Author Archives: aneconomicsense

Unknown's avatar

About aneconomicsense

Economist

The Performance of the Stock Market During Trump’s Term in Office: Not So Special

Posted on by

A.  Introduction

Stock market performance is often taken to be a good measure of how the economy as a whole is performing.  But it is not.  For most Americans it is simply irrelevant, as the overwhelming share of investments in the stock markets are held by only a small segment of the population (the wealthy).  And its track record as a broader indicator of how the economy is performing is imperfect at best.

Still, many do focus on stock market returns, and Trump brags that the performance of the market during his term in office has been spectacular.

That is not the case.  This post will look at how the stock market has performed during Trump’s term in office thus far, and compare it to what that performance was under presidents going back to Reagan up to the same point in their terms.

First, however, we will briefly discuss to what extent one should expect stock market prices to reflect actions a president might be taking.  And the answer is some, but there is much more going on.

B.  Presidential Policies and the Stock Market

Owning shares of a firm entitles the owner to a share of the profits generated by that firm, both now and into the future.  And while there are many complications, a simple metric commonly used to assess the price of a share in a firm, is the price/earnings ratio.  If earnings (profits) go up, now and into the future, then for a given price/earnings ratio the price of the stock would go up in proportion.

Economic policies affect profits.  And in a thriving economy, profits will also be rising.  The policies of a presidential administration will affect this, and although the link is far from a tight one (with important lags as well), policies that are good for the economy as a whole will generally also lead to a rising stock market.

But there is also a more specific link to policy.  What accrues to the shareholders are not overall profits, but profits after taxes.  And this changed significantly as a result of the new tax law pushed through Congress by Trump and the Republicans in December 2017.  It resulted in the effective corporate profits (income) tax being cut by more than half:

This chart is an update of one prepared for an earlier post on this blog (where one can see a further discussion of what lies behind it).  It shows corporate profit taxes at the federal level as a share of corporate profits (calculated from figures in the national income accounts issued by the BEA).  While Trump and the Republicans in Congress asserted the 2017 tax bill would not lead to lower corporate profit taxes being paid (as loopholes would be closed, they asserted), in fact they did.  And dramatically so, with the effective corporate tax rate being slashed by more than half –  from around 15 to 16% prior to 2017, to just 7% or so since the beginning of 2018 (and to just 6.3% most recently).

This cut therefore led to a significant increase in after-tax profits for any given level of before-tax profits, which has accrued to the shareholders.  Note that this would not be due to the corporations becoming more productive or efficient, but rather simply from taxing profits less and shifting the tax burden then on to others (i.e. a redistributive effect).  And based on a reduction in the taxes from 16% of corporate profits to 7%, after-tax profits would have gone from 84% of profits to 93%, an increase of about 11%.  For any given price/earnings ratio, one would then expect stock prices, for this reason alone, to have gone up by about 11%.

[Side note:  Technically one should include in this calculation also the impact of taxes on profits by other government entities – primarily those of state and local governments.  These have been flat at around 3 1/2% of profits, on average.  With these taxes included, after-tax profits rose from 80 1/2% of before-tax profits to 89 1/2%, an increase that is still 11% within round-off.]

One should therefore expect that stock prices following this tax cut (or in anticipation of it) would have been bumped up by an additional 11% above what they otherwise would have been.  Other things equal, the performance of the stock market under Trump should have looked especially good as a result of the shift in taxes away from corporations onto others.  But what has in fact happened?

C.  Trump vs. Obama

The chart at the top of this post compares the performance of the stock market during Trump’s term in office thus far (through December 31, 2019) to that under Obama to the same point in his first term in office.  The difference is clear.  Other than during Obama’s first few months in office, when he inherited from George W. Bush an economy in freefall, stock market performance under Obama was always better than it has been under Trump.  Even after slashing corporate profit taxes by more than half, the stock market under Trump did not do exceptionally well.

The S&P500 Index is being used as the measure of the US stock market.  Most professionals use this index as the best indicator of overall stock market performance, as it is comprehensive and broad (covering the 500 largest US companies as measured by stock market value, with the companies weighted in the index based on their market valuations).  The data were downloaded from Yahoo Finance, where it is conveniently available (with daily values for the index going back to 1927), but can be obtained from a number of sources.  The chart shows end-of-month figures, starting from December 31 of the month before inauguration, and going through to December 31 of their third year in office.  The index is scaled to 100.0 on exactly January 20 (with this presented as “month” 0.65).

So if one wants to claim “bragging rights” for which president saw a better stock market performance, Obama wins over Trump, at least so far in their respective terms.

D.  Trump vs. All Presidents Since Reagan

A comparison to just one president is limited.  How does the performance under Trump compare to that under other US presidents up to the same points in their terms in office?  Trump is roughly in the middle:

This chart tracks the performance under each president since Reagan up through the third year of their first terms in office.  I have adjusted here for inflation (using the CPI), as inflation was substantially higher during the Reagan and Bush Sr. terms in office than it has been since.  (I left the chart at the top of this post of just Obama vs. Trump in nominal terms as inflation in recent years has been steady and low.  But for those interested in the impact of this, one can see the Obama and Trump numbers in real terms in the current chart.)  I have included in this chart only the first terms of each president (with one exception) as the chart is already cluttered and was even more so when I had all the presidential terms.

The exception is that I included for perspective the stock market performance during Clinton’s second term in office.  The stock market rose over that period by close to 80% in real terms, which was substantially higher than under any other president since at least before Reagan in either their first or second terms.  The performance in Obama’s first term (of 146% in real terms) was the second-highest.  There was then a set of cases which, at the three-year mark, showed surprising uniformity in performance, with increases of between 32% and 34% in the second Reagan term, the first Clinton term, the second Obama term, and Trump’s term so far.  Bush Sr. was not far behind this set with an increase of 28%.

The worst performances were under Bush Jr. ( a fall of 22% to the third-year point in his first term), and Reagan (an increase of just 8% to that point in his first term).

So the performance of the market under Trump is in the middle – not the worst, but well below the best.

E.  Single Year Increases in the S&P500 from 1946 to 2019

Finally, was the increase under Trump in his best single year so far (2019) a record?  No, it was not.  Looking at the single year performances (in real terms) since 1946, the top 15 were:

The increase in 2019, of 25.9%, was good, but only the sixth-highest of the 74 years between 1946 and 2019 (inclusive).  The stock market rose by more in 2013 during Obama’s term in office (by 27.7%), and in 1997 (28.8%) and 1995 (30.8%) which were both Clinton years.  And the highest increases were in 1958 (35.7%) and 1954 (45.6%) when Eisenhower was president.

The market also rose substantially in 2017, in Trump’s first year in office, by 16.9%.  But it then fell by 8.0% in 2018, in Trump’s second year in office.  Overall, the average rank (out of the 74 years from 1946 to 2019) of the individual year performances over the three years Trump has been in office so far, would place Trump in the middle third.  Not the worst, but also far from the best.  And comparing the three-year average while Trump has been president to rolling three-year averages since 1946, Trump’s average (of 11.6%) is well below the best.  The highest was an average return of 25.3% in 1995-97 during Clinton’s term in office.  And the three-year average return was also higher at 16.7% in 2012-14 during Obama’s term.

F.  Summary and Conclusion

Trump likes to brag that the performance of the stock market during his term in office has been exceptional.  But despite a slashing of corporate profit taxes (which, other things being equal would be expected to increase stock prices by 11%), the performance of the market during Trump’s term in office would put him in the middle.  Specifically:

a)  The market rose by more during the first three years of Obama’s term in office than it has under Trump;

b)  Compared to the first three years in office of all presidents since Reagan (whether first terms only, or first and second terms) would place Trump in the middle.  Indeed, the increase under Trump so far was almost exactly the same as the increases seen (at the three-year point) in Obama’s second term, in Reagan’s second term, and in Clinton’s first term.  And the return under Trump was well below that seen in Obama’s first term, and especially far below that in Clinton’s second term.

c)  The individual year performances during Trump’s three years have also not been exceptional.  While the performance in 2019 was good, it was below that of a number of other years since World War II, and below that of individual years during Obama’s and Clinton’s terms in office.

But as noted at the start of this post, stock market returns should not be over-emphasized.  An increase in the stock market does little for those who do not have the wealth to have substantial holdings in the stock market, and as a broader indicator of how the overall economy is performing, stock market returns are imperfect at best.

Still, one should be accurate in one’s claims.  And as on many things, Trump has not been.

The High and Rising Cost of Health Care in the US

Posted on by

A.  Introduction

What to do about health care has been a central issue debated by the candidates seeking the Democratic nomination for the presidency.  Different proposals have been made, but two underlying factors drive the issue.  One is that health care costs in the US are high – far more than what they were not so long ago, and far higher than what health care costs in other countries.  The second and related issue is that despite such a high amount being spent, and despite as well a marked improvement under the Obamacare reforms, the US still has a substantial share of its population who are uninsured and hence suffer from a lack of effective access to the health care system.

This post will look at those issues at a macro level, basically through a series of charts.  I have been working on a post that examines the Medicare-for-All proposals, where the plans now issued by Elizabeth Warren are the most specific and detailed.  But it is useful first to set the context by reviewing the high costs for the US (while yielding only mediocre health outcomes), as medical care and how to pay for it would not be the prominent, and difficult, issues that they now are if we were not spending so much.

B.  Health Care Expenditures as a Share of GDP

A commonly used measure, which allows comparisons both over time and between countries, is the amount being spent on health care as a share of GDP.  The chart at the top of this post shows this for the US, with national health expenditures (whether from private sources or from public, and for investment as well as for current consumption of the services) going back to 1960.

The figures on national health expenditures are as published by the Centers for Medicare and Medicaid Services (CMS), which provides each year authoritative and detailed figures on such expenditures, both historical and projected.  The most recent historical figures cover the years through 2017, while the accompanying forecasts (for 2018 to 2027) were issued in February 2019.  The figures used in the chart above for 1960 to 2017 come from the historical tables, while the 2018 and 2019 figures come from the first two years of the 2018 to 2017 forecasts.  The GDP figures come from the Bureau of Economic Analysis (BEA), where the 2019 figure is based on a simple extrapolation of the current estimates for the first three quarters of the year.

It was not so long ago that the US spent far less on health care than it does now.  In 1960 it was only 5.0% of GDP.  But this then rose, fairly steadily, to 8.9% in 1980, 13.4% in 2000, and close to 18% of GDP now.  Note that by examining this as a share of GDP one is taking into account that as a country grows richer, it can spend more on health care (as well as on everything else) with a share that is unchanged.  But in the US the share itself has increased dramatically.

Note also that as a share of GDP one is taking a ratio of a nominal magnitude (what is being spent on health care, in current dollars) to a nominal magnitude (overall GDP, also in current dollars).  There is therefore no issue of whether the price indices for health care services are being measured correctly (and in particular whether they are reflecting quality changes correctly), as price indices do not enter.  Rather, one is simply adding up all that is being spent on health care services, and comparing this to a broad concept of national income and output.

But while this chart is an accurate portrayal of how much is being spent on health care as a share of actual GDP in any period, fluctuations in the curve could be due either to changes in health care spending from one year to the next, or due to changes in GDP from one year to the next.  One can see that there were particularly sharp upward movements in the share in those years when the economy went into recession (such as in the early years of the Reagan presidency, again when Bush Sr. was president, again in the early years of the Bush Jr. presidency, and then especially sharply in his last year in office).  The curve flattened out in the years of strong economic growth (such as during most of the Clinton presidency).

One can control for these business cycle fluctuations in GDP by calculating the health care expenditures as a share of potential GDP rather than as a share of actual GDP:

Potential GDP is an estimate of what GDP would have been in any given year, had employment been at full employment levels.  The estimates of potential GDP used here come from the Congressional Budget Office, which comes up with careful estimates of potential GDP as part of its budget work.

Potential GDP grows at a steady pace, unlike the business cycle fluctuations of actual GDP.  And with this, one does not see the degree of fluctuations in the curve for health care spending as one has with actual GDP.  What this implies is that health care spending grows at a relatively steady rate, whether the economy is growing well or has gone into a downturn.  The rise in the share during economic downturns, when measured in terms of actual GDP, thus is primarily due to the reduction in those years of the denominator (GDP), rather than a sudden rise in the numerator (health care spending).

This should not be surprising, except to those who think consumers treat health care spending as a discretionary expense which they can control.  That is, conservative analysts (and Republican congressmen) have advocated for health funding systems where the patients face a high share of their health care costs (and a 100% share in high deductible health insurance plans, except in catastrophes).  The presumption is that patients have a good deal of choice in whether to seek health treatments or not.  If this were the case, one would find health care spending falling along with GDP when the economy goes into a downturn, due to higher unemployment in such times (with the unemployed losing their previous health insurance cover) and money in general being tight.  But one does not see this.  Rather, health care is a necessity, which one needs regardless of the state of the economy.

There are still some modest fluctuations in the curve, coinciding with periods where administrations have focused on health care costs.  Thus one sees a flattening of the curve (relative to the overall trend) during the Clinton years (1993 to 2000), and again after the Obamacare reforms were passed in 2010.  But overall, the curve has steadily risen, from 5% of GDP in 1960 to 18% now.  This increase, by a factor of 3.6, is extraordinary.

C.  US Health Care Spending and Outcomes Compared to Other Countries

Not only has US health care spending increased by an extraordinary amount over the last several decades, but it is also far higher than what other countries spend:

 

This chart, as well as those immediately below with other cross-country comparisons, are updated versions of several from an earlier post on this blog.  That earlier post also has additional comparisons and detail which readers may find of interest, but have not been replicated here due to space.  But they still apply.

The earlier charts were for 2011, while these are now for 2017.  Both sets are based on health care data assembled by the OECD for its member countries.  Standard definitions are set by the OECD to allow cross-country comparisons.

As the chart above shows, the US still stands out, as before, by spending far more for health care (as a share of GDP) than any other country.  Note that the figures here only include expenditures on current health care services.  The OECD had earlier included also health care expenditures for investments (such as for hospital buildings, for research, and so on), but now places investment expenditures in a separate, and only partial, table.  While the investment figures are readily available for the US, it appears they are difficult to obtain for a number of others.  Hence the OECD now treats it separately, in a table with a large number of blanks.

But the US spending on current health care services, at 17.1% of GDP, is still far ahead of such spending by the second-ranking country – Switzerland here.  And the US spends close to double the OECD average.

Such exceptionally high spending does not, however, yield exceptionally good health outcomes.  Indeed, by a number of standard measures US health outcomes are among the worst in the OECD, where the only countries that are worse are those with a far lower income than what the US enjoys.

First, one can look at infant mortality rates:

Infant mortality rate, 2017, US compared to other OECD member countries

Only Mexico, Turkey, and Chile have a higher infant mortality rate than the US does.  And there is a huge room for improvement.  The infant mortality rate in Japan is only one-third that of the US.  Yet Japan spends less than 11% of GDP on health care services, compared to the more than 17% of GDP in the US.

Life expectancy is also a standard measure of health care outcomes:

Only a number of countries from Central Europe. as well as Turkey and Mexico, have a lower life expectancy than that of the US.  And they all have far lower incomes than the US.

Finally, a measure more commonly used by health care professionals takes into account how death rates vary by age.  That is, given the age profile at which deaths occur in a country, the “potential years of life lost” measure multiplies the number of deaths that occur at any given age by the difference between that age and a reference age (where the OECD uses age 70 for this benchmark).  It then takes the sum across all ages, and computes this per 100,000 of population.  Thus a death at age 50 will receive twice the weight of a death at age 60, as a person dying of some disease at age 50 would have had 20 more years to live to age 70, while a person dying at age 60 would have had only 10 more years to live to age 70.

The US once again comes out poorly, even by this more sophisticated measure:

Only several countries of Central Europe, plus Mexico, are worse than the US.  Even Turkey is better.  All the countries of Western Europe, as well as Japan, Korea, Australia, and Canada, are far better.

D.  The Impact of the Uninsured

At least part of the reason for these poor health outcomes in the US compared to the high-income countries of Europe and Asia is that the US, despite spending more, has a high share of its population left uninsured.  All those other high-income countries have universal health insurance cover (with the partial exception of Belgium, where coverage is 98.7%):

US health insurance coverage (under the standardized definition of the OECD) is only at 90.6%.  Only Mexico, among OECD members, is worse.

Taking this into account, the US spending figures are even worse than they at first appear.  Not only does the US spend on health care services a far higher share of GDP than other OECD member countries do, but that higher spending is concentrated on less than the full population due to the high share of uninsured.  The uninsured do, of course, have certain health care costs, which they pay for, to the extent they can, out of pocket, by charity, or for certain services under various government programs (such as for public health) that cover everyone.  But overall, the uninsured are more limited in what they can obtain in health care services than what an insured person can.

With the Obamacare reforms, the share of the population in the US without health insurance fell significantly, for the first time in a generation.  While the Trump administration has done all it legally can to reverse these gains, the impact so far (up to 2018, where note the figures here are annual averages) has been only partial:

The figures come from the American Community Survey (ACS) of the US Census Bureau, by way of a Kaiser Family Foundation analysis of the ACS results (and with 2018 based on figures in the 2018 Census Bureau report on Health Insurance Coverage).

With Obamacare (more formally, the Affordable Care Act), the share of the US population without any health insurance fell from 15% in 2010 to 9% in 2015 and 8.3% in 2016.  The Affordable Care Act was passed in 2010, and while certain of the reforms went soon into effect (such as the requirement, effective in 2011, that sons and daughters up to age 26 could remain on their family health insurance plans), the primary changes entered into effect in 2014, with the opening of the Obamacare market exchanges and the Medicaid expansion.

With this significant reduction in the number uninsured, an interesting question is whether this might account for some or all of the increase in observed national health expenditures as a share of GDP in the years of that expansion in coverage.  This chart will help address this:

The lower curve in the chart, in blue, shows national health expenditures as a share of GDP, and is the same, and from the same sources, as in the chart at the top of this post.  The only difference is that the focus now is only on the period since 2008.  The curve was basically flat (indeed falling a bit) from 2009 to 2013.  It then rose from 2014 to 2016, which is the period that coincides with the substantial reduction in the share of the population that did not have health insurance.

An analogous curve for just the insured is then shown as the top curve in the chart, in red, and plots what national health expenditures were per insured person in the US, with this then taken as a share of GDP per capita.  Note that national health expenditures as a share of GDP (the blue curve) is the same as national health expenditures per person, with this then taken as a share of GDP per capita (i.e. with both numerator and denominator divided by the population).  The sources are the previously cited Kaiser Family Foundation report and ACS for the number uninsured, the Census Bureau for population, and the BEA for GDP.

If 100% of the population were insured, the two curves would coincide.  As the share who were uninsured fell following the Obamacare reforms, the two curves approached each other.  And it is interesting that the costs per insured person fell after 2010, with an especially sharp fall in 2014 and a smaller reduction in 2015.

But as was noted previously, to see the underlying trends one should remove the impact of the cyclical fluctuations of GDP, by calculating the shares in terms of potential GDP.  When one does this one finds:

National health care costs per insured person (as a share of potential GDP per capita) fell in 2014, the year the number of uninsured fell sharply, and in 2015 was still below earlier levels.  But it has risen fairly steadily from 2015 onwards.  The increase in national health care costs (as a share of GDP) in 2014 and 2015 therefore can be attributed to the expanded coverage achieved in those years.  But after 2015, the steady and similar increases in both curves suggest a return to the earlier trends of rising health care costs.

E.  Conclusion

The US spends an extraordinary amount on health care.  No other country comes close.  Furthermore, what the US spends, as a share of its GDP, has risen dramatically over recent decades, from just 5% of GDP in 1960 to almost 18% of GDP now.  And while other developed countries have been able to attain universal health insurance coverage despite spending far less, the US has not.  The Obamacare reforms helped, but the Trump administration is trying to do all it legally can to reverse those achievements.

It is these high costs that are basically driving the need for fundamental reform in the US health care funding system.  The current path is not sustainable.  But whether reforms will be enacted to reverse those cost trends, or at least flatten them out, remains to be seen.

How Fast is GDP Growing?: A Curiosum

Posted on by

A.  How Fast is GDP Growing?

The Bureau of Economic Analysis released today its first estimate (what it calls it’s Advance Estimate) for the growth of GDP and its components for the third quarter of 2019.  Most of it looked basically as one would expect, with an estimate of real GDP growth of 1.9% in the quarter, or about the same as the 2.0% growth rate of the second quarter.  There has been a continued slowdown in private investment (which I will discuss below), but this has been offset by an expansion in government spending under Trump, coupled with steady growth in personal consumption expenditures (as one would expect with an economy now at full employment).

But there was a surprise on the last page of the report, in Appendix Table A.  This table provides growth rates of some miscellaneous aggregates that contribute to GDP growth, as well as their contribution to overall GDP growth.  One line shown is for “motor vehicle output”.  What is surprising is that the growth rate shown, at an annualized rate, is an astounding 32.6%!  The table also indicates that real GDP excluding motor vehicle output would have grown at just 1.2% in the quarter.  (I get 1.14% using the underlying, non-rounded, numbers, but these are close.)  The difference is shown in the chart above.

Some points should be noted.  While all these figures provided by the BEA are shown at annualized growth rates, one needs to keep in mind that the underlying figures are for growth in just one quarter.  Hence the quarterly growth will be roughly one-quarter of the annual rate, plus the effects of compounding.  For the motor vehicle output numbers, the estimated growth in the quarter was 7.3%, which if compounded over four quarters would yield the 32.6% annualized rate.  One should also note that the quarterly output figures of this sector are quite volatile historically, and while there has not been a change as large as the 32.6% since 2009/10 (at the time of the economic downturn and recovery) there have been a few quarters when it was in the 20s.

But what appears especially odd, but also possibly interesting to those trying to understand how the GDP accounts are estimated, is why there should have been such a tremendously high growth in the sector, of 32.6%, when the workers at General Motors were on strike for half of September (starting on September 15).  GM is the largest car manufacturer in the US, its production plummeted during the strike, yet the GDP figures indicate that motor vehicle output not only soared in the quarter, but by itself raised overall GDP growth to 1.9% from a 1.2% rate had the sector been flat.

This is now speculation on my part, but I suspect the reason stems from the warning the BEA regularly provides that the initial GDP estimates that are issued just one month after the end of the quarter being covered, really are preliminary and partial.  The BEA receives data on the economy from numerous sources, and a substantial share of that data is incomplete just one month following the end of a quarter.  For motor vehicle production, I would not be surprised if the BEA might only be receiving data for two months (July and August in this case), in time for this initial estimate.  They would then estimate the third month based on past patterns and seasonality.

But because of the strike, past patterns will be misleading.  Production at GM may have been ramped up in July and August in anticipation of the strike, and a mechanical extrapolation of this into September, while normally fine, might have been especially misleading this time.

I stress that this is speculation on my part.  Revised estimates of GDP growth in the third quarter, based on more complete data, will be issued in late November and then again, with even more data, in late December.  We will see what these estimates say.  I would not be surprised if the growth figure for GDP is revised substantially downwards.

B.  Growth in Nonresidential Private Fixed Investment

The figures released by the BEA today also include its estimates for private fixed investment.  The nonresidential portion of this is basically business investment, and it is interesting to track what it has been doing over the last few years.  The argument made for the Trump/Republican tax cuts pushed through Congress in December 2017 were that they would spur business investment.  Corporate profit taxes were basically cut in half.

But the figures show no spur in business investment following their taxes being slashed.  Nonresidential private fixed investment was growing at a relatively high rate already in the fourth quarter of 2017 (similar to rates seen between mid-2013 and mid-2014, and there even was growth of 11.2% in the second quarter of 2014).  This continued through the first half of 2018.  But growth since has fallen steadily, and is now even negative, with a decline of 3.0% in the third quarter of 2019:

There is no indication here that slashing corporate profit taxes (and other business taxes) led to greater business investment.