, with major funding from the Koch brothers), the Cato Institute (conservative – libertarian), lobby groups with a vested interest, and publications that link back to these analyses.  But these arguments are flawed.  Indeed, some of the responses to the assertions are so obvious that one must assume that ideology (a view that it is impossible for government to be more efficient) was the primary driver.

These critics make three primary arguments:

1)  First, several contend that Medicare does not pay for, nor include in its recorded administrative costs, the costs incurred by Social Security and other government agencies that provide services that are essential to Medicare’s operations.  For example, initial enrollment in Medicare at age 65 is handled through the Social Security Administration, and Medicare premia payments (for Parts B and D) are normally collected out of Social Security checks.

However, while it is true that Social Security provides such services to Medicare, it is not true that Medicare does not pay for this.  A simple look at the Medicare income and expenses tables in the Medicare Trustees Annual Report will show what those payments are.  For example, for fiscal year 2017, Tables V.H1 and V.H.2 (on pages 217 and 218 of the 2018 report) indicate that $980,805,000 was paid to the Social Security Administration under the Medicare HI Trust Fund (“Hospital Insurance”, for Part A) and $1,247,226,000 under the Medicare SMI Trust Fund (“Supplementary Medical Insurance”, for Parts B and D).  These are substantial amounts, and they are not hidden.

And the tables similarly show the amounts paid by Medicare (as components in its administrative costs) to other government agencies for services they provide to Medicare.  These include payments made to the FBI and the Department of Justice (for fraud and abuse control), to the Office of the Secretary of Health and Human Services (HHS, for oversight) as well as to other HHS offices (such as the Inspector General), to the US Treasury, and to a number of others.  They are all shown.  The conservative critics who assert Medicare expenses do not include payments for such services simply never looked.

2)  Second, the critics argue that while private insurers must raise the capital they need to fund their operations, and that that capital has a cost, the costs of funding Medicare’s “capital” are not counted but rather are hidden away in the overall government budget.

But this reflects a fundamental confusion on the capital requirements of established insurers, whether private or public.  Insurers are not banks.  Banks raise funds (at a cost) and then lends them out.  Insurers take in premia payments from those insured, and at some later time make payments out under the insurance policies for covered costs.  On average, the payments they make come later than the payments they receive in premia, and hence they have capital to invest.  That capital is invested in stocks and bonds, real estate, commodities, or whatever, they make a return on those investments, and that return is factored into, and can reduce (not raise), the premia they need to charge to cover their overall costs.

Private insurers hence generate earnings from their capital, as it is invested as an asset.  It is not a cost.  Furthermore, Medicare operates in fundamentally the same way as other insurers.  The Medicare Trust Funds (HI and SMI) reflect funds that have been paid in and not yet expended in covered claims or other expenses, and they earn interest on the balances in those trust funds (at the long-term US Treasury bond rate).  The accounting is all there to be seen, for those interested, in the Medicare Trustees Annual Report.

3)  Probably most importantly, the conservative critics of Medicare assert that it is incorrect to calculate administrative costs as a share of the total costs paid.  Rather, they say those costs should be calculated per person enrolled.  Since older people have far higher medical costs each year than younger people do (which is certainly true), they argue that the low administrative cost share seen in Medicare (when taken as a share of total costs) is actually a reflection of the high health care costs of the elderly.

But there are two problems with this.  First, when elderly people see doctors at a pace of say 10 times a year rather than perhaps once a year when younger, they will be generating 10 times as many bills that need to be recorded and properly paid.  Each bill must go through the system, checked for possible fraud, and then paid in the correct amount.  That will cost more, indeed one should expect it will cost 10 times as much.  And if anything, medical procedures are more complicated for the elderly (as they have more complicated medical conditions), so it should be expected that the costs to process the more complex bills will indeed go up more than in proportion to the amount spent.  The conservative critics assert the costs of administering this do not go up with the more frequent billing, but rather are the same, flat, rate per person regardless of how many, how complex, and how costly the medical interventions are that they have in any given year.

Second, one has data.  The Medicare Parts C (Medicare Advantage) and D (for drugs) are managed via private health insurers.  And this Medicare is for the same elderly population that government-managed Medicare covers.  If what the conservative critics assert is correct, then the cost of administering these privately-managed Medicare programs should be similar to the cost of administering the portion of Medicare that government manages directly.  But this is not the case.  Government-managed Medicare spent only 2.4% on administration in 2016, while privately-managed Medicare spent 12.6%.  These are far from the same.

Indeed, the 12.6% administrative cost share for the privately-managed portion of Medicare is similar to, but a bit more than, the 11.5% share seen with regular private health insurance.  This is what one would expect, where the somewhat higher cost share might well be because of the greater complexity of the medical interventions required for the elderly population.

The government-managed portion of Medicare has a far low administrative cost share than private health insurance.  The conservative critics have not looked at the data.

Taxes on Corporate Profits Have Collapsed

A.  Introduction:  The Plunge in Corporate Profit Tax Revenues

Corporate profit tax revenues have collapsed following the passage by Congress last December of the Trump-endorsed Republican tax plan.  And this is not because corporate profits have decreased:  They have kept going up.  The initial figures, for the first half of 2018, show federal corporate profit taxes (also referred to as corporate income taxes) collected have fallen to an annual rate of roughly just $150 billion.  This is only half, or less, of the $300 to $350 billion collected (at annual rates) over the past several years.  See the chart above.

The estimates on corporate profit taxes actually being paid through the first half of 2018 come from the National Income and Product Accounts (NIPA, and commonly also referred to as the GDP accounts) produced by the Bureau of Economic Analysis.  The figures are collected as part of the process of producing the GDP accounts, but for various reasons the figures on corporate profit taxes are not released with the initial GDP estimates (which come out at the end of the month that follows the end of each quarter), but rather one month later (i.e. on August 29 this time, for the estimates for the April to June quarter).  The quarterly estimates are seasonally adjusted (which is important, as tax payments have a strong seasonality to them), and are then shown at annual rates.  While we already saw such a collapse in corporate tax revenues in the figures for the first quarter of 2018 (first published in May), it is always best with the estimates of GDP and its components to wait until a second quarter’s figures are available to see whether any change is confirmed.  And it was.

This initial data on what is actually now being collected in taxes following the passage of the Republican tax plan last December suggests that the revenue losses will be substantially higher than the $1.5 trillion over ten years that the staff at the Joint Committee on Taxation (the official arbiters for Congress on such matters) forecast.  Indeed, the plunge in corporate profit tax collections alone looks likely to well exceed this.  On top of this, there were also sharp cuts in non-corporate business taxes and in income taxes for those in higher income groups.

This blog post will look at what the initial figures are revealing on the tax revenues being collected, as estimated in the GDP accounts.  The focus will be on corporate income taxes, although in looking at the total tax revenue losses we will also look briefly at what the initial data is indicating on reductions in individual income taxes being paid.

The chart above shows what the reduction has been in corporate profit taxes in dollar terms.  In the next section below we will look at this in terms of the taxes as a share of corporate profits.  That implicit average actual tax rate is more meaningful for comparisons over time, and it has also plunged.  And the implicit actual rate now being paid, of only about 7% for the taxes at the federal government level, shows how misleading it is to focus on the headline rate of tax on corporate profits of 21% (down from 35% before the new law).  The actual rate being paid is only one-third of this, as a consequence of the numerous loopholes built into the law.  The Republican proponents of the bill had argued that while they were cutting the headline rate from 35% to 21%, they were also (they asserted) ending many of the loopholes which allowed corporations to pay less.  But in fact numerous loopholes were added or expanded.

The next section of the post will then look at this in the longer term context, with figures on the implicit corporate profit tax rate going back to 1950.  The implicit rate has fallen steadily over time, from a rate that reached over 50% in the early 1950s, to just 7% now.  While Trump and his Republican colleagues argued the cut in corporate taxes was necessary in order for the economy to grow, the economy in fact grew at a faster pace in the 1950s and 1960s, when the rate paid varied between 30 and 50%, than it has in recent decades despite the now far lower rates corporations face.

But this is for the federal tax on corporate profits alone.  There are also taxes on corporate profits imposed at the state and local level, as well as by foreign governments (although such foreign taxes are then generally deductible from the taxes due domestically).  This overall tax burden is more meaningful for understanding whether the overall burden is too high.  But, as we shall see below, that rate has also fallen steadily over time.  There is again no evidence that lower rates lead to higher growth.

The final substantive section of the post will then look more closely at the magnitude of the revenue losses from the December bill.  They are massive, and based on the initial evidence could very well total over $2 trillion over ten years for the losses on the corporate profit tax alone.  The losses from the other tax cuts in the new law, primarily for the wealthy and for non-corporate business, will add to this.  A very rough estimate is that the losses in individual income tax revenues may total an additional $1 trillion, bringing the total to over $3 trillion.  This is double the $1.5 trillion loss in revenues originally forecast.

But first, an analysis of what we see from the initial evidence on what is being paid.

B.  Profit Taxes as a Share of Corporate Profits

The chart at the top of this post shows what has been collected, by quarter (but shown at an annual rate), by the federal tax on corporate profits over the last several years.  Those figures are in dollars, and show a fall in the first half of 2018 of a half or more compared to what was collected in recent years.  But for comparisons over time, it is more meaningful to look at the implicit corporate tax rate, as corporate profits change over time (and generally grow over time).  And this can be done as the National Income and Product Accounts include an estimate of what corporate profits have been, as part of its assessment of how national income is distributed among the major functional groups.

That share since 2013 has been:

Between 2013 and 2016, the implicit rate (quarter by quarter) varied between about 15 and 17%.  It came down to about 14% for most of 2017 for some reason (possibly tied to the change in administration in Washington, with its new interpretation of regulatory and tax rules), but one cannot know from the aggregate figures alone.  But the rate then fell sharply, by half, to just 7% after the new tax law entered into effect.

A point to note is that the corporate profit figures provided here are corporate profits as estimated in the National Income and Product Accounts.  They are a measure of what corporate profits actually are, in an economic sense, and will in general differ from what corporate profits are as defined for tax purposes.  Thus, for example, accelerated depreciation allowed for tax purposes will reduce taxable corporate profits.  But the BEA estimates for the NIPA accounts will reflect not the accelerated depreciation allowed for tax purposes, but rather an estimate of what depreciation actually was.  Thus the figures as shown in the chart above will be a measure of what the true average corporate tax rate actually was, before the adjustments made (as permitted under tax law) to arrive at taxable corporate profits.

That average rate is now just 7%.  That is only one-third of the headline rate under the new law of 21%.  Provisions in the tax code allow corporations to pay far less in tax than what the headline rate would suggest.  This is not new (the headline rate previously was 35%, but the actual average rate paid was just 15 to 17% between 2013 and 2016, and 14% in most of 2017).  But Trump administration officials had asserted that many of the loopholes allowing for lower taxes would be ended under the new tax law, so that the actual rate paid would be closer to the headline rate.  But this clearly did not happen.  As many independent analysts pointed out before the bill was passed, the new tax law had numerous provisions which would allow the system to be gamed.  And we now see the result of that in the figures.

C.  Corporate Taxes in a Longer Term Context

The cuts in corporate profit taxes are not new.  Taxes on corporate profits in the US used to be far higher:

In the early 1950s, the federal tax on corporate profits (actually paid, not the headline rate) reached over 50%.  While it then fell, it kept to a rate of between about 30% and 50% through the 1950s and 60s.  And this was a period of good economic growth in the US – substantially faster than it has been since.  A high tax rate on corporate profits did not block growth.  Indeed, if one looked just at the simple correlation, one might conclude that a higher tax on corporate profits acts as a spur to growth.  But this would be too simplistic, and I would not argue that.  But what one can safely conclude is that a high rate of tax on corporate profits does not act as a block to more rapid growth.

There have also been important distributional consequences, however.  Corporate wealth is primarily owned by the wealthy (duh), and the sharp decline in taxes paid on corporate profits means that a larger share of the overall tax burden has been shifted to taxes on individual incomes, which are primarily borne by the middle classes.  Based on figures in the NIPA accounts, in 1950 taxes on individual incomes (including Social Security taxes) accounted for 47% of total federal taxes, while taxes on corporate profits accounted for 35% (with the rest primarily various excise taxes such as on fuels, liquor, tobacco, etc., plus import duties).  By 2017, however, the share of taxes on individual incomes had grown to 87.4%, while the share on corporate profits had declined to just 8.6%.  There was a gigantic shift away from taxes on wealth to taxes on individual incomes – taxes that are borne primarily by the middle class.  And that share will now fall further in 2018, by about half.

The chart above is for federal corporate profit taxes alone.  It could be argued that what matters to growth is not just the corporate profit taxes paid at the federal level, but all such taxes, including those paid at the state and local level, as well as to foreign governments (although the taxes paid abroad are generally deductible on their domestic taxes, so that will be a wash).

That chart looks like:

This follows the same path as the chart for federal corporate profit taxes alone, with a similar decline.  With the federal share of such taxes averaging 84% over the period (up to 2017), this is not surprising.  The federal share will now fall sharply in 2018, due to the new tax law.  But over the 1950 to 2017 period, the chart covering all taxes on corporate profits is basically a close to proportionate increase over what the tax has been at the federal level alone.

So the same pattern holds, and the total of the taxes on corporate profits varied between 33% to over 50% in the 1950s and 60s, to between 15 and 20% in recent years before the plunge in the first half of 2018 to just 10%.  But the relatively high taxes in the 1950s and 60s did not lead to slow growth in those years, nor did the low taxes in recent decades lead to more rapid growth.  Rather, one had the reverse.

D.  An Estimate of the Revenue Losses Due to the Tax Bill

These initial figures on the taxes actually being paid following the passage of the Republican tax bill allow us to make an estimate of what the revenue losses will turn out to be.  These will be very rough estimates, as we only have data for half a year, and one should be cautious in extrapolating this to what the losses will be over a decade.  But they can give us a sense of the magnitude.  And it is large.  As we will see below, based on the evidence so far the revenue losses (from the cuts in both corporate taxes and in personal income taxes) might be over $3 trillion over ten years, or about double the $1.5 trillion loss estimate originally forecast.

First, for the federal taxes on corporate profits, as the largest changes are there:  As was discussed before (and seen in the charts above), corporate profit taxes paid as a share of corporate profits were relatively flat between 2013 and 2016, varying between 15 and 17% each quarter, before falling to 14% for most of 2017.  For the full 2013 to 2017 period, the simple average was 15.3%.  The implicit rate then fell to just 7.0% in the first half of 2018.  Had the rate instead remained at 15.3%, corporate profit taxes collected in 2018 would have been $184 billion higher (on an annual basis).

This is not small, and is twice as high as the estimate of the staff of the Joint Committee on Taxation of revenue losses of $91 billion in FY2019 (the first full year under the new tax regime) from all the tax measures affecting businesses (including non-corporate businesses, and covering both domestic business and overseas business).  It is three times as high as the estimated loss of $60 billion in FY18, but the new tax law did not affect the first quarter of FY2018 (October to December).

One should be cautious with any extrapolation of this loss estimate going forward, as not only is the time period of actual experience under the new tax regime short (only a half year), but the law is also a complicated one, with certain provisions changing over time.  But a simple extrapolation over ten years, based on the assumption that corporate profits grow at just a modest 3% a year in nominal terms (meaning 1% a year in real terms if inflation is 2% a year), and that the tax rate on corporate profits will be 7.0% a year (as seen so far in 2018) rather than the 15.3% of recent years, implies that the reduction in corporate profit tax revenues will sum to about $2.1 trillion.

Note that the losses would be greater (everything else equal) if the assumed growth rate of corporate profits is higher.  But the results are not very sensitive to this.  The total losses over ten years would be $2.2 trillion, for example, at an assumed nominal growth rate of 4% (i.e. with inflation still at 2%, then with corporate profits growing at 2% a year in real terms, or double the 1% rate of the base scenario).  Note this also counters the argument of some that such tax cuts will lead to such a large spurt in growth that total tax collections will rise despite the cut in the rates.  As will be discussed below, there is no evidence that this has ever been the case in the US.  But even assuming there were, the argument is undermined by the basic arithmetic.  In the example here, a doubling of the assumed growth rate of profits (from 1% in real terms to 2%) would imply taxes on corporate profits would still fall by $2.0 trillion over the next ten years.  This is not far from the $2.1 trillion loss if there is no rise at all in the growth of corporate profits.  And a doubling of the real growth rate is far above what anyone would reasonably assume could follow from such a cut in the tax rate.

Second, there were also substantial cuts in individual income taxes, although primarily for the wealthy.  While far less in proportional terms, the substantially higher taxes that are now paid by individuals than by corporations means that this is also significant for the totals.

Specifically, individual (federal) income taxes as a share of GDP in the NIPA accounts were quite steady in the quarterly GDP accounts for the period from 2015Q1 to 2017Q4, varying only between 8.22% and 8.44%.  The average was 8.31%.  But then this fell to an average of 7.89% in the first half of 2018 (7.90% in the first quarter, and 7.87% in the second quarter).  Had the rate remained at 8.31%, then $86 billion more in revenues (at an annual rate) would have been collected.

Extrapolating this out for ten years, assuming again just a modest rate of growth for GDP of 3% a year in nominal terms (i.e. just 1% a year in real terms if inflation is 2% a year), the total loss would be $1.0 trillion.  With a higher rate of growth, and everything else the same, the losses would again be larger.  This extrapolation is, however, particularly fraught, as the Republicans wrote into their bill that the cuts in individual taxes would be reversed in 2026.  They did this to keep the forecast cost of the tax bill to the $1.5 trillion envelope they had set, and an effort is already underway to make this permanent (Speaker Paul Ryan has said he will schedule a vote on this in September).  But even if we left out the tax revenue losses in the final two years of the period, the losses in individual taxes would still reach about $0.8 trillion.

Adding the lower revenues from the taxes on corporate profits and the taxes on individual incomes, the total revenue losses would come, over the ten years, to about $3 trillion.  This is double the $1.5 trillion loss that had been forecast.  It is not a small difference.

To give a sense of the magnitude, the loss in 2018 alone (a total of $270 billion) would allow a doubling of the entire budgets (based on FY2017 actual outlays) of the Departments of Education, Housing and Urban Development, and Labor; the Environmental Protection Agency; all international assistance programs (foreign aid); NASA; the National Science Foundation; the Army Corps of Engineers (civil works); and the Small Business Administration.  Note I am not arguing that all of their budgets should necessarily be doubled (although many should, indeed, be significantly increased).  Rather, the point is simply to give readers a sense of the size of the revenues lost as a consequence of the tax cut bill.

As another comparison to give a sense of the magnitude, just half of the lost revenue (now and into the future) would suffice to fund fully the Social Security Trust Fund for the foreseeable future.  If nothing is done, the Social Security Trust Fund will run out at some point around 2034.  Republicans have asserted that nothing can be done for Social Security except to scale back (already low) Social Security pensions.  This is not true.  Just half of the revenues that will be lost under the tax cut bill would suffice to ensure the pensions can be paid in full for at least 75 years (the forecast period used by the Social Security trustees).

But as noted above, proponents of the tax cuts argue that the lower taxes will spur growth.  This has been discussed in earlier posts on this blog, where we have seen that there is no evidence that this will follow.  There are not only basic conceptual problems with this argument (a misreading of basic economics), but also no indication in what we have in fact observed for the economy that this has ever been the case (whether in the years immediately following the major tax cuts of Reagan or Bush, nor if one focuses on the longer term).

Administration officials have not surprisingly argued that the relatively rapid pace of growth in the second quarter of 2018 (of 4.2% at an annual rate in the end-August BEA estimates) is evidence of the tax cut working as intended.  But it is not.  Not only should one not place much weight on one quarter’s figures (the quarterly figures bounce around), but this followed first-quarter figures which were modest at best (with GDP growth of an estimated 2.2% at an annual rate).

But more fundamentally, one should dig into the GDP figures to see what is going on.  The argument that tax cuts (especially cuts in corporate profit taxes) will spur growth is based on the presumption that such cuts will spur business investment.  More such investment, especially in equipment, could lead to higher productivity and hence higher growth.  But growth in business investment in equipment has slowed in the first half of 2018.  Such investment grew at the rates of 9.1%, 9.7%, 9.8%, and 9.9% through the four quarters of 2017 (all at annual rates).  It then decelerated to a pace of 8.5% in the first quarter of 2018 and to a pace of 4.4% in the second quarter.  While still early (these figures too bounce around a good deal), the evidence so far is the exact opposite of what proponents have argued the tax cut bill would do.

So what might be going on?  As noted before, there is first of all a good deal of volatility in the quarterly figures for GDP growth.  But to the extent growth has accelerated this year, a more likely explanation is simple Keynesian stimulus.  Taxes were cut, and while most of the cuts went to the rich, some did go to the lower and middle classes.  In addition, government spending is now rising, while it been kept flat or falling for most of the Obama years (since 2010).  It is not surprising that such stimulus would spur growth in the short run.

The problem is that with the economy now running at or close to full capacity, such stimulus will not last for long.  And when it was needed, in the years from 2010 until 2016, as the economy recovered from the 2008/09 downturn (but slowly), such stimulus measures were repeatedly blocked by a Republican-controlled Congress.  This sequence for fiscal policy is the exact opposite of the path that should have been followed.  Contractionary policies were followed after 2010 when unemployment was still high, while expansionary fiscal policies are being followed now, when unemployment is low.  The result is that the fiscal deficit is rising soon to exceed $1 trillion in a year (5% of GDP), which is unprecedented for a period with the economy at full employment.

E.  Conclusion

We now have initial figures on what is being collected in taxes following the tax cut bill of last December.  While still early, the figures for the first two quarters of 2018 are nonetheless clear for corporate profit taxes:  They have fallen by half.  Corporate profit taxes paid would be an estimated $184 billion higher in 2018 had the tax rate remained at the level it had been over the last several years.

While this post has not focused on personal income taxes, they too were cut.  The reduction here was more modest – only by about 5% overall (although certain groups got far more, while others less).  But with their greater importance in overall federal tax collections, this 5% reduction is leading to an estimated $86 billion reduction in revenues (in 2018) from this source.

Based on what has been observed in the first two quarters of 2018, the two taxes together (corporate and individual) will see a combined reduction in taxes paid of about $270 billion in 2018.  Extrapolating over ten years, the combined losses may be on the order of $3 trillion.

These losses are huge.  And they are double what had been earlier forecast for the tax bill.  Just half of what is being lost would suffice to ensure Social Security would be fully funded for the foreseeable future.  And the rest could fund programs to rebuild and strengthen the physical infrastructure and human capital on which growth ultimately depends.  Or some could be used to reduce the deficit and pay down the public debt.  But instead, massive tax cuts are going to the rich.

Why Do the Quarterly GDP Figures Bounce Around So Much?: Econ 101

A.  Introduction

The Bureau of Economic Analysis (BEA) released on July 27 its initial estimate of GDP growth in the second quarter of 2018 (what it technically calls its “advance estimate”).  It was a good report:  Its initial estimate is that GDP grew at an annualized rate of 4.1% in real terms in the quarter.  Such growth, if sustained, would be excellent.

But as many analysts noted, there are good reasons to believe that such a growth rate will not be sustained.  There were special, one-time, factors, such as that the second quarter growth (at a 4.1% annual rate) had followed a relatively modest rate of growth in the first quarter of 2.2%.  Taking the two together, the growth was a good, but not outstanding, rate of 3.1%.

More fundamentally, with the economy now at full employment, few (other than Trump) expect growth at a sustained rate of 4% or more.  Federal Reserve Board members, for example, on average expect GDP growth of 2.8% in 2018 as a whole, with this coming down to a rate of 1.8% in the longer run.  And the Congression Budget Office (in forecasts published in April) is forecasting GDP growth of 3.0% in 2018, coming down to an average rate of 1.8% over 2018 to 2028.  The fundamental issue is that the population is aging, so the growth rate of the labor force is slowing.  As discussed in an earlier post on this blog, unless the productivity of those workers started to grow at an unprecedented rate (faster than has ever been achieved in the post-World War II period), we cannot expect GDP to grow for a sustained period going forward at a rate of 3%, much less 4%.

But there will be quarter to quarter fluctuations.  As seen in the chart at the top of this post covering the period just since 2006, there have been a number of quarters in recent years where GDP grew at an annualized rate of 4% or more.  Indeed, growth reached 5.1% in the second quarter of 2012, with this followed by an also high 4.9% rate in the next quarter.  But it then came back down.  And there were also two quarters (setting aside the period of the 2008/09 recession) which had growth of a negative 1.0%.  On average, GDP growth was around 2% (at an annual rate) during Obama’s two terms in office (2.2% annually from the end of the recession in mid-2009).

Seen in this context, the 4.1% rate in the initial estimate for the second quarter of 2018 was not special.  There have been a number of such cases (and with even substantially higher growth rates for a quarter or even two) in the recent past, even though average growth was just half that.  The quarterly rates bounce around.  But it is of interest to examine why they bounce around so much, and that is the purpose of this blog post.

B.  Reasons for this Volatility

There are a number of reasons why one should not be surprised that these quarter to quarter growth rates in GDP vary as they do.  I will present several here.  And note that these reasons are not mutually exclusive.  Several of them could be acting together and be significant factors in any given quarter.

a)  There may have been actual changes in growth:

To start, and to be complete, one should not exclude the possibility that the growth in the quarter (or the lack of it) was genuine.  Perhaps output did speed up (or slow down) as estimated.  Car plants might go on extra shifts (or close for a period) due to consumers wanting to buy more cars (or fewer cars) in the period for some reason.  There might also be some policy change that might temporarily spur production (or the opposite).  For example, Trump’s recent trade measures, and the response to them from our trading partners, may have brought forward production and trade that would have been undertaken later in the year, in order to avoid tariffs threatened to be imposed later.  This could change quarterly GDP even though GDP for the year as a whole will not be affected positively (indeed the overall impact would likely be negative).

[Side note:  But one special factor in this past quarter, cited in numerous news reports (see, for example, here, here, here, here, and here), was that a jump in exports of soybeans was a key reason for the higher-than-recently-achieved rate of GDP growth.  This was not correct.  Soybean exports did indeed rise sharply, with this attributed to the response threatened by China and others to the new tariffs Trump has imposed.  China and others said they would respond with higher tariffs of their own, targeted on products such as soybeans coming from the US.  There was thus a rush to export soybeans in the period between when China first announced they would impose such retaliatory tariffs (in late March) and when they were then imposed (ultimately on July 6).

But while soybean exports did indeed increase sharply in the April to June quarter, soybeans are a crop that takes many months to grow.  Whatever increase in shipments there was had to come out of inventories.  An increase in exports would have to be matched by a similar decrease in inventories, with this true also for corn and other such crops.  There would be a similar issue for any increase in exports of Kentucky bourbon, also a target of retaliatory tariffs.  Any decent bourbon is aged for at least three years.

One must keep in mind that GDP (Gross Domestic Product) is a measure of production, and the only production that might have followed from the increased exports of soybeans or similar products would have been of packing and shipping services.  But packing and shipping costs are only a relatively small share of the total value of the products being exported.

Having said that, one should not then go to the opposite extreme and assume that the threatened trade war had no impact on production and hence GDP in the quarter.  It probably did.  With tariffs and then retaliatory tariffs being threatened (but to be imposed two or three months in the future), there were probably increased factory orders to make and ship various goods before such new tariffs would enter into effect.  Thus there likely was some impact on GDP, but to an extent that cannot be quantified in what we see in the national level accounts.  And with such factory orders simply bringing forward orders that likely would have been made later in the year, one may well see a fallback in the pace of GDP growth in the remainder of the year.  But there are many other factors as well affecting GDP growth, and we will need to wait and see what the net impact will be.]

So one should not exclude the possibility that the fluctuation in the quarterly growth rate is real.  But it could be due to many other factors as well, as we will discuss below.

b)  Change at an Annualized Rate is Not the Change in a Quarter:

While easily confused, keep in mind also that in the accounts as normally published and presented in the US, the rates of growth of GDP (and of the other economic variables) are shown as annual equivalent rates.  The actual change in the quarter is only about one-fourth of this (a bit less due to compounding).  That is, in the second quarter of 2018, the BEA estimated that GDP (on a seasonally adjusted basis, which I will discuss below as a separate factor) grew by 1.00% (and yes, exactly 1.00% within two significant digits).  But at an annualized rate (some say “annual rate”, and either term can be used), this would imply a rate of growth of 4.06% (which rounded becomes 4.1%).  It is equal to slightly more than 4.0% due to compounding.  [Technically, 1% growth in the quarter means 1.00 will grow to 1.01, and taking 1.01 to the fourth power yields 1.0406, or an increase of 4.06%.]

Thus it is not correct to say that “GDP grew by 4.1% in the second quarter”.  It did not – it grew by 1.0%.  What is correct is to say that “GDP grew at an annualized rate of 4.1% in the second quarter”.

Not all national statistical agencies present such figures in annualized terms.  European agencies, for example, generally present the quarterly growth figures as simply the growth in the quarter.  Thus, for example, Eurostat on June 7 announced that GDP in the eurozone rose by an estimated 0.4% in the first quarter of 2018.  This 0.4% was the growth in the quarter.  But that 0.4% growth figure would be equivalent to growth of 1.6% on an annualized basis (actually 1.61%, if the growth had been precisely 0.400%).  Furthermore, the European agencies will generally also focus on the growth in GDP over what it had been a year earlier in that same quarter.  In the first quarter of 2018, this growth over the year-earlier period was an estimated 2.5% according to the Eurostat release.  But the growth since the year-earlier period is not the same as the growth in the current quarter at an annualized rate.  They can easily be confused if one is not aware of the conventions used by the different agencies.

c)  Don’t confuse the level of GDP with the change in GDP:

Also along the lines of how we might misleadingly interpret figures, one needs to keep in mind that while the quarterly growth rates can, and do, bounce around a lot, the underlying levels of GDP are really not changing much.  While a 4% annual growth rate is four times as high as a 1% growth rate, for example, the underlying level of GDP in one calendar quarter is only increasing to a level of about 101 (starting from a base of 100 in this example) with growth at a 4% annual rate, versus to a level of 100.25 when  growth is at an annual rate of 1%.  While such a difference in growth rates matters a great deal (indeed a huge deal) if sustained over time, the difference in any one quarter is not that much.

Indeed, I personally find the estimated quarter to quarter levels of GDP in the US (after seasonal adjustment, which will be discussed below) to be surprisingly stable.  Keep in mind that GDP is a flow, not a stock.  It is like the flow of water in a river, not a stock such as the body of water in a reservoir.  Flows can go sharply up and down, while stocks do not, and some may mistakenly treat the GDP figures in their mind as a stock rather than a flow.  GDP measures the flow of goods and services produced over some period of time (a calendar quarter in the quarterly figures).  A flow of GDP to 101 in some quarter (from a base of 100 in the preceding quarter) is not really that different to an increase to 100.25.  While this would matter (and matter a good deal) if the different quarterly increases are sustained over time, this is not that significant when just for one quarter.

d)  Statistical noise matters:

Moving now to more substantive reasons why one should expect a significant amount of quarter to quarter volatility, one needs to recognize that GDP is estimated based on surveys and other such sources of statistical information.  The Bureau of Economic Analysis (BEA) of the US Department of Commerce, which is responsible for the estimates of the GDP accounts in the US (which are formally called the National Income and Product Accounts, or NIPA), bases its estimates on a wide variety of surveys, samples of tax returns, and other such partial figures.  The estimates are not based on a full and complete census of all production each quarter.  Indeed, such an economic census is only undertaken once every five years, and is carried out by the US Census Bureau.

One should also recognize that an estimate of real GDP depends on two measures, each of which is subject to sampling and other error.  One does not, and cannot, measure “real GDP” directly.  Rather, one estimates what nominal GDP has been (based on estimates in current dollars of the value of all economic transactions that enter into GDP), and then how much prices have changed.  Price indices are estimated based on the prices of surveyed samples, and the components of real GDP are then estimated from the nominal GDP of the component divided by the relevant price index.  Real GDP is only obtained indirectly.

There will then be two sets of errors in the measurements:  One for the nominal GDP flows and one for the price indices.  And surveys, whether of income flows or of prices, are necessarily partial.  Even if totally accurate for the firms and other entities sampled, one cannot say with certainty whether those sampled in that quarter are fully representative of everyone in the economy.  This is in particular a problem (which the BEA recognizes) in capturing what is happening to newly established firms.  Such firms will not be included in the samples used (as they did not exist when the samples were set up) and the experiences of such newly established firms can be quite different from those of established firms.

And what I am calling here statistical “noise” encompasses more than simply sampling error.  Indeed, sampling error (the fact that two samples will come up with different results simply due to the randomness of who is chosen) is probably the least concern.  Rather, systemic issues arise whenever one is trying to infer measures at the national level from the results found in some survey.  The results will depend, for example, on whether all the components were captured well, and even on how the questions are phrased.  We will discuss below (in Section C, where we look at a comparison of estimates of GDP to estimates of Gross Domestic Income, or GDI, which in principle should be the same) that the statistical discrepancy between the estimates of GDP and GDI does not vary randomly from one quarter to the next but rather fairly smoothly (what economists and statisticians call “autocorrelation” – see Section C).  This is an indication that there are systemic issues, and not simply something arising from sample randomness.

Finally, even if that statistical error was small enough to allow one to be confident that we measured real GDP within an accuracy of just, say, +/- 1%, one would not then be able to say whether GDP in that quarter had increased at an annualized rate of about 4%, or decreased by about 4%.  A small quarterly difference looms large when looked at in terms of annualized rates.

I do not know what the actual statistical error might be in the GDP estimates, and it appears they are well less than +/- 1% (based on the volatility actually observed in the quarter to quarter figures).  But a relatively small error in the estimates of real GDP in any quarter could still lead to quite substantial volatility in the estimates of the quarter to quarter growth.

e)  Seasonal adjustment is necessary, but not easy to do:

Economic activity varies over the course of the year, with predictable patterns.  There is a seasonality to holidays, to when crops are grown, to when students graduate from school and enter the job market, and much much more.  Thus the GDP data we normally focus on has been adjusted by various statistical methods to remove the seasonality factor, making use of past data to estimate what the patterns are.

The importance of this can be seen if one compares what the seasonally adjusted levels of GDP look like compared to the levels before seasonal adjustment.  Note the level of GDP here is for one calendar quarter – it will be four times this at an annual rate:

There is a regular pattern to GDP:  It is relatively high in the last quarter of each year, relatively low in the first quarter, and somewhere in between in the second and third quarters.  The seasonally adjusted series takes account of this, and is far smoother.  Calculating quarterly growth rates from a series which has not been adjusted for seasonality would be misleading in the extreme, and not of much use.

But adjusting for seasonality is not easy to do.  While the best statisticians around have tried to come up with good statistical routines to do this, it is inherently difficult.  A fundamental problem is that one can only look for patterns based on what they have been in the past, but the number of observations one has will necessarily be limited.  If one went back to use 20 years of data, say, one would only have 20 observations to ascertain the statistical pattern.  This is not much.  One could go back further, but then one has the problem that the economy as it existed 30 or 40 years ago (and indeed even 20 years ago) was quite different from what it is now, and the seasonal patterns could also now be significantly different.  While there are sophisticated statistical routines that have been developed to try to make best use of the available data (and the changes observed in the economy over time), this can only be imperfect.

Indeed, the GDP estimates released by the BEA on July 27 incorporated a number of methodological changes (which we will discuss below), one of which was a major update to the statistical routines used for the seasonal adjustment calculations.  Many observers (including at the BEA) had noted in recent years that (seasonally adjusted) GDP growth in the first quarter of each year was unusually and consistently low.  It then recovered in the second quarter.  This did not look right.

One aim of the update to the seasonal adjustment statistical routines was to address this issue.  Whether it was fully successful is not fully clear.  As seen in the chart at the top of this post (which reflects estimates that have been seasonally adjusted based on the new statistical routines), there still appear to be significant dips in the seasonally adjusted first quarter figures in many of the years (comparing the first quarter GDP figures to those just before and just after – i.e. in 2007, 2008, 2010, 2011, 2014, and perhaps 2017 and 2018.  This would be more frequent than one would expect if the residual changes were now random over the period).  However, this is an observation based just on a simple look at a limited sample.  The BEA has looked at this far more carefully, and rigorously, and believes that the new seasonal adjustment routines it has developed have removed any residual seasonality in the series as estimated.

f)  The timing of weekends and holidays may also enter, and could be important:

The production of the goods and services that make up the flow of GDP will also differ on Saturdays, Sundays, and holidays.  But the number of Saturdays, Sundays, and certain holidays may differ from one year to the next.  While there are normally 13 Saturdays and 13 Sundays in each calendar quarter, and most holidays will be in the same quarter each year, this will not always be the case.

For example, there were just 12 Sundays in the first quarter of 2018, rather than the normal 13.  And there will be 14 Sundays in the third quarter of 2018, rather than the normal 13.  In 2019, we will see a reversion to the “normal” 13 Sundays in each of the quarters.  This could have an impact.

Assume, just for the sake of illustration, that production of what goes into GDP is only one-half as much on a Saturday, Sunday, or holiday, than it is on a regular Monday through Friday workday.  It will not be zero, as many stores, as well as certain industrial plants, are still open, and I am just using the one-half for illustration.  Using this, and based on a simple check of the calendars for 2018 and 2019, one will find there were 62 regular, Monday through Friday, non-holiday workdays in the first quarter of 2018, while there will be 61 such regular workdays in the first quarter of 2019.  The number of Saturdays, Sundays, and holidays were 28 in the first quarter of 2018 (equivalent to 14 regular workdays in terms of GDP produced, assuming the one-half figure), while the number of Saturdays, Sundays, and holidays will be 29 in the first quarter of 2019 (equivalent to 14.5 regular workdays).  Thus the total regular work-day equivalents will be 76 in 2018 (equal to 62 plus 14), falling to 75.5 in 2019 (equal to 61 plus 14.5).  This will be a reduction of 0.7% between the periods in 2018 and 2019 (75.5/76), or a fall of 2.6% at an annualized rate.  This is not small.

The changes due to the timing of holidays could matter even more, especially for certain countries around the world.  Easter, for example, was celebrated in March (the first quarter) in 2013 and 2016, but came in April (the second quarter) in 2014, 2015, 2017, and 2018.  In Europe and Latin America, it is customary to take up to a week of vacation around the Easter holidays.  The change in economic activity from year to year, with Easter celebrated in one quarter in one year but a different one in the next, will make a significant difference to economic activity as measured in the quarter.

And in Muslim countries, Ramadan (a month of fasting from sunrise to sunset), followed by the three-day celebration of Eid al-Fitr, will rotate through the full year (in terms of the Western calendar) as it is linked to the lunar cycle.

Hence it would make sense to adjust the quarterly figures not only for the normal seasonal adjustment, but also for any changes in the number of weekends and holidays in some particular calendar quarter.  Eurostat and most (but not all) European countries make such an adjustment for the number of working days in a quarter before they apply the seasonal adjustment factors.  But I have not been able to find how the US handles this.  The adjustment might be buried somehow in the seasonal adjustment routines, but I have not seen a document saying this.  If no adjustment is made, then this might explain part of the quarterly fluctuations seen in the figures.

g)  There have been, and always will be, updates to the methodology used:

As noted above, the GDP figures released on July 27 reflected a major update in the methodology followed by the BEA to arrive at its GDP estimates.  Not only was there extensive work on the seasonal adjustment routines, but there were definitional and other changes.  The accounts were also updated to reflect the findings from the 2012 Economic Census, and prices were changed from a previous base of 2009 to now 2012.  The July 27 release summarized the changes, and more detail on what was done is available from a BEA report issued in April.  And with the revisions in definitions and certain other methodological changes, the BEA revised its NIPA figures going all the way back to 1929, the first year with official GDP estimates.

The BEA makes such changes on a regularly scheduled basis.  There is normally an annual change released each year with the July report on GDP in the second quarter of the year.  This annual change incorporates new weights (from recent annual surveys) and normally some limited methodological changes, and the published estimates are normally then revised going back three and a half years.  See, for example, this description of what was done in July 2017.

On top of this, there is then a much larger change once every five years.  The findings from the most recent Economic Census (which is carried out every five years) are incorporated, seasonal adjustment factors are re-estimated, and major definitional or methodological changes may be incorporated.  The July 2018 release reflected one of those five-year changes.  It was the 15th such comprehensive revision to the NIPA accounts undertaken by the BEA.

I stress this to make the point that the GDP figures are estimates, and as estimates are always subject to change.  The professionals at the BEA are widely admired around the world for the quality of their work, and do an excellent job in my opinion.  But no estimates will ever be perfect.  One has to recognize that there will be a degree of uncertainty surrounding any such estimates, and that the quarter to quarter volatility observed will derive at least in part from the inherent uncertainty in any such estimates.

C.  Estimates of GDP versus Estimates of GDI

One way to develop a feel for how much the changes in quarterly GDP may be due to the inherent uncertainty in the estimates is to compare it to the estimated quarterly changes in Gross Domestic Income (GDI).  GDP (Gross Domestic Product) measures the value of everything produced.  GDI measures the value of all incomes (wages, profits, rents, etc.) generated.  In principle, the totals should be the same, as the value of whatever is produced accrues to someone as income.  They should add up to the same thing.

But the BEA arrives at its estimates of GDP and of GDI by different routes.  As a consequence, the estimates of the totals will then differ.  The differences are not huge in absolute amount, nor have they grown over time (as a share of GDP or of GDI).  That is, on average the estimates match each other over time, with the same central tendency.  But they differ by some amount in any individual quarter, and hence the quarter to quarter growth rates will differ.  And for the reasons reviewed above, those slight changes in the levels in any individual quarter can translate into often major differences in the growth rates from one quarter to the next.  And these differences may appear to be particularly large when the growth rates are then presented in annualized terms.

For the period since 2006, the two sets of growth rates were (where the initial estimate for the second quarter of 2018 will not be available until the end-August figures come out):

As is seen, the alternative estimates of growth in any individual quarter can be quite different.  There was an especially large difference in the first quarter of 2012, when the estimated growth in GDP was 3.2% at an annual rate, while the estimated growth in GDI was a giant 8.7%.

Which is correct?  Was the growth rate in the first quarter of 2012 3.2% (as found with the GDP estimate) or 8.7% (as found with the GDI estimate)?  The answer is we do not know, and indeed that probably neither is correct.  What is most likely is that the true figure is probably somewhere in between.

Furthermore, and also moderating what the impact on the differences in the respective estimated growth rates will be, it is not the case that the estimates of GDP and GDI are statistically independent of each other, with the two bouncing around randomly with respect to each other.  Rather, if one looks at what the BEA calls the “statistical discrepancy” (the difference between GDP and GDI), one finds that if, say, the estimate of GDP were above the estimate of GDI in one quarter, then it likely would also be above in the next quarter.  Not by the same amount, and the differences would evolve over time, but moving more like waves than as balls ricocheting around.  Economists and statisticians refer to this as “autocorrelation”, and it indicates that there is some systemic error in the estimates of GDP and of GDI, which carries over from one quarter to the next.  What the source of that is, we do not know.  If we did know, then it would be eliminated.  But the fact such autocorrelation exists tells us that there is some source of systemic error in the measures of GDP and GDI, and we have not been able to discover the source.

Estimates are estimates.  We need to recognize that there will be statistical uncertainty in any such figures.  Even if they even out over time, the estimated growth from one quarter to the next will reflect such statistical volatility.  The differences seen in the estimated rates of growth in any one quarter for total output (estimated by way of GDP versus by way of GDI) provides a useful benchmark for how to judge the reported changes seen in growth for GDP in any individual quarter.  The true volatility (for purely statistical reasons) is likely to be at least as much, if not more.

D.  Conclusion

There are many reasons, then, to expect the quarterly growth figures to bounce around.  One should not place too much weight on the estimates from any individual quarter.  It is the longer term trends that matter.  The estimated figure for growth in GDP of 4.1% in the second quarter was not out of line with what has been seen in a number of quarters in recent years.  But growth since mid-2009 has only been about one half as much on average, despite several quarters when estimated growth was well in excess of 4.1%.

To conclude, some may find of interest three country cases I am personally familiar with which illustrate why one needs to exercise care, and with an understanding of the country context, when considering what is meaningful or not for reported figures on GDP growth.  The countries are Japan, China, and an unidentified, but newly independent, former colony in the 1960s.

a)  Japan:  In the late 1990s / early 2000s, while holding a position within the World Bank Group, I was responsible for assessments of the prospects and risks of the countries of East Asia where the World Bank was active.  This was not long after the East Asia crisis of 1997, and the countries were just beginning to recover.  Japan was important, both as a trading partner to the others and because Japan itself had gone through a somewhat similar crisis following 1990, when the Japanese financial bubble burst.

As part of this, I followed closely the quarterly GDP growth figures for Japan.  But as many analysts at the time noted, the quarter to quarter figures behaved in ways that were difficult to understand.  Components went up when one would have thought they would go down (and vice versa), the quarterly changes were far more extreme than seen elsewhere, and in general the quarter to quarter fluctuations were difficult to make sense of.  The volatility in the figures was far greater than one would have expected for an economy such as Japan’s.

This view among analysts was such a common one that the government agency responsible for the estimates felt it necessary to issue a news release in June 2000 defending its work and addressing a number of the concerns that had been raised.

I have no doubt that the Japanese government officials responsible for the estimates were well-qualified and serious professionals.  But it is not easy to estimate GDP and its components, the underlying data on which the statisticians relied might have had problems (including sample sizes that were possibly too small), and there may have been segments of the economy (in the less formal sectors) which might not have been captured well.

I have not followed closely in recent years, and do not know if the issues continue.  But Japan’s case illustrates that even a sophisticated agency, with good professionals, can have difficulty in arriving at GDP estimates that behave as one would expect.

b)  China:  The case of China illustrates the mirror image problem of what was found in Japan.  While the Japanese GDP estimates bounced around far too sharply from one quarter to the next, the GDP estimates for China showed remarkable, and not believable, stability.

Chinese growth rates have normally been presented as growth of GDP in the current period over what it was in the same period one year ago.  Seasonal adjustment is then not needed, and indeed China only started to present seasonally adjusted figures in 2011.  However, these estimates are still not fully accepted by many analysts.  Comparing GDP in the current quarter to what it was in the same quarter a year before overcomes this, but at the cost that it does not present information on growth just in the quarter, as opposed to total growth over the preceding year.

And the growth rates reported over the same quarter in the preceding year have been shockingly smooth.  Indeed, in recent years (from the first quarter of 2015 through to the recently released figures for the second quarter of 2018), China’s reported growth of its GDP over the year-earlier period has not been more than 7.0% nor less than 6.7% in each and every quarter.  Specifically, the year on year GDP growth rates from the first quarter of 2015 through to the second quarter of 2018 were (in sequence):  7.0%, 7.0%, 6.9%, 6.8%, 6.7%, 6.7%, 6.7%, 6.8%, 6.9%, 6.9%, 6.8%, 6.8%, 6.8%, and 6.7% (one can find the figures in, for example, the OECD database).  Many find this less than credible.

There are other problems as well in the Chinese numbers.  For example, it has often been the case that the reported growth in provincial GDP of the 31 provincial level entities in China was higher in almost all of the 31 provinces, and sometimes even in all of the provinces, than GDP growth was in China as a whole.  This is of course mathematically impossible, but not surprising when political rewards accrue to those with fast reported growth.

With such weak credibility, analysts have resorted to coming up with proxies to serve as indicators of what quarter to quarter might have been.  These might include electricity consumption, or railway tonnage carried, or similar indicators of economic production.  Indeed, there is what has been labeled the “Li index”, named after Li Keqiang (who was vice premier when he formulated it, and later China’s premier).  Li said he did not pay much attention to the official GDP statistics, but rather focused on a combination of electricity production, rail cargo shipments, and loan disbursements.  Researchers at the Federal Reserve Bank of San Francisco who reproduced this and fitted it through some regression analysis found that it worked quite well.

And the index I found most amusing is calculated using nighttime satellite images of China, with an estimation of how much more night-time illumination one finds over time.  This “luminosity” index tracks well what might be going on with China’s GDP.

c)  An unidentified, newly independent, former colony:  Finally, this is a story which I must admit I received third hand, but which sounds fully believable.  In the mid-1970s I was working for a period in Kuala Lumpur, for the Government of Malaysia.  As part of an economic modeling project I worked closely with the group in the national statistical office responsible for estimating GDP.  The group was led by a very capable, and talkative, official (of Tamil origin), who related a story he had heard from a UN consultant who had worked closely with his group in the early 1970s to develop their system of national accounts.

The story is of a newly independent country in the mid-1960s (whose name I was either not told or cannot remember), and its estimation of GDP.  An IMF mission had visited it soon after independence, and as is standard, the IMF made forecasts of what GDP growth might be over the next several years.  Such forecasts are necessary in order to come up with estimates for what the government accounts might be (as tax revenues will depend on GDP), for the trade accounts, for the respective deficits, and hence for what the financing needs might be.

Such forecasts are rarely very good, especially for a newly independent country where much is changing.  But something is needed.

As time passed, the IMF received regular reports from the country on what estimated GDP growth actually was.  What they found was that reported GDP growth was exactly what had been forecast.  And when asked, the national statisticians responded that who were they to question what the IMF officials had said would happen!

What Has Been Happening to Real Wages? Sadly, Not Much

A.  Introduction

There is little that is more important to a worker than his or her wages.  And as has been discussed in an earlier post on this blog, real wages in the US have stagnated since around 1980.  An important question is whether this has changed recently.  Trump has claimed that his policies (of lifting regulations, slashing corporate taxes, and imposing high tariffs on our trading partners) are already leading to higher wages for American workers.  Has that been the case?

The answer is no.  As the chart at the top of this post shows, real wages have been close to flat.  Nominal wages have grown with inflation, but once inflation is taken into account, real wages have barely moved.  And one does not see any sharp change in that trend after Trump took office in January 2017.

It is of course still early in Trump’s term, and the experience so far does not mean real wages will not soon rise.  We will have to see.  One should indeed expect that they would, as the unemployment rate is now low (continuing the path it has followed since 2010, first under Obama and now, at a similar pace, under Trump).  But the primary purpose of this blog post is to look at the numbers on what the experience has been in recent years, including since Trump took office.  We will see that the trend has not much changed.  And to the extent that it has changed, it has been for the worse.

We will first take an overall perspective, using the chart at the top of this post and covering the period since 2006.  This will tell us what the overall changes have been over the full twelve years.  For real wages, the answer (as noted above) is that not much has changed.

But the overall perspective can mask what the year to year changes have been.  So we will then examine what these have been, using 12 month moving averages for the changes in nominal wages, the consumer price index, and then the real wage.  And we will see that changes in the real wage have actually been trending down of late, and indeed that the average real wage in June 2018 was below where it had been in June 2017.

We will then conclude with a short discussion of whether labor market trends have changed since Trump took office.  They haven’t.  But those trends, in place since 2010 as the economy emerged from the 2008/09 downturn, have been positive.  At some point we should expect that, if sustained, they will lead to rising real wages.  But we just have not seen that yet.

B.  Nominal and Real Wages Since 2006

It is useful first to start with an overall perspective, before moving to an examination of the year to year changes.  The chart at the top of this post shows average nominal wages in the private sector, in dollars per hour, since March 2006, and the equivalent in real terms, as deflated by the consumer price index (CPI).  The current CPI takes the prices of 1982-84 (averaged over that period) as the base, and hence the real wages shown are in terms of the prices of 1982-84.  For June 2018, for example, average private sector wages were $26.98 per hour, equivalent to $10.76 per hour in terms of the prices of 1982-84.

The data series comes from the Current Employment Survey of the Bureau of Labor Statistics, which comes out each month and is the source of the closely watched figures on the net number of jobs created each month.  The report also provides figures on average private sector wages on a monthly basis, but this particular series only started being reported in March 2006.  That is part of the reason why I started the chart with that date, but it is in any case a reasonable starting point for this analysis as it provides figures starting a couple of years before the economic collapse of 2008, in the last year of Bush’s presidential term, through to June 2018.

The BLS report also only provides figures on average wages in the private sector.  While it would be of interest also to see the similar figures on government wages, they are not provided for some reason.  If they had been included, the overall average wage would likely have increased at an even slower pace than that shown for the private sector only, as government wages have been increasing at a slower pace than private wages over this period.  But government employment is only 15% of total employment in the US.  Private wages are still of interest, and will provide an indication of what the market pressures have (or have not) been.

The chart shows that nominal wages have increased at a remarkably steady pace over this period.  Many may find that lack of fluctuation surprising.  The economy in 2008 and early 2009 went through the sharpest economic downturn since the Great Depression, and unemployment eventually hit 10.0% (in October 2009).  Yet nominal private sector wages continued to rise.  As we will discuss in more detail below, nominal wages were increasing at about a 3% annual pace through 2008, and then continued to increase (but at about a 2% pace) even after unemployment jumped.

But while nominal wages rose at this steady pace, it was almost all just inflation.  After adjusting for inflation, average real wages were close to flat for the period as a whole.  They were not completely flat:  Average real wages over the period (March 2006 to June 2018) rose at an annual rate of 0.57% per year.  This is not much.  It is in fact remarkably similar to the 0.61% growth in the average real wage between 1979 and 2013 in the data that were discussed in my blog post from early 2015 that looked at the factors underlying the stagnation in real wages in the decades since 1980.

But as was discussed in that blog post, the average real wage is not the same as the median real wage.  The average wage is the average across all wage levels, including the wages of the relatively well off.  The median, in contrast, is the wage at the point where 50% of the workers earn less and 50% earn more.  Due to the sharp deterioration in the distribution of income since around 1980 (as discussed in that post), the median real wage rose by less than the average real wage, as the average was pulled up by the more rapid increase in wages of those who are relatively well off.  And indeed, the median real wage rose by almost nothing over that period (just 0.009% per year between 1979 and 2013) when the average real wage rose at the 0.61% per year pace.  If that same relationship has continued, there would have been no increase at all in the median real wage in the period since 2006.  But the median wage estimates only come out with a lag (they are estimated through a different set of surveys at the Census Bureau), are only worked out on an annual basis, and we do not yet have such estimates for 2018.

C.  12 Month Changes in Nominal Wages, the Consumer Price Index, and Real Wages Since 2006

While the chart at the top of this post tracks the cumulative changes in wages over this period, one can get a better understanding of the underlying dynamics by looking at how the changes track over time.  For this we will focus on percentage changes over 12 month periods, worked out month by month on a moving average basis.  Or another way of putting it, these will be the percentage changes in the wages or the CPI over what it had been one year earlier, worked out month by month in overlapping periods.

For average nominal wages (in the private sector) this is:

Note that the date labels are for the end of each period.  Thus the point labeled at the start of 2008 will cover the percentage change in the nominal wage between January 2007 and January 2008.  And the starting date label for the chart will be March 2007, which covers the period from March 2006 (when the data series begins) to March 2007.

Prior to the 2008/09 downturn, nominal wages were growing at roughly 3% a year.  Once the downturn struck they continued to increase, but at a slower pace of roughly 2% a year or a bit below.  And this rate then started slowly to rise over time, reaching 2.7% in the most recent twelve-month period ending in June 2017.  The changes are remarkably minor, as was also noted above, and cover a period where unemployment was as high as 10% and is now just 4%.  There has been very little year to year volatility.

[A side note:  There is a “bump” in late 2008/early 2009, with wage growth over the year earlier period rising from around 3% to around 3 1/2%.  This might be considered surprising, as the bump up is precisely in the period when jobs were plummeting and unemployment increasing, in the worst period of the economic collapse.  But while I do not have the detailed microdata from the BLS surveys to say with certainty, I suspect this is a compositional effect.  When businesses start to lay off workers, they will typically start with the least experienced, and lowest paid, workers.  That will leave them with a reduced labor force, but one whose wages are on average higher.]

There have been larger fluctuations in the consumer price index:

But note that “larger” should be interpreted in a relative sense.  The absolute changes were generally not all that large (with some exceptions), and can mostly be attributed to changes in the prices of a limited number of volatile commodities, namely for food items and energy (oil).  The prices of such commodities go up and down, but over time they even out.  Thus for understanding inflationary trends, analysts will often focus instead on the so-called “core CPI”, which excludes food and energy prices.  For the full period being examined here, the regular CPI rose at a 1.88% annual pace while the core CPI rose at a 1.90% pace.  Within round-off, these are essentially the same.

But what matters to wage earners is what their wages earn, including for food and energy.  Thus to examine the impact on real living standards, what matters is the real wage defined in terms of the regular CPI index.  And this was:

With the relatively steady changes in average nominal wages, year to year, the fluctuations will basically be the mirror image of what has been happening to inflation.  When prices fell, real wages rose, and when prices rose more than normal, real wages fell.

Prices are now again rising, although still within the norm of the last twelve years.  For the 12 months ending in June 2018, the CPI (using the seasonally adjusted series) rose at a 2.8% rate.  The average nominal wage rate rose at a rate of 2.74% and thus the real wage fell slightly by 0.05% (calculated before rounding).  Average real wages are basically the same as (and formally slightly below) where they were a year ago.

D.  Employment and Unemployment

There is thus no evidence that the measures Trump has trumpeted (of deregulation, slashing taxes for corporations, and launching a trade war) have led to a step up in real wages.  This should not be surprising.  Deregulation which spurs industry consolidation increases the power of firms to raise prices while holding down wages.  And there is no reason to believe that tax cuts will lead quickly to higher wages.  Corporations do not pay their workers out of generosity or out of some sense of charity.  In a market economy they pay their employees what they need to in order to get the workers in the number and quality they need.  And although there can be winners in a trade war, there will also certainly be losers, and overall there will be a loss.  Workers, on average, will lose.

But what is surprising is that wages are not now rising by more in an economy that has reached full employment.  Federal Reserve Chair Jerome Powell, for example, has called this “a puzzle”.  And indeed it is.

The labor market turned around in the first two years of the Obama administration, and since then employment has grown consistently:

This has continued (although at a slightly slower pace) since Trump took office in January 2017.  The same trend as before has continued.  And this trend growth in net jobs each month has meant a steady fall in the unemployment rate:

Again, the pace since Trump took office is similar to (but a bit slower than) the pace when Obama was still in office.  But the somewhat slower pace should not be surprising.  With the economy at close to full employment, one should expect the pace to slow.

Indeed, the unemployment rate cannot go much lower.  There is always a certain amount of “churn” in the job market, which means an unemployment rate of zero is impossible.  And many economists in fact have taken a somewhat higher rate of unemployment (or at least 5.0%) as the appropriate target for “full employment”, arguing that anything lower will lead to a wage and price spiral.

But we have not seen any sign of that so far.  Nominal wages are rising at only a modest pace, and indeed over the last year at a pace less than inflation.

E.  Conclusion

There has been no step up in real wages since Trump took office.  Indeed, over the past twelve months, they fell slightly.  But while there is no reason to believe there should have been a jump in real wages following from Trump’s economic policies (of deregulation, tax cuts for corporations, and trade war), it is surprising that the economy is not now well past the point where low unemployment should have been spurring more substantial wage gains.

This very well could change, and indeed I would expect it to.  There is good reason to believe that the news for the real wage will be a good deal more positive over the next year than it has been over the past year.  But we will have to wait and see.  So far it has not happened.

The Simple Economics of What Determines the Foreign Trade Balance: Econ 101

“There’s no reason that we should have big trade deficits with virtually every country in the world.”

“We’re like the piggybank that everybody is robbing.”

“the United States has been taken advantage of for decades and decades”

“Last year,… [the US] lost  … $817 billion on trade.  That’s ridiculous and it’s unacceptable.”

“Well, if they retaliate, they’re making a mistake.  Because, you see, we have a tremendous trade imbalance. … we can’t lose”

Statements made by President Trump at the press conference held as he left the G-7 meetings in, Québec, Canada, June 9, 2018.

 

A.  Introduction

President Trump does not understand basic economics.  While that is not a surprise, nor something necessarily required or expected of a president, one should expect that a president would appoint advisors who do understand, and who would tell him when he is wrong.  Unfortunately, this president has been singularly unwilling to do so.  This is dangerous.

Trump is threatening a trade war.  Not only by his words at the G-7 meetings and elsewhere, but also by a number of his actions on trade and tariffs in recent months, Trump has made clear that he believes that a trade deficit is a “loss” to the nation, that countries with trade surpluses are somehow robbing those (such as the US) with a deficit, that raising tariffs can and will lead to reductions in trade deficits, and that if others then also raise their tariffs, the US will in the end necessarily “win” simply because the US has a trade deficit to start.

This is confused on many levels.  But it does raise the questions of what determines a country’s trade balance; whether a country “loses” if it has a trade deficit; and what is the role of tariffs.  This Econ 101 blog post will first look at the simple economics of what determines a nation’s trade deficit (hint:  it is not tariffs); will then discuss what tariffs do and where do they indeed matter; and will then consider the role played by foreign investment (into the US) and whether a trade deficit can be considered a “loss” for the nation (a piggybank being robbed).

B.  What Determines the Overall Trade Deficit?

Let’s start with a very simple case, where government accounts are aggregated together with the rest of the economy.  We will later then separate out government.

The goods and services available in an economy can come either from what is produced domestically (which is GDP, or Gross Domestic Product) or from what is imported.  One can call this the supply of product.  These goods and services can then be used for immediate consumption, or for investment, or for export.  One can call this the demand for product.  And since investment includes any net change in inventories, the goods and services made available will always add up to the goods and services used.  Supply equals demand.

One can put this in a simple equation:

GDP + Imports = Domestic Consumption + Domestic Investment + Exports

Re-arranging:

(GDP – Domestic Consumption) – Domestic Investment = Exports – Imports

The first component on the left is Domestic Savings (what is produced domestically less what is consumed domestically).  And Exports minus Imports is the Trade Balance.  Hence one has:

Domestic Savings – Domestic Investment = Trade Balance

As one can see from the way this was derived, this is simply an identity – it always has to hold.  And what it says is that the Trade Balance will always be equal to the difference between Domestic Savings and Domestic Investment.  If Domestic Savings is less than Domestic Investment, then the Trade Balance (Exports less Imports) will be negative, and there will be a trade deficit.  To reduce the trade deficit, one therefore has to either raise Domestic Savings or reduce Domestic Investment.  It really is as straightforward as that.

Where this becomes more interesting is in determining how the simple identity is brought about.  But here again, this is relatively straightforward in an economy which, like now, is at full employment.  Hence GDP is essentially fixed:  It cannot immediately rise by either employing more labor (as all the workers who want a job have one), nor by each of those laborers suddenly becoming more productive (as productivity changes only gradually through time by means of either better education or by investment in capital).  And GDP is equal to labor employed times the productivity of each of those workers.

In such a situation, with GDP at its full employment level, Domestic Savings can only rise if Domestic Consumption goes down, as Domestic Savings equals GDP minus Domestic Consumption.  But households want to consume, and saving more will mean less for consumption.  There is a tradeoff.

The only other way to reduce the trade deficit would then be to reduce Domestic Investment.  But one generally does not want to reduce investment.  One needs investment in order to become more productive, and it is only through higher productivity that incomes can rise.

Reducing the trade deficit, if desirable (and whether it is desirable will be discussed below), will therefore not be easy.  There will be tradeoffs.  And note that tariffs do not enter directly in anything here.  Raising tariffs can only have an impact on the trade balance if they have a significant impact for some reason on either Domestic Savings or Domestic Investment, and tariffs are not a direct factor in either.  There may be indirect impacts of tariffs, which will be discussed below, but we will see that the indirect effects actually could act in the direction of increasing, not decreasing, the trade deficit.  However, whichever direction they act in, those indirect effects are likely to be small.  Tariffs will not have a significant effect on the trade balance.

But first, it is helpful to expand the simple analysis of the above to include Government as a separate set of accounts.  In the above we simply had the Domestic sector.  We will now divide that into the Domestic Private and the Domestic Public (or Government) sectors.  Note that Government includes government spending and revenues at all levels of government (state and local as well as federal).  But the government deficit is primarily a federal government issue.  State and local government entities are constrained in how much of a deficit they can run over time, and the overall balance they run (whether deficit or surplus) is relatively minor from the perspective of the country as a whole.

It will now also be convenient to write out the equations in symbols rather than words, and we will use:

GDP = Gross Domestic Product

C = Domestic Private Consumption

I = Domestic Private Investment

G = Government Spending (whether for Consumption or for Investment)

X = Exports

M = Imports

T = Taxes net of Transfers

Note that T (Taxes net of Transfers) will be the sum total of all taxes paid by the private sector to government, minus all transfers received by the private sector from government (such as for Social Security or Medicare).  I will refer to this as simply net Taxes (T).

The basic balance of goods or services available (supplied) and goods or services used (demanded) will then be:

GDP + M = C + I + G + X

We will then add and subtract net Taxes (T) on the right-hand side:

GDP + M = (C + T) + I + (G – T) + X

Rearranging:

GDP – (C + T) – (G – T) – I = X – M

(GDP – C – T) – I + (T – G) = X – M

Or in (abbreviated) words:

Dom. Priv. Savings – Dom. Priv. Investment + Govt Budget Balance = Trade Balance

Domestic Private Savings (savings by households and private businesses) is equal to what is produced in the economy (GDP), less what is privately consumed (C), less what is paid in net Taxes (T) by the private sector to the public sector.  Domestic Private Investment is simply I, and includes investment both by private businesses and by households (primarily in homes).  And the Government Budget Balance is equal to what government receives in net Taxes (T), less what Government spends (on either consumption items or on public investment).  Note that government spending on transfers (e.g. Social Security) is already accounted for in net Taxes (T).

This equation is very much like what we had before.  The overall Trade Balance will equal Domestic Private Savings less Domestic Private Investment plus the Government Budget Balance (which will be negative when a deficit, as has normally been the case except for a few years at the end of the Clinton administration).  If desired, one could break down the Government Budget Balance into Public Savings (equal to net Taxes minus government spending on consumption goods and services) less Public Investment (equal to government spending on investment goods and services), to see the parallel with Domestic Private Savings and Domestic Private Investment.  The equation would then read that the Trade Balance will equal Domestic Private Savings less Domestic Private Investment, plus Government Savings less Government Investment.  But there is no need.  The budget deficit, as commonly discussed, includes public spending not only on consumption items but also on investment items.

This is still an identity.  The balance will always hold.  And it says that to reduce the trade deficit (make it less negative) one has to either increase Domestic Private Savings, or reduce Domestic Private Investment, or increase the Government Budget Balance (i.e. reduce the budget deficit).  Raising Domestic Private Savings implies reducing consumption (when the economy is at full employment, as now).  Few want this.  And as discussed above, a reduction in investment is not desirable as investment is needed to increase productivity over time.

This leaves the budget deficit, and most agree that it really does need to be reduced in an economy that is now at full employment.  Unfortunately, Trump and the Republican Congress have moved the budget in the exact opposite direction, primarily due to the huge tax cut passed last December, and to a lesser extent due to increases in certain spending (primarily for the military).  As discussed in an earlier post on this blog, an increase in the budget deficit to a forecast 5% of GDP at a time when the economy is at full employment is unprecedented in peacetime.

What this implies for the trade balance is clear from the basic identity derived above.  An increase in the budget deficit (a reduction in the budget balance) will lead, all else being equal, to an increase in the trade deficit (a reduction in the trade balance).  And it might indeed be worse, as all else is not equal.  The stated objective of slashing corporate taxes is to spur an increase in corporate investment.  But if private investment were indeed to rise (there is in fact little evidence that it has moved beyond previous trends, at least so far), this would further worsen the trade balance (increase the trade deficit).

Would raising tariffs have an impact?  One might argue that this would raise net Taxes paid, as tariffs on imports are a tax, which (if government spending is not then also changed) would reduce the budget deficit.  While true, the extent of the impact would be trivially small.  The federal government collected $35.6 billion in all customs duties and fees (tariffs and more) in FY2017 (see the OMB Historical Tables).  This was less than 0.2% of FY2017 GDP.  Even if all tariffs (and other fees on imports) were doubled, and the level of imports remained unchanged, this would only raise 0.2% of GDP.  But the trade deficit was 2.9% of GDP in FY2017.  It would not make much of a difference, even in such an extreme case.  Furthermore, new tariffs are not being pushed by Trump on all imports, but only a limited share (and a very limited share so far).  Finally, if Trump’s tariffs in fact lead to lower imports of the items being newly taxed, as he hopes, then tariffs collected can fall.  In the extreme, if the imports of such items go to zero, then the tariffs collected will go to zero.

Thus, for several reasons, any impact on government revenues from the new Trump tariffs will be minor.

The notion that raising tariffs would be a way to eliminate the trade deficit is therefore confused.  The trade balance will equal the difference between Domestic Savings and Domestic Investment.  Adding in government, the trade balance will equal the difference between Domestic Private Savings and Domestic Private Investment, plus the equivalent for government (the Government Budget Balance, where a budget deficit will be a negative).  Tariffs have little to no effect on these balances.

C.  What Role Do Tariffs Play, Then?

Do tariffs then matter?  They do, although not in the determination of the overall trade deficit.  Rather, tariffs, which are a tax, will change the price of the particular import relative to the price of other products.  If applied only to imports from some countries and not from others, one can expect to see a shift in imports towards those countries where the tariffs have not been imposed.  And in the case when they are applied globally, on imports of the product from any country, one should expect that prices for similar products made in the US will then also rise.  To the extent there are alternatives, purchases of the now more costly products (whether imported or produced domestically) will be reduced, while purchases of alternatives will increase.  And there will be important distributional changes.  Profits of firms producing the now higher priced products will increase, while the profits of firms using such products as an input will fall.  And the real incomes of households buying any of these products will fall due to the higher prices.

Who wins and who loses can rapidly become turn into something very complicated.  Take, for example, the new 25% tariff being imposed by the Trump administration on steel (and 10% on aluminum).  The tariffs were announced on March 8, to take effect on March 23.  Steel imports from Canada and Mexico were at first exempted, but later the Trump administration said those exemptions were only temporary.  On March 22 they then expanded the list of countries with temporary exemptions to also the EU, Australia, South Korea, Brazil, and Argentina, but only to May 1.  Then, on March 28, they said imports from South Korea would receive a permanent exemption, and Australia, Brazil, and Argentina were granted permanent exemptions on May 2.  After a short extension, tariffs were then imposed on steel imports from Canada, Mexico, and the EU, on May 31.  And while this is how it stands as I write this, no one knows what further changes might be announced tomorrow.

With this uneven application of the tariffs by country, one should expect to see shifts in the imports by country.  What this achieves is not clear.  But there are also further complications.  There are hundreds if not thousands of different types of steel that are imported – both of different categories and of different grades within each category – and a company using steel in their production process in the US will need a specific type and grade of steel.  Many of these are not even available from a US producer of steel.  There is thus a system where US users of steel can apply for a waiver from the tariff.  As of June 19, there have been more than 21,000 petitions for a waiver.  But there were only 30 evaluators in the US Department of Commerce who will be deciding which petitions will be granted, and their training started only in the second week of June.  They will be swamped, and one senior Commerce Department official quoted in the Washington Post noted that “It’s going to be so unbelievably random, and some companies are going to get screwed”.  It would not be surprising to find political considerations (based on the interests of the Trump administration) playing a major role.

So far, we have only looked at the effects of one tariff (with steel as the example).  But multiple tariffs on various goods will interact, with difficult to predict consequences.  Take for example the tariff imposed on the imports of washing machines announced in late January, 2018, at a rate of 20% in the first year and at 50% should imports exceed 1.2 million units in the year.  This afforded US producers of washing machines a certain degree of protection from competition, and they then raised their prices by 17% over the next three months (February to May).

But steel is a major input used to make washing machines, and steel prices have risen with the new 25% tariff.  This will partially offset the gains the washing machine producers received from the tariff imposed on their product.  Will the Trump administration now impose an even higher tariff on washing machines to offset this?

More generally, the degree to which any given producer will gain or lose from such multiple tariffs will depend on multiple factors – the tariff rates applied (both for what they produce and for what they use as inputs), the degree to which they can find substitutes for the inputs they need, and the degree to which those using the product (the output) will be able to substitute some alternative for the product, and more.  Individual firms can end up ahead, or behind.  Economists call the net effect the degree of “net effective protection” afforded the industry, and it can be difficult to figure out.  Indeed, government officials who had thought they were providing positive protection to some industry often found out later that they were in fact doing the opposite.

Finally, imposing such tariffs on imports will lead to responses from the countries that had been providing the goods.  Under the agreed rules of international trade, those countries can then impose commensurate tariffs of their own on products they had been importing from the US.  This will harm industries that may otherwise have been totally innocent in whatever was behind the dispute.

An example of what can then happen has been the impact on Harley-Davidson, the American manufacturer of heavy motorcycles (affectionately referred to as “hogs”).  Harley-Davidson is facing what has been described as a “triple whammy” from Trump’s trade decisions.  First, they are facing higher steel (and aluminum) prices for their production in the US, due to the Trump steel and aluminum tariffs.  Harley estimates this will add $20 million to their costs in their US plants.  For a medium-sized company, this is significant.  As of the end of 2017, Harley-Davidson had 5,200 employees in the US (see page 7 of this SEC filing).  With $20 million, they could pay each of their workers $3,850 more.  This is not a small amount.  Instead, the funds will go to bolster the profits of steel and aluminum firms.

Second, the EU has responded to the Trump tariffs on their steel and aluminum by imposing tariffs of their own on US motorcycle imports.  This would add $45 million in costs (or $2,200 per motorcycle) should Harley-Davidson continue to export motorcycles from the US to the EU.  Quite rationally, Harley-Davidson responded that they will now need to shift what had been US production to one of their plants located abroad, to avoid both the higher costs resulting from the new steel and aluminum tariffs, and from the EU tariffs imposed in response.

And one can add thirdly and from earlier, that Trump pulled the US out of the already negotiated (but still to be signed) Trans-Pacific Partnership agreement.  This agreement would have allowed Harley-Davidson to export their US built motorcycles to much of Asia duty-free.  They will now instead be facing high tariffs to sell to those markets.  As a result, Harley-Davidson has had to set up a new plant in Asia (in Thailand), shifting there what had been US jobs.

Trump reacted angrily to Harley-Davidson’s response to his trade policies.  He threatened that “they will be taxed like never before!”.  Yet what Harley-Davidson is doing should not have been a surprise, had any thought been given to what would happen once Trump started imposing tariffs on essential inputs needed in the manufacture of motorcycles (steel and aluminum), coming from our major trade partners (and often closest allies).  And it is positively scary that a president should even think that he should use the powers of the state to threaten an individual private company in this way.  Today it is Harley-Davidson.  Who will it be tomorrow?

There are many other examples of the problems that have already been created by Trump’s new tariffs.  To cite a few, and just briefly:

a)  The National Association of Home Builders estimated that the 20% tariff imposed in 2017 on imports of softwood lumber from Canada added nearly $3,600 to the cost of building an average single-family home in the US and would, over the course of a year, reduce wages of US workers by $500 million and cost 8,200 full-time US jobs.

b)  The largest nail manufacturer in the US said in late June that it has already had to lay off 12% of its workforce due to the new steel tariffs, and that unless it is granted a waiver, it would either have to relocate to Mexico or shut down by September.

c)  As of early June, Reuters estimated that at least $2.5 billion worth of investments in new utility-scale solar installation projects had been canceled or frozen due to the tariffs Trump imposed on the import of solar panel assemblies.  This is far greater than new investments planned for the assembly of such panels in the US.  Furthermore, the jobs involved in such assembly work are generally low-skill and repetitive, and can be automated should wages rise.

So there are consequences from such tariffs.  They might be unintended, and possibly not foreseen, but they are real.

But would the imposition of tariffs necessarily reduce the trade deficit, as Trump evidently believes?  No.  As noted above, the trade deficit would only fall if the tariffs would, for some reason, increase domestic savings or reduce domestic investment.  But tariffs do not enter directly into those factors.  Indirectly, one could map out some chains of possible causation, but these changes in some set of tariffs (even if broadly applied to a wide range of imports) would not have a major effect on overall domestic savings or investment.  They could indeed even act in the opposite direction.

Households, to start, will face higher prices from the new tariffs.  To try to maintain their previous standard of living (in real terms) they would then need to spend more on what they consume and hence would save less.  This, by itself, would reduce domestic savings and hence would increase the trade deficit to the extent there was any impact.

The impacts on firms are more various, and depend on whether the firm will be a net winner or loser from the government actions and how they might then respond.  If a net winner, they have been able to raise their prices and hence increase their profits.  If they then save the extra profits (retained earnings), domestic savings would rise and the trade deficit would fall.  But if they increase their investments in what has now become a more profitable activity (and that is indeed the stated intention behind imposing the tariffs), that response would lead to an increase in the trade deficit.  The net effect will depend on whether their savings or their investment increases by more, and one does not know what that net change might be.  Different firms will likely respond differently.

One also has to examine the responses of the firms who will be the net losers from the newly imposed tariffs.  They will be paying more on their inputs and will see a reduction in their profits.  They will then save less and will likely invest less.  Again, the net impact on the trade deficit is not clear.

The overall impact on the trade deficit from these indirect effects is therefore uncertain, as one has effects that will act in opposing directions.  In part for this reason, but also because the tariffs will affect only certain industries and with responses that are likely to be limited (as a tariff increase today can be just as easily reversed tomorrow), the overall impact on the trade balance from such indirect effects are likely to be minor.

Increases in individual tariffs, such as those being imposed now by Trump, will not then have a significant impact on the overall trade balance.  But tariffs still do matter.  They change the mix of what is produced, from where items will be imported, and from where items will be produced for export (as the Harley-Davidson case shows).  They will create individual winners and losers, and hence it is not surprising to see the political lobbying as has grown in Washington under Trump.  Far from “draining the swamp”, Trump’s trade policy has made it critical for firms to step up their lobbying activities.

But such tariffs do not determine what the overall trade balance will be.

D.  What Role Does Foreign Investment Play in the Determination of the Trade Balance?

While tariffs will not have a significant effect on the overall trade balance, foreign investment (into the US) will.  To see this, we need to return to the basic macro balance derived in Section B above, but generalize it a bit to include all foreign financial flows.

The trade balance is the balance between exports and imports.  It is useful to generalize this to take into account two other sources of current flows in the national income and product accounts which add to (or reduce) the net demand for foreign exchange.  Specifically, there will be foreign exchange earned by US nationals working abroad plus that earned by US nationals on investments they have made abroad.  Economists call this “factor services income”, or simply factor income, as labor and capital are referred to as factors of production.  This is then netted against such income earned in the US by foreign nationals either working here or on their investments here.  Second, there will be unrequited transfers of funds, such as by households to their relatives abroad, or by charities, or under government aid programs.  Again, this will be netted against the similar transfers to the US.

Adding the net flows from these to the trade balance will yield what economists call the “current account balance”.  It is a measure of the net demand for dollars (if positive) or for foreign exchange (if a deficit) from current flows.  To put some numbers on this, the US had a foreign trade deficit of $571.6 billion in 2017.  This was the balance between the exports and imports of goods and services (what economists call non-factor services to be more precise, now that we are distinguishing factor services from non-factor services).  It was negative – a deficit.  But the US also had a surplus in 2017 from net factor services income flows of $216.8 billion, and a deficit of $130.2 billion on net transfers (mostly from households sending funds abroad).  The balance on current account is the sum of these (with deficits as negatives and surpluses as positives) and came to a deficit of $485.0 billion in 2017, or 2.5% of GDP.  As a share of GDP, this deficit is significant but not huge.  The UK had a current account deficit of 4.1% of GDP in 2017 for example, while Canada had a deficit of 3.0%.

The current account for foreign transactions, basically a generalization of the trade balance, is significant as it will be the mirror image of the capital account for foreign transactions.  That is, when the US had a current account deficit of $485.0 billion (as in 2017), there had to be a capital account surplus of $485.0 billion to match this, as the overall purchases and sales of dollars in foreign exchange transactions will have to balance out, i.e. sum to zero.  The capital account incorporates all transactions for the purchase or sale of capital assets (investments) by foreign entities into the US, net of the similar purchase or sale of capital assets by US entities abroad.  When the capital account is a net positive (as has been the case for the US in recent decades), there is more such investment going into the US than is going out.  The investments can be into any capital assets, including equity shares in companies, or real estate, or US Treasury or other bonds, and so on.

But while the two (the current account and the capital account) have to balance out, there is an open question of what drives what.  Look at this from the perspective of a foreigner, wishing to invest in some US asset.  They need to get the dollars for this from somewhere.  While this would be done by means of the foreign exchange markets, which are extremely active (with trillions of dollars worth of currencies being exchanged daily), a capital account surplus of $485 billion (as in 2017) means that foreign entities had to obtain, over the course of the year, a net of $485 billion in dollars for their investments into the US.  The only way this could be done is by the US importing that much more than it exported over the course of the year.  That is, the US would need to run a current account deficit of that amount for the US to have received such investment.

If there is an imbalance between the two (the current account and the capital account), one should expect that the excess supply or demand for dollars will lead to changes in a number of prices, most directly foreign exchange rates, but also interest rates and other asset prices.  These will be complex and we will not go into here all the interactions one might then have.  Rather, the point to note is that a current account deficit, even if seemingly large, is not a sign of disequilibrium when there is a desire on the part of foreign investors to invest a similar amount in US markets.  And US markets have traditionally been a good place to invest.  The US is a large economy, with markets for assets that are deep and active, and these markets have normally been (with a few exceptions) relatively well regulated.

Foreign nationals and firms thus have good reason to invest a share of their assets in the US markets.  And the US has welcomed this, as all countries do.  But the only way they can obtain the dollars to make these investments is for the US to run a current account deficit.  Thus a current account deficit should not necessarily be taken as a sign of weakness, as Trump evidently does.  Depending on what governments are doing in their market interventions, a current account deficit might rather be a sign of foreign entities being eager to invest in the country.  And that is a good sign, not a bad one.

E.  An “Exorbitant Privilege”

The dollar (and hence the US) has a further, and important, advantage.  It is the world’s dominant currency, with most trade contracts (between all countries, not simply between some country and the US) denominated in dollars, as are contracts for most internationally traded commodities (such as oil).  And as noted above, investments in the US are particularly advantageous due to the depth and liquidity of our asset markets.  For these reasons, foreign countries hold most of their international reserves in dollar assets.  And most of these are held in what have been safe, but low yielding, short-term US Treasury bills.

As noted in Section D above, those seeking to make investments in dollar assets can obtain the dollars required only if the US runs a current account deficit.  This is as true for assets held in dollars as part of a country’s international reserves as for any other investments in US dollar assets.  Valéry Giscard d’Estaing in the 1960s, then the Minister of Finance of France, described this as an “exorbitant privilege” for the US (although this is often mistakenly attributed Charles de Gaulle, then his boss as president of France).

And it certainly is a privilege.  With the role of the dollar as the preferred reserve currency for countries around the world, the US is able to run current account deficits indefinitely, obtaining real goods and services from those countries while providing pieces of paper generating only a low yield in return.  Indeed, in recent years the rate of return on short-term US Treasury bills has generally been negative in real terms (i.e. after inflation).  The foreign governments buying these US Treasury bills are helping to cover part of our budget deficits, and are receiving little to nothing in return.

So is the US a “piggybank that everybody is robbing”, as Trump asserted to necessarily be the case when the US is has a current account deficit?  Not at all.  Indeed, it is the precise opposite.  The current account deficit is the mirror image of the foreign investment inflows coming into the US.  To obtain the dollars needed to do this those countries must export more real goods to the US than they import from the US.  The US gains real resources (the net exports), while the foreign entities then invest in US markets.  And for governments obtaining dollars to hold as their international reserves, those investments are primarily in the highly liquid and safe, short-term US Treasury bills, despite those assets earning low or even negative returns.  This truly is an “exorbitant privilege”, not a piggybank being robbed.

Indeed, the real concern is that with the mismanagement of our budget (tax cuts increasing deficits at a time when deficits should be reduced) plus the return to an ideologically driven belief in deregulating banks and other financial markets (such as what led to the financial and then economic collapse of 2008), the dollar may lose its position as the place to hold international reserves.  The British pound had this position in the 1800s and then lost it to the dollar due to the financial stresses of World War I.  The dollar has had the lead position since.  But others would like it, most openly by China and more quietly Europeans hoping for such a role for the euro.  They would very much like to enjoy this “exorbitant privilege”, along with the current account deficits that privilege conveys.

F.  Summary and Conclusion

Trump’s beliefs on the foreign trade deficit, on the impact of hiking tariffs, and on who will “win” in a trade war, are terribly confused.  While one should not necessarily expect a president to understand basic economics, one should expect that a president would appoint and listen to advisors who do.  But Trump has not.

To sum up some of the key points:

a)  The foreign trade balance will always equal the difference between domestic savings and domestic investment.  Or with government accounts split out, the trade balance will equal the difference between domestic private savings and domestic private investment, plus the government budget balance.  The foreign trade balance will only move up or down when there is a change in the balance between domestic savings and domestic investment.

b)  One way to change that balance would be for the government budget balance to increase (i.e. for the government deficit to be reduced).  Yet Trump and the Republican Congress have done the precise opposite.  The massive tax cuts of last December, plus (to a lesser extent) the increase in government spending now budgeted (primarily for the military), will increase the budget deficit to record levels for an economy in peacetime at full employment.  This will lead to a bigger trade deficit, not a smaller one.

c)  One could also reduce the trade deficit by making the US a terrible place to invest in.  This would reduce foreign investment into the US, and hence the current account deficit.  In terms of the basic savings/investment balance, it would reduce domestic investment (whether driven by foreign investors or domestic ones).  If domestic savings was not then also reduced (a big if, and dependant on what was done to make the US a terrible place to invest in), this would lead to a similar reduction in the trade deficit.  This is of course not to be taken seriously, but rather illustrates that there are tradeoffs.  One should not simplistically assume that a lower trade deficit achieved by any means possible is good.

d)  It is also not at all clear that one should be overly concerned about the size of the trade and current account deficits, at where they are today.  The US had a trade deficit of 2.9% of GDP in 2017 and a current account deficit of 2.5% of GDP.  While significant, these are not huge.  Should they become much larger (due, for example, to the forecast increases in government budget deficits to record levels), they might rise to problematic levels.  But at the current levels for the current account deficit, we have seen the markets for foreign exchange and for interest rates functioning pretty well and without overt signs of concern.  The dollars being made available through the current account deficit have been bought up and used for investments in US markets.

e)  Part of the demand for dollars to be invested and held in the US markets comes from the need for international reserves by governments around the world.  The dollar is the dominant currency in the world, and with the depth and liquidity of the US markets (in particular for short-term US Treasury bills) most of these international reserves are held in dollars.  This has given the US what has been called an “exorbitant privilege”, and permits the US to run substantial current account deficits while providing in return what are in essence paper assets yielding just low (or even negative) returns.

f)  The real concern should not be with the consequences of the dollar playing such a role in the system of international trade, but rather with whether the dollar will lose this privileged status.  Other countries have certainly sought this, most openly by China but also more quietly for the euro, but so far the dollar has remained dominant.  But there are increasing concerns that with the mismanagement of the government budget (the recent tax cuts) plus ideologically driven deregulation of banks and the financial markets (as led to the 2008 financial collapse), countries will decide to shift their international reserves out of the dollar towards some alternative.

g)  What will not reduce the overall trade deficit, however, is selective increases in tariff rates, as Trump has started to do.  Such tariff increases will shift around the mix of countries from where the imports will come, and/or the mix of products being imported, but can only reduce the overall trade deficit to the extent such tariffs would lead somehow to either higher domestic savings and/or lower domestic investment.  Tariffs will not have a direct effect on such balances, and indirect effects are going to be small and indeed possibly in the wrong direction (if the aim is to reduce the deficits).

h)  What such tariff policies will do, however, is create a mess.  And they already have, as the Harley-Davidson case illustrates.  Tariffs increase costs for US producers, and they will respond as best they can.  While the higher costs will possibly benefit certain companies, they will harm those using the products unless some government bureaucrat grants them a special exemption.

But what this does lead to is officials in government picking winners and losers.  That is a concern.  And it is positively scary to have a president lashing out and threatening individual firms, such as Harley-Davidson, when the firms respond to the mess created as one should have expected.