Random Critical Analysis

Random analysis of stuff that interests me

More silliness related to corporate profits

I was pointed to this work by Hussman through Business Insider.

The implication here is that total dissaving is not only strongly correlated with corporate profits, but is directly causative.

Although he doesn’t fully specify this methods, it’s obvious that Corporate Profits is after-tax corporate profits (including foreign profits) and I was able to approximate his results using this FRED2 link.

Update: I re-charted this using the NIPA corporate profits inventory & capital adjusted data that he clearly used (CPROFIT).  It doesn’t really change the outcome here, but it matches his chart more precisely.

Corporate profits is, in other words, after-tax and including foreign profits.

Savings is approximately personal savings (PSAVE) + the Federal deficit/surplus (FGRECPT-FGEXPND) (multiplied by -1 to match to shape of the profit line)

There are many issues with this analysis

1: By showing after-tax profits he’s exposing his analysis to changes in the tax regime and business reporting (as I’ve alluded to in prior posts).   The corporate taxes paid has direct mechanical effect on after-tax profits and the tax regime has changed.  Further, in this most recent recession we have actually cut corporate taxes in the short term by allowing businesses to accelerate deprecation more rapidly.

2: Including foreign profits in this discussion also obscures the role that deficit spending (or consumer dissaving) might be playing here.

3: Cutting it off at 1970 obscures what is obviously a much worse correlation in earlier years, i.e., when savings were much higher and corporate profits were much higher.

4: Lagging quarterly data by a full year just doesn’t make much sense to me.

Same method, i.e., 4Q lead, using PRE-tax profits

Same method over a longer time span (1947-2011)

Scatter chart savings vs after-tax

 

Scatter chart: savings vs pre-tax profits

Observe: When you remove taxation from the equation the slope and correlation virtually disappears.

After tax profits vs Corporate Profits 

It’s almost as-if corporate taxes actually has a direct impact on after-tax profits (and thus accounts for most of this apparent correlation)!  Who would have thought???

 

 

 

Note: This includes both foreign and S-Corporation profits.

 

Comparing domestic pre-tax profits with savings  (notice: pre-tax corporate profits are NOT at record highs)

Quick Scatter chart with annual domestic corporate profits vs savings (this is annual, not quarterly, since I don’t have historic quarterly data)

If anything, this would suggest that higher savings are correlated with higher profits.

I am including the CBO chart to show that:

1: the taxation of domestic corporate profits hasn’t changed that much. If anything, this under-states the recent  burden, relative to the the 80s, since a large and growing share of profits are being taxed at the individual level (S-Corporations), i.e., they are being factored in the denominator (profits) but not the numerator (corporate taxes paid).

2: my analysis of the NIPA corporate profit data matches their analysis almost exactly for the period they produced this data.

In short, the analysis conducted by BI and company is extremely misleading.  Deficit spending and dissaving may have some short term stimulative impact, but it’s ultimately a net negative.  These funds will have to be repaid in the future, both by tax payers and individual consumers, which will slow future economic growth.

Some misc. charts and data on wealth shifts





 

 

 

 

 

 

 

 

Real net worth of the lower end of the income and wealth distribution

 

 

Saez data in Excel

source for SCF data in excel

A look at college board data as it relates to affirmative action

The college board provides some summary data regarding the distribution of SAT scores by ethnicity and other measures.

Reading | Math | Writing
Black: 428 (98) | 427 (97) | 417 (94)
White: 528 (103) | 535 (102) | 516 (103)
Asian: 517 (125) | 595 (125) | 528 (127)

Mean (Standard Deviation)

With this data we can produce some quick analysis using Excel like so:

Note: This is only an estimate, assuming a normal distribution.  The left and right tails appear to be non-continuous because the SAT top codes the very highest and very lowest results (< 200 and > 800).

You might be able to tell, if you look carefully, that there are some pronounced differences to the right-hand side of the distributions.

One thing you might infer here is that if the elite universities decide to pick the top part of any distribution with preferences for various ethnic groups, i.e., to ensure that, say, 12% of the class is black, no more than 30% asian, this actually will tend to vastly amplify these differences amongst the ethnic groups within each institution.

If the “elite” schools constitute, say, 15K entering students total and the schools collectively pick the highest scoring students across these this is what it will look like (roughly).  [Yes, I know Harvard and the like will theoretically reject a non-trivial number of "perfect" scoring students, but this is largely along ethnic lines (against asians) and other better schools tend to pick them up]

Top 15K students with racial preferences

 

At 30K students

As a direct result of the preferences driving relatively poorly qualified black people into these schools there is relatively little overlap amongst black people and whites/asians.

It’s worth noting, too, that the 25th percentile of most of the elite schools (for all groups) is about 660-700 across all categories.

So what essentially happens is that more elite schools need to poach blacks and hispanics/latinos from schools several tiers below if they want to make their targets and this has a clear ripple effect through the entire system (i.e., those lower tier schools also need to reach further down to meet their targets too).

It is thus not very surprising when Richard Sander and others find very similar results when they look at law school class rank data when some handful of these students move on to law school several years later and similar race-based preferences are exercised so that the law schools can meet their own numbers.

Misc healthcare stats

 Just a collection of links and images for now so I don’t have to dredge them up later….

source

source

source

 

 

 

 

 

 

 

Source: OECD 2012 Frequently Requested Data

HEALTH STATUS, HEALTH CARE AND INEQUALITY: CANADA VS. THE U.S.

Smaller welfare state?

Someone recently told me that welfare programs are smaller than they’ve ever been.

Here is some actual data from the Whitehouse’s Office of Management and Budget (OMB): See table 3.1 – Outlays by Superfunction and Function: 1940–2017

Some data on IRS changing income concept

IRS report on changing AGI income definitions

The Tax Reform Act of 1986 (TRA 86) made extensive changes to the calculation of AGI beginning with 1987. These changes made necessary a revision of the calculation of the 1979 Income Concept, in order to make tax years beginning with 1987 comparable to the base years, 1979 through 1986. The law changes limited the deduction of passive losses and eliminated unreimbursed employee business expenses and moving expenses as “adjustments” (moving expenses changed back for 1994) in figuring AGI beginning with Tax Year 1987. Since passive losses had been fully deductible for both income measures prior to 1987, the disallowed passive losses had to be deducted in the 1979 Income Concept calculation for tax years after 1986. Some income items, such as capital gains, that had been partially excluded from AGI under prior law were fully included. The new law also eliminated or restricted some deductions. Therefore, if AGI is used to measure income, comparisons between 1986 income and tax data with that for years after 1986 are misleading. A more accurate comparison can be made using the 1979 Income Concept because it measures income in the same way for all years. Table B shows total income and selected tax items for 2009 using AGI and the 1979 Income Concept, classified by size of 2009 income. Before TRA 86 became effective, a comparison of income measured by AGI with that measured by the 1979 Income Concept showed significant differences at income levels of $200,000 or more.

But, with the elimination of preferential treatment of various income items by TRA 86, such as the exclusion of a portion of capital gains, much of the difference disappeared. Under tax law prior to 1987, the capital gains exclusion accounted for the largest difference at the higher income levels between the two income measures. For 2009, 1979 Concept income was 2.2 percent higher than income as calculated using AGI. This difference was primarily attributed to the inclusion of more than $343.4 billion in nontaxable pensions and annuities (including IRA distributions) in the 1979 Income Concept. Income for all returns, using the 1979 Income Concept, decreased 8.2 percent for 2009; income for the $200,000 and above group decreased 20.0 percent. Total income tax for all returns decreased 16.1 percent in 2009 after an decrease of 7.5 percent in 2008; and total income tax reported for the $200,000 and above income group decreased 19.3 percent for 2009, down from the 12.0 percent decrease for 2008. The average tax rates (income tax as a percentage of total income) for each income class and both income concepts for years 1986 through 2009 are shown in Figure 4. For the population as a whole, average tax rates for 2009 (based on the 1979 Income Concept) were 1.1 percentage points lower than those for 2008. Between 1986 and 2009, the average tax rates declined in all income categories except the $1 million or more category.

Average income tax rates using consistent 1979 Income Concept (direct from IRS data table)

Observation: The very top income groups are paying roughly similar taxes as they paid in 1986 when we actually use a consistent methodology like this.  Lower to middle income groups are paying substantially less and the methodology makes much less difference for them (AGI and TIC render similar results).

Comparison between AGI and Total Income Concept tax rate for selected groups

Notice how AGI and TIC lines actually converge much more closely after-tax reform.

Total income tax as percentage of AGI

Note: This is taxes paid as a percentage of the constantly changing Adjusted-Gross-Income concept that the IRS warns against.  It’s also, not coincidentally, usually the one used by advocates of going back to the supposedly golden era when high marginal tax rates presumably collected much more taxes from the very rich [despite all the evidence].

Notice how similar the high income lines look here to the data reported by this CRS chart

Although he doesn’t spell out his methods or even which taxes he’s including (payroll, CIT,…?), it’s highly likely that his analysis is simply naively dividing taxes paid by the AGI.

Some other charts produced directly by the IRS for the top 0.2%

Note: The first chart (figure I, 1951-1986) is the income tax paid as a percentage of AGI [which changed significantly, but not nearly to the same extent that it did between 86 and 87].  The second chart (figure H, 1916-1950) is “net income” — which isn’t perfectly analogous to AGI.   Nevertheless, you might observe that between the mid-40s and 1986 the average rate of the 0.2% averaged around 40% of AGI, which roughly corresponds to the more recent AGI chart above [and which, the data shows, AGI way over reported the effective tax rates in economic terms ( at least it did by 1986 )

Piketty and Saez  — see table A3 (individual + payroll)

IRS data on effective rates.  See figure E…

More IRS: see 1979 income concept, etc

IRS calculations of AETR for top 0.2% 1916-1950 and 1951-1986 @ Figure E, pg 46

United States physician income in context

One of the issues that I have when people assert that United States physician compensation is much higher than other countries is that they make terribly naive comparison.  They compare, say, PPP-adjusted incomes to PPP-adjusted incomes in other countries without accounting for the fact that the “average” person in this country has a much higher PPP-adjusted income by most measures.   Likewise, they’ll compare physician income to “average income” or “average wage” ratios without comparing it to the more relevant labor pool in each country, i.e., at least college graduates (or better). example

Average Physician Gross Income to Average College Grad Gross Income [apples-to-apples]

Note: In both cases, “gross” is pre-tax income, including social security/payroll contributions.

The same data as ratios (gross vs gross)

Average Physician Gross Income to Average College-grad LABOR cost

Note: The college labor cost data (X-axis) is pre-tax and includes the cost of benefits like healthcare, which must be baked into the costs of healthcare itself in this country to pay for employee salaries.  So this tells us something too [whereas in many other countries with socialized medicine this is payed for by (higher) taxes]…

Average Physician Gross Income to Average Net Income of College grads

Note: This is after-tax data.   The average college grads in the United States has much more disposable income available on average (not only are we paid more pre-tax (even gross), but we get to keep a much larger chunk of what we do make).   Although they we do have to pay more for education out of pocket, these costs over the course of a career are relatively small.

 

In any event, comparing how much more physicians make in proportional terms to their more equivalent peers (college graduates) is, in my opinion, a far better way to analyze the situation.   Most countries pay their physicians around 2x what a typical college grad makes.  If you think about it, this makes sense because there are real opportunity costs involved in going to medical school, going through residency programs, etc (probably even more in this country since our programs are longer and more expensive)…

 

 

Note: The data for college graduates income (various measures) can be retrieved directly from this OECD spreadsheet.

Detailed analysis of Piketty and Saez top income growth data

Recently I have been scrutinizing the data produced by Piketty and Saez on the evolution of top incomes in the United States (“Income Inequality in the United States 1913-2002″)

Their data can be downloaded directly here in Excel format that covers 1913-2012.

Their analysis is not 100% comparable to the CBO for several reason:

  1. They are analyzing family income (not households, as the CBO does, and not individual earners, and not necessarily even tax units)
  2. They do not adjust their rankings by unit size (the CBO divides household income by the square root of the household size).
  3. Their “income” is essentially identical to IRS’s AGI figures.  (Unlike the CBO et. al, they do not include health benefits, payroll taxes, corporate income taxes, etc)
  4. They also are forced to make a lot of assumptions to create a historical series stretching back this far (again, unlike the CBO)
  5. Their “real” incomes are deflated with the CPI, whereas the CBO use the PCE index.

So while this data set has its issue and probably isn’t very relevant to income distribution per se, it is a useful and probably relatively accurate picture of the distribution of 90th percentile of top AGIs.

—-

My thesis here is that most of the apparent divergence that we have seen over the past several decades is a function of several things:

  1. A changing tax code and, especially, top marginal rates
  2. A large increase from the mid-80s of business owners converting to or starting up as pass-through entities instead of C Corporations (e.g., S Corporations, LLCs, etc).  [Note: This is probably substantially a result of the fact that these pass-through entities started  paying lower effective rates relative to similar investments in C Corporations).]
  3. A significant change in the household composition and types of income earned at the lower end (fewer people per household, a smaller proportion of income as cash wages, increase in payroll taxes, etc)

The data from P&S provide some pretty powerful evidence for my first two arguments.

The following charts probably sum this up best:

[This is all income EXCEPT for capital gains index in real dollar terms, indexed to the year 1917]

[Or, a little cleaner: just wage and entrepreneurial income since that's the biggest delta]

Actual “real” dollars on a log  10 scale

Observations: Between the late 30s and the mid-60s we see a period of rising corporate profitability and rising incomes at the more modest income levels, e.g., 90th and 95th percentile.   We see essentially no rise in non-capital gain income for the very top income levels for almost 50 years!!  In fact, the percentiles appear to converge during this period.   The top incomes only trend significantly up and gain ground on the lesser groups when the top marginal rate is cut in the 80s and they do so implausibly rapidly — jumping almost two to three times their period levels in the space of just a few very short.

I think the reasons for this are very clear:  the very highest income earners would pay a much larger share of their real incomes at the top marginal rates (if they were to actually do so at least) and they typically have more flexibility to plan when and how they report their incomes (relative to people that get most of their incomes from a regular paycheck).  They can, for instance,  spread their accumulated earnings out over several years or to otherwise find tax advantaged ways to minimize the impact of very high rates.   In short: they demonstrate very high elasticity (at least with respect to taxable income).

Observation: Capital gains comprise a large and volatile share of top incomes so looking at the total number or absoute share can be a misleading.  The temporary spike in capital gains realizations in the mid 80s obscures the large and permanent jump in “wage” and “entrepreneurial” income that happened between 1986 and 1988.   That said, dividend, rental income, and interest income — usually passive types of income, decrease in both relative and absolute terms.   I rather suspect that after 1987 a growing share of entrepreneurs (especially owners of closely held businesses structured as pass-through entities) displaced the previous generation of passive “old money” types– their incomes simply got pushed into a relatively lower AGI group.

By income type

Wage Income

Note: “Wages”, in the P&S series, includes stock options and probably a significant fraction of other pass-through type income as well.  To me, at least, the movement of the highest income “wages” seems to track broadly with the profitability of domestic corporate activity (especially post 1987) which suggests that it’s comprised largely of options, 

Dividend Income

Observation: next to no dividend growth in real terms and, in fact, the highest incomes show less growth than relatively lower groups.  One would think that if the highest income groups were merely getting vastly richer in real terms that they would diversify and accumulate more stock in less closely held companies…

All non-capital gains income (includes entrepreneurial income as well as wage income–two very large components of income)

Interest Income (note: this comprises a very small fraction of total income for all groups):

Capital gains alone

Observation: I don’t know about this data … it seems a little too neat to me.  I suspect that they just allocated total capital gains realizations back to the various income groups according to their assumption (e.g., Pareto distribution or some such).  That said, if it is accurate, it would indicate that capital gains realizations have not been driving inequality.  [Perhaps I'll investigate this later when I have more time]

In any event, what I find interesting is that since 1990, after the most significant reductions to the top marginal rate and the tax code was reformed, there has only been a 30-40% increase in real (CPI-adjusted) non-capital gain incomes income at the highest reaches.

 Observation: Not much more when you include the highly volatile capital gains component.

Observation: “Wages” alone are only up ~30% for the two highest groups here and these are both quite clearly linked to the market.

Observation: The entrepreneurial component is the only significant part that’s dramatically up (~2x) for the highest income groups and it in a relatively stable fashion.

Income composition of P99.99-P99.99

Now if we remove entrepreneurial and capital gains….

We get about a 16% “real” gain against their CPI index since 1990.   I won’t claim it’s exactly this simple, since these various components are not necessarily as distinct as it might seem at the high end and some “entrepreneurial” income could be very similar to this “wage” income, but even allowing some fudge factor-it seems be roughly comparable to the comprensive income growth witnessed by most working people over a similar period of time (i.e., especially if we include healthcare and similar in-kind benefits).

 

Update

Please note that timing and composition of income gains amongst the top earners are very different along the grain you slice this (the top 1% itself is not nearly enough).

 

 

 

Observation: Even though you can see a significant response to tax rates the amongst the entire top 1% the fact the growing wages at the lower reaches obscures the extent of this…see below

 

Observation: The professionals, management, top sales people and the like occupying the lower reaches of the top 1% see a steady growth in their “wages”  incomes throughout the entire period.   The jump post-tax change there, but it’s relatively muted and the trend doesn’t seem to change terribly much afterwards either.  Now compare this to the very highest (P99.99-100) group P&S offer up….

 

Enough said.

 

Issues with the CBO income distribution data

The Congressional Budget Office periodically produces income distribution and effective tax rate data for households by income group.

One interesting, but little known fact, is they produce their “income categories” (quintiles, top 1%, etc) with a weighting according to the household size.

Here is their definition:

Income categories are defined by ranking all people by their income adjusted for household size—that is, divided by the square root of a household’s size. (A household consists of the people who share a housing unit, regardless of their relationships.) Quintiles, or fifths, contain equal numbers of people, as do percentiles, or hundredths. Households with negative income (business or investment losses larger than other income) are excluded from the lowest income category but are included in totals.

What this means is that a household with 1 person and 50K of income would be ranked identically to a household with 100K of income and 4 people, as would a household with 150K in income and 9 people, and so on.

Although I think this is, in some respects, a useful and perhaps necessary way of approximating the welfare of each individual household, I suspect they unintentionally mislead a lot of people with respect to both the effective tax rates and the actual distribution of income since few people probably know that they do this in the first place and fewer still understand the implication of this.

Consider, for instance, that if the wealthiest 0.5% of households (unadjusted for size) adopted 1 child each, it would surely produce a more “unequal” distribution since highest reaches of the income distribution would account for that much more of the population (CBO income groups always account for similar shares of the entire population), despite the fact that they have less discretionary income and haven’t (for the sake of argument) increased their incomes by one dime.

The CBO further confuses this issue by then quoting the average income, pre- and post- tax, across these adjusted-income groups without actually quoting the adjusted-incomes (which seems very strange to me indeed).    Thus, say, middle 20% of households may shrink dramatically in size and may include people with very different raw-income levels, but their published results do not give you any hint of this at all.

This is not an academic argument since, in fact, the households have never been identically sized and there has been a significant shift in the distribution of population (and, in fact, earners) amongst the households.

Below I have calculated the approximate size using their household count data (they round the numbers so there is a small amount of error between years).

Average number of people per CBO household income group (scaled to 1979)

Observation: The very richest and very poorest grew or stayed the roughly the same, whereas the middle income groups and the like dropped dramatically in size.  (Remember: this is after their weighting method so “middle” can mean very different pre-weighted incomes… the effects are probably even more dramatic w/o this weighting)

Average number of people per household income group – unscaled

Percentage of (unadjusted) households represented by the CBO’s various income groups – scaled to 1979’s values

Clearly these adjusted-income groups look VERY different now than they once did.

Below is what the results of these UNadjusted incomes look-like according to the adjusted-household income groups:

Now I can sort of correct for this by adjusting each adjusted group’s mean income according to the mean number of people in them…

UPDATE: Now let’s look at 1990 – 2009 [post-tax reform]

 

UNadjusted….

 

Adjusted….

Observation: After 1990, the adjusted incomes (which are probably a decent measure of the individual welfare of the household members) of all of the CBO’s income groups climbed 30 to 36% for ALL of their income groups [only a 5% difference between the top 1% and the middle quintile].  Their unadjusted  figure produces a 20% spread between the middle quintile and the top 1% (but notice that the middle quintile drops in size in this same period and, not coincidentally, lags behind everyone else in these unadjusted averages….)

Fortunately, the CBO does publish the minimum income levels for each income group according to household size, so we can remove some of this bias….

1979 – 2009

1987-2009

What I see is that similarly sized households in the income distribution saw similar gains.  Everyone gained appreciably, but higher income groups at all levels saw their income rise and fall more with the business cycle.

I included domestic corporate profit, the dotted green line on the right axis to make this relationship more apparent.  This domestic corporate profit measure is direct from the BEA and includes all private C-corporations as well as S corporations.  This, I believe, is important since higher income  fates are tied to this profit measure in several respects: market returns, senior management pay, and (crucially) it includes most of the gross income of pass-through entities that’s not normally reported elsewhere [the CBO also imputes corporate taxes to pre-tax income].

I also ran separate charts for 1979 and 1987 since the trend is much different, post 1987, and because I suspect (but cannot yet prove) that part of the trend may have been be linked how AGIs were reported pre-1986 (what or how the CBO corrected for this, I do not know).  In any event, tax reform clearly plays a large role (just compare the trends from between 1979 and 86 or 90– VERY different).

Anyways- the highest income groups still saw their incomes grow somewhat more rapidly than the less well off, but this impact is MUCH less dramatic than that reported by the CBO and the like and it seems to be clearly linked to domestic corporate profitability…

P.S., I realize that this does not account for the fat tail past the 99th percentile, but even still it seems pretty clear to me that much of the apparent inequality is a byproduct of their methodology that is highly sensitive to compositional issues.

Assessing corporate profits in context

To add some more persective to claims like this, i.e., assertions that corporate profits are at an all time high while labor costs are at an all time low…

These analysis suffer from several major issues:

1) The “wages” component is much too narrow — it excludes benefits like healthcare that are a large and growing share of total compensation.

2) The “profits” themselves are essentially just financial accounting profits — they do not account for inventory or capital depletion which boosts apparent profitability during recoveries and makes it seem much lower in the lead up.  more here

3) The profits used by Henry Blodget et. al are after-tax profit which, in presence of an apparent shift in corporate ETR, further exxagerates these trends since the apparent ETR is different now [much larger shares of NIPA corporate profits are earned foreign receipts and a much larger share of these profits are now also derived from S-corporations that do NOT pay corporate income taxes]

4) These profits include ALL national corporate profits — including a large and growing amount of non-repatriated foreign profits.   These profits are not, for the most part, generated with US labor (except, perhaps, for some corporate overhead), so it’s very misleading to blend that in….

Below I’ve attempted to correct for the above mentioned issues by using NIPA’s pre-tax corporate profit calculations with inventory and capital adjustments.   I have further focused primarily on domestic profits (both financial and non-financial) to see how these trends compare to historical levels.

I think the evidence is quite clear….

Domestic corporate profits as percentage of GDP and as ratio of total compensation of employees

Corporate_profit_labor_ratio

Composition of corporate profits by major source

Comporate_profit_shares

Note: This does not include the distiguish by corporate entity type, i.e., whether it’s a C-corp or an S-Corp [which is relevant to the tax rate debate... since they don't pay corporate taxes and they're a growing share of NIPA corporate profits   more here]

Aggregate labor costs vs aggregate domestic corporate profits

Corporate_profit_nominal_log_scale

Real labor costs and profit per civilian employee (EMRATIO * POP)

Corporate_profit_per_civilian_employee

Real labor cost and profit per capita (US population)

Corporate_profit_real_per_capita

Total real domestic labor cost and domestic profit

Corporate_profit_real_total

Total nominal labor cost vs nominal corporate profits (w/o scale this time)

Corporate_profit_nominal_totals
Note:
You can download all of the above data from the BEA directly at this link.  You will want both “Section 6″ files (pre and post 1969).
The other data can be extracted from FRED2, measures:
COE: Compensation of all employees (some of this is also accessible from these same tables, but it’s a little easier with FRED2)
PCEPI: PCE Index, with index set to 2011
POP: US population
EMRATIO: Civilian employment ratio of US population
Follow

Get every new post delivered to your Inbox.