Relevant and even prescient commentary on news, politics and the economy.

Guest post: Want a Flat Tax? I Got a Flat Tax for You

Guest post by Steve Roth

Want a Flat Tax? I Got a Flat Tax for You
crossposted with Asymptosis

One percent of financial assets. Personal and corporate. Annually.

With somewhere north of $55 trillion in U.S. financial assets out there (2009, down from $63 trillion in 2007), a Financial Assets Tax would generate more than $550 billion in annual revenue.

What could we do with that revenue? Here are some options:

• Eradicate taxes on corporate profits, dividends, and capital gains, and cut income taxes by between 22% and 51%. (Depends on which tax year you’re looking at; this for 2007-09. NIPA Table 3.2.)

• Pay off some of our national debt, invest in productive infrastructure to build true national wealth, greatly expand the EITC (and index it to unemployment) to turbocharge the real economy, or…

• Some equitable and economically efficient combination of the above

Why should we do this?

Simple: greater prosperity and greater equality. Both.

This idea seems to have far greater upsides than downsides. But I’ve undoubtedly missed some things, which I hope my gentle readers will fill in.

Here are some of the issues involved:

Innovation and growth. Naysayers (mostly — surprise! — large holders of financial assets) will scream that THIS PROPOSAL TAXES SAVINGS!!!, which for not-very-well-hidden and supposedly moralistic reasons is seen as BAD!!! (As my friend Gabriel in England said to me once, “Yes, well, we shipped all our Puritans over to you, now didn’t we?”)

Their post-hoc rationalization for that faux moralizing: savings are (more accurately: can be) used for investment, by which they mean (in this instance) fixed investment in productive assets — structures, equipment, software, etc. We want to encourage fixed investment, right? It drives growth (and long term, employment), and builds national wealth and prosperity, right? The answer to those questions is “Yes.”

But when they start objecting to the tax because it “will hurt poor people,” raise at least one eyebrow. First — as usual — they’re confusing flows with stocks. Savings is a flow. Money from savings goes into the stock of financial assets — cash in mattresses, bank account balances, CDs, stocks, bonds, collateralized debt obligations, etc.

But that misconception aside, talk about being wrong by 180 degrees. This proposal does indeed discourage saving — in favor of real investment.

Remember: there are only five things a person or company can do with a dollar of income:

1. Spend it on consumption (buy food or office supplies, pay wages for ongoing operations, etc.)

2. Invest it (to create or purchase — hence spur creation of — real assets)

3. Save it (“invest”/store it in financial assets — effectively or actually lending it out)

4. Pay off loans (basically saving, but on the other side of the balance sheet)

5. Pay taxes

If a dollar is “saved,” it is by definition not invested. (Though it or an identical, fungible equivalent might flow back out of the pool of financial assets to be spent on real investment –or on consumption, loan payoffs, or taxes.)

If people and companies sock away their income in financial “investments” (save it) and just enjoy the returns, one percent of those returns will be skimmed off every year. (Not terribly onerous, given that hedge-fund “investors” pay large multiples of that and still make piles of money for doing nothing.)

If, on the other hand, people and companies use that money to build houses, apartment buildings, malls, office buildings, amusement parks, and factories, invest in new equipment and software, or spend it on those deucedly hard-to-measure but massive contributors to our national asset base — education, training, research, and development — the assets they create won’t be hit by this tax.

And the ongoing income generated by those real assets will be taxed at a far lower rate.

Alternatively, the income that isn’t saved might be spent on consumption — increasing monetary velocity and aggregate demand, and making the whole swirling pie that is the economy, bigger.

If you think collateralized debt obligations are valuable national assets, you should hate this tax. If you think — correctly — that as Kuznets and many others have pointed out, real assets constitute true national wealth (though many of the most important real assets, like ideas, knowledge, skills, and “organizational capital,” are intangible and unmeasurable), you should love this tax.

What do we mean by financial assets? There are some gray areas that would need to be sorted out (insurance, pensions, etc.), but most financial assets quack like ducks. For a quick list, take a look at the column headings in Table 2A the of the Fed’s analysis of the Census’s 2009 Panel Survey of Consumer Finances (XLS):

Transaction accounts
Certificates of deposit
Savings bonds Bonds
Stocks
Pooled
investment funds
Retirement accounts
Cash value life insurance
Other managed assets Other

We might even want to include cash (actual dollar or euro bills). Why should we encourage cash in mattresses? TBD.

My friend Steve tentatively suggested “anything that’s traded on an exchange,” which seems like a good idea except it would encourage Wall Streeters to move towards assets that are traded off-exchange, over the counter. We’ve seen the effects of that.

What about incentives and economic distortions? How much would this tax distort economic decisions? Think: Nobody, ever, says “I’m not going to get wealthy because I’d have to pay taxes.” Why? Ask any of Jane Austen’s heroines: because there’s no substitute for wealth.

Earned income? Quite otherwise, because there’s a very good substitute for working to earn money: leisure. Spending time with the kids, playing golf, writing overly long and abstruse economic blog posts, watching NASCAR, assembling intricate miniatures of Civil War battle scenes.

In economic terms, the demand for employment is quite elastic because there’s an attractive substitute. The demand for wealth is quite inelastic, because there just ain’t no substitute for being rich.

The Financial Assets Tax would provide a very slightly lower incentive to earn money every year (anyone care to do the arithmetic?), but nothing like the disincentive that results (in theory, to some greater or lesser degree) from high marginal income tax rates, corporate profit and dividend taxes, etc.

The biggest incentive — arguably an economic “distortion,” but every tax except maybe land value taxes is distortionary — would be to spend on real investment (and consumption) instead of saving.

But that incentive would actually compensate for an inherent distortion resulting from the nature of financial assets, which are at root an artificial creation: Financial assets don’t decay and depreciate like real assets do. After ten years (or whatever), you still have the initial capital, plus the returns, which is not true with fixed investments. So fixed investments have a big disadvantage when they’re competing for “investment” dollars. A Financial Assets Tax would to some degree correct for that inherent market distortion/inefficiency.

Too big to fail. I’ve pointed out that the financial economy — the trade in financial assets — is many time larger (40x, 50x?) than the real economy — trade in goods and services. And it’s arguably many times larger than is necessary to lubricate and intermediate the real economy. And innumerable wise voices have pointed out the negative externalities of this excessive size: systemic risk of financial-market meltdowns that trash the real economy, gross misallocation of financial and human resources, etc.

There have been some salutary if rather timid proposals to address this via taxes on financial transactions — the flows — to compensate for those externalities and shrink the sector. But this proposal for taxing the stock of financial assets could be a superior alternative. I’ll leave it to others, for now, to analyze the pros and cons of those two alternatives.

What about private residences? This is both a large segment of fixed investment, and kind of a special case — different from business investments because the value derived isn’t in the ability to produce more, saleable goods and services, but having a roof over your head.

Here’s the functional scenario: you’ve got half a million dollars in financial assets, and you want to build a half-million-dollar house. Here are the two ends of the spectrum — you can land anywhere in between:

1. Sell all your financial assets and spend the money to build the house.

2. Sell $50,000 of financial assets for the 10% down payment on a loan, and borrow the rest.

Your taxes on the house asset are the same either way. But in scenario 2, under the Financial Asset Tax proposal you also pay tax every year on the $450,000 in financial assets you’re still holding. So it’s less attractive. It effectively increases the interest rate on your loan by 1%.

Putting aside for the moment all the other factors you personally consider (liquidity, risk, return, peace of mind, etc.): which of these scenarios — if repeated by millions of people over decades — contributes more to national wealth and prosperity? Which should the tax system encourage? (Because every tax system encourages something.)

The first scenario reduces the value of financial assets by reducing demand for them; the second increases it. As Dirk Bezemer has explained, borrowing-driven booms in financial asset values drain resources from the real economy, and so are associated with slower real-sector growth. The second scenario also effectively prints $450,000 in new money, causing more inflation pressure, which puts pressure on the Fed to raise interest rates, which discourages real-sector investment.

Which one should we encourage — borrowing or real investment? You decide. (And yes: we should also eradicate the insanely economically inefficient — and regressive — mortgage interest deduction, which makes the second scenario so much more attractive.)

What about volatility? The value of U.S. financial assets (nominal — not inflation-adjusted) fell by 13% from 2007 to 2009 (assets held by U.S. households, nonprofits, and nonfarm, nonfinancial businesses — Fed Flow of Funds TFAABSNNCB + TFAABSHNO). Under this proposal, that would result in a huge hole in the Federal budget.

Personal income went up by 2% (nominal) over that period (NIPA Table 2.1). This tells you: the revenues from a Financial Assets Tax would be much more volatile than from income taxes, because asset values are far more volatile than incomes.

But let’s step back: Every reasonable person (this excludes large portions of the Republican and Tea Parties) agrees that it’s smart for government to spend more in the bad times (causing a government deficit), and less in the good times (causing a surplus). It’s intuitively obvious (thanks to Keynes), we’ve seen it work (1939 passim), and all sorts of old and new economic theory, notably Modern Monetary Theory, supports it in spades.

The problem, of course, is politicians. When bad times hit and government revenues are down, they panic like scared children. They don’t remember (or look forward to) the good times, like when the federal debt was plummeting during the Clinton boom/tax-increase/surplus years. So they do exactly the opposite of what common sense and prudent wisdom prescribe: they cut spending. The volatility of the Financial Asset Tax could contribute greatly to this panic-driven policymaking.

My only reply: it is to be hoped that the economic efficiency of the Financial Assets Tax — the growth and prosperity it engenders, especially in the real sector of the economy — would over the long term overwhelm the negative effects of political mismanagement. Other suggestions to overcome this difficulty, much appreciated.

What about evasion and offshoring? Stocks of financial assets are easier to track and harder to hide than flows of income. They have to be stored somewhere. Since they’re not ongoing flows, they can’t be laundered and hidden as easily through multiple international pipelines and entities. People will still use secret accounts in the Bahamas to hide their money (as long as we allow the Bahamas bankers to get away with it; the Swiss can’t anymore…). And yes, people and businesses will figure out new schemes to evade this tax. People will always figure out ways to commit fraud. A simple tax on an easier-to-track item will make it harder for them to do so.

Since we’ll tax domestic entities’ financial assets no matter what country those assets happen to be stored in (just as we tax worldwide income now — at least of natural humans) — people and businesses will have less incentive to keep piling up their treasure in financial (and real) assets overseas. They can bring it home at no cost (they’re paying taxes on it in either place) hopefully investing it in real assets here.

Why not a consumption tax or VAT instead? I don’t actually know much about these, so take this for what you will.

Those taxes are good because they don’t penalize or kill incentives for work, innovation, and real earnings (personal earned income and real — nonfinancial — corporate profits). But both of them, as I understand the proposals and real-world examples I know of, tax investment spending at the same rate as consumption spending. So they’re not really just consumption taxes. They’re also investment taxes. That’s not good. I assume there are carve-outs to correct for that, but the more carve-outs you put in place for various and sundry reasons, the more messed up, gamed, and inefficient tax regimes become. The Financial Assets Tax makes things much more clear and simple. That’s one value of a flat tax.

Also: to make consumption taxes reasonably progressive, the top marginal rates have to be high, maybe even greater than 100%. Imagine a million-dollar consumer paying two million dollars in taxes on that consumption. You think that’s gonna happen?

How do you phase it in? This would be a big change in the rules of the game. It’s both fair and efficient to give people time to adapt. I would do it based on: 1. person versus company, 2. age, and 3. quantity of financial assets.

For people: Give a few year’s warning, then in the first year, people 30 and under with more than $5 million in financial assets would be subject to the tax. (I am rather unconcerned with these people’s well-being, or with their ability to adapt to the new regime over the course of their lives.) Increase the age and reduce the cutoff — perhaps ending at around four to six times median income — over ten or fifteen years. Other personal tax rates should decline in concert. For companies: Give two or three years warning, then replace the tax on C-corp profits with the Financial Assets Tax. They’ll adapt just fine — mainly by shifting from financial investment to real investment, but also probably by increasing dividends, putting the money back out there where it can be intelligently allocated by the wisdom of the crowds rather than by CEOs’ purported omniscience. Maybe this will also encourage corporations to hire CEOs who are real business managers, rather than practitioners of financial prestidigitation.

Oh yeah: equity! Don’t change the channel, America. It matters. It especially matters to those who don’t have it. (Few of whom are reading this.) But excessive inequity hurts the rich too, in the long run, because it kills long-term national prosperity. Economically efficient policies that deliver greater equity also deliver greater long-term prosperity. Even the rich get richer under progressive policies (except maybe the very, very rich). The poor and the middle class get far richer.

What do we mean by equity and inequity? Are we talking about income inequality? Feh.

Our world in pictures (regular readers will recognize some of these, but some are all new):

DESCRIPTIONSource: Piketty, T. and Saez, E. 2007. Income and Wage Inequality in the United States 1913-2002. In Atkinson, A. B. and Piketty, T. Top Incomes Over the Twentieth Century: A Contrast Between Continental European and English-Speaking Countries, Oxford University Press, Chapter 5; series updated by the same authors. Hat tip Catherine Rampell.

Source (pdf). The tax info here is just an added bonus feature, showing that above about $60K or 80K in income, our tax system (local, state, federal combined) isn’t progressive at all. The people making $160K pay the same share of their income pool as the people making $160 million. DESCRIPTIONSources (PDF): Sylvia A. Allegretto, Economic Policy Institute; Edward Wolff, unpublished 2010 analysis of the U.S. Federal Reserve Board, Survey of Consumer Finances and Federal Reserve Flow of Funds, prepared for the Economic Policy Institute. Another hat tip to Catherine Rampell.

Figure 2. The actual United States wealth distribution plotted against the estimated and ideal distributions across all respondents. Ariely and Norton, 2010 (PDF).

The bottom 80% gets about 40% of the income.

The bottom 80% owns about 15% of the wealth.

You want freedom? Look to your bank balance. You want opportunity? Look to your bank balance. You want time and space enough to innovate and be an entrepreneur, without the disincentive (not to mention embarrassment and inconvenience) of potential financial catastrophe? You want to buy birthday presents for your kids, get their teeth fixed, or put them through good colleges, maybe take a family vacation every year or two? You know where to look.

But if you’re not in the top 20%, you’re not looking at shit. So who pays the tax? You can see the ownership breakout for financial assets sliced various ways in the aforementioned Fed/Census Bureau table (XLS). Here’s the breakout for personal holdings, by levels of income:

This would be a very progressive tax, because the the distribution of wealth is very regressive. It would compensate quite effectively for all our country’s regressive taxes, like payroll taxes, sales taxes, property taxes, and cut-rate taxes on financial investments.

I don’t have the wherewithal to calculate the total resulting progressivity. Perhaps it would be excessive, to the point of economic inefficiency (though I doubt it).
If we thought that was the case, we could reduce the rate from 1% to .75% or .5%, and continue income and other taxes at higher rates. Subject to discussion.

Bottom line: If we want widespread freedom, opportunity, innovation, entrepreneurship, and healthy, happy living, all feeding on itself in a virtuous, self-perpetuating cycle, then broadly based wealth distribution is at least one necessary condition. A tax on financial assets would be an (economically) efficient and effective means to move towards that goal.

Oh, and not that this matters, but in the short and medium term, while pies are growing and boats are rising, tens of millions of people get to suffer less. But maybe it does matter, because in the long run, you know…

Tags: , Comments (34) | |

Drop the corporate saving rate, please…

Update: The term corporate savings below refers to excess saving, gross saving over gross domestic investment, as a percentage of GDP. This is the defined 3-sector financial balance model (referred to below).

The Federal Reserve Flow of Funds showed a third quarter shift in the financial sector balances: the corporate saving rate declined 0,25% to 2,7% of GDP; the household saving rate fell 0,13% to 3,8% of GDP; the current account fell 0,11% to 3,5% of GDP; and the government increased its saving rate 0,27% to -10,0% of GDP.

Basically, the government was able to increase saving slightly, even as foreigners increased surpluses against the US, at the cost of reduced household and firm saving.


The chart above illustrates the 3-sector financial balances approach, which is the identity that the private sector and public sector saving rates must equal that of the foreign sector (the current account). The private sector is broken into the household and corproate sectors. For a discussion of the 3-sector financial balances, see Scott Fullwiler, Rob Parenteau; and I’ve written on this as well.

(Note: I am in Deutschland, where the keyboard and number system are slightly different from that in the US – so for this post, I can write ß whenever I want to, but I won’t, and all numbers with “,” represent an American decimal point, “.”. Funny thing is, when I use the Blogger spellcheck here, everything is highlighted yellow, so I plead “in Deutschland” for any misspellings :))

Some people may see the large government deficit, still -10% of GDP, as the ‘problem child’ of the sectoral financial balances. Me, I see the government deficit as a red herring of the corporate saving rate, which remains stickily in the 2-3% range. Until the corporate saving rate falls markedly, let’s say to zero or below, the unemployment rate is to remain high, and the household deleveraging process slower than would otherwise be if wages and disposable incomes were growing more quickly.


The chart illustrates the corporate saving rate and the unemployment rate, both have an 84% correlation. Therefore, adjusting for the standard deviations, corporate saving and the unemployment rate move roughly in sync. When the corporate saving rate is negative, firms purchase new capital goods and hire labor for production faster than they accumulate financial assets, thereby reducing the unemployment rate. In contrast, when the saving rate is positive, firms are investing in financial assets (or consuming capital at a higher rate) faster than they are increasing the capital stock and labor force, thereby increasing or leveling the unemployment rate.

In a very simple linear regression model (chart below), the relationship betwen the corporate saving rate and the unemployment rate exhibits an R2 of 70%. Accordingly, reducing the corporate saving rate to zero corresponds with a near-3ppt drop in the unemployment rate to 7%, all else equal, of course.


So one way to quicken the household deleveraging process is to reduce the corporate saving rate. Reducing the corporate saving rate corresponds to a falling unemployment rate, so that workers accrue SOME pricing power (they have none at this point).

Another way is to increase the fiscal deficit, to Mark Thoma’s point in The Fiscal Times this week. The correlation between the government financial deficit and the unemployment rate is -92%.

This is where the two come together: fiscal policy needs to provide incentives to lower the corporate saving rate.

Rebecca Wilder

Tags: , , , Comments (22) | |

Evaluating the "excess" in the US corporate financial balance

In a NY Times op-ed, Rob Parenteau and Yves Smith reminded us that the private sector financial balance is a function of the household financial balance and the corporate financial balance. They concluded the following regarding excess corporate saving:

So instead of pursuing budget retrenchment, policymakers need to create incentives for corporations to reinvest their profits in business operations.

In my view, it’s not that simple (not that passing this type policy would be easy at all). As I illustrate below, firms, like households, are in a deleveraging cycle, where corporate excess saving is likely to persist for some time. (Note: US total corporate financial balance, excess saving if the balance is positive, is roughly undistributed profits minus gross corporate domestic investment)

In their NY Times article, Rob and Yves cite a 2005 JP Morgan study, “Corporates are driving the global saving glut”. In that study, JP Morgan argues that the global saving glut has been driven largely by G6 excess corporate saving, and to a lesser extent emerging economies. In the US, positive corporate excess saving persisted through the latest print, 2010 Q2.

The illustration above plots the total corporate financial balance as a percentage of GDP. I calculate the Total Corporate Financial Balance (TCFB) as in the JP Morgan study, which is the residual of the national accounting identity of the Current Account Balance minus the Household Financial Balance minus the Government Financial Balance. According to this measure, the TCFB was roughly +3% in 2010 Q2, or about +1% above the 2008-2010Q2 average (2.1%).

About the same time as JP Morgan published their research, the IMF and the OECD were wondering why global TCFBs were rising. Several factors are attributed the upward trend in the first half of the 2000’s, including (this is not a complete list of factors):

  • Repurchase of stock shares relative to dividend payouts
  • The falling relative price of capital goods dragging nominal investment spending as a share of GDP (see Table 3.2 of the OECD publication)
  • The overhang of leverage build in the 1990’s
  • Rising profits via falling taxes and low interest payments (especially in other OECD economies)

Although firms likely worked out much of the debt overhang from the 1990’s, the debt accumulation spanning the second half of the 2000’s was precipitous.

It’s very unlikely that the excess corporate saving will fall anytime soon, as non-financial business leverage is high just as household leverage is high. Total non-financial business debt peaked in 2009 Q1 at 79.5% of GDP and is now trending downward, hitting 74.9% in 2010 Q2.

If history is any guide, then the “excessive” borrowing spanning 2005-2008 will take some time to repair. Spanning 2002 to 2004, the non-financial business sector dropped leverage 2.5% to 64%. If this 2-year period of de-leveraging indicates an “equilibrium” level of leverage, then non-financial businesses are likely to run consecutive financial surpluses (excess saving) in order to reduce debt levels by another 11 percentage points of GDP for a decade more.

If firms run excess saving balances, then they’re not investing in future profitability via capital expenditures nor increasing marginal costs, like wages and hiring, relative to profit growth. So while it is true that some fiscal policy should be targeted directly at investment incentives (Rob Parenteau and Yves Smith article), these measures may prove less effective since the non-financial business sector’s desire to “save” and repair balance sheets is high.

I leave you with one final chart to inspire more discussion: a breakdown of the total corporate financial balance into its two parts, financial-business and non-financial business.

The financial balances in illustration 2 are computed directly from the Flow of Funds Accounts, Table F.8, rather than taking the residual as calculated in illustration 1. Thus, the total corporate financial balance will not match that in illustration 1.

The point is simple: the small drop in excess saving in total corporate financial balance in illustration 1 is stemming from the financial side. The non-financial corporate sector continues to raise excess saving by investing retained earnings into liquid financial assets relative to capital investment.

Fiscal policy should be targeted at the high desired saving by the corporate and household sectors alike. The idea is to pull forward the deleveraging process by “helping” households and firms lower debt burden via direct liquidity transfers (lower taxes or subsidies, for example). Only then will healthy private-sector growth resume.

Rebecca Wilder

Tags: , Comments (16) | |

Household leverage: what does the US have that the UK does not?

Earlier this week I compared household saving rates across the US, UK, Canada, and Germany. My conclusion was pretty simple:

So generally, this simple analysis would suggest that Menzie Chinn’s skepticism of a “status quo” of US consumer imports is worthy. But with the status quo firmly in place in Germany, the household saving data paint a foreboding picture – certainly for the Eurozone, but possibly for the global economy as well.

The financial circumstances of US and UK households are very similar despite their diverging saving rates over the last two quarters (see saving rate chart here): leverage is high.

The chart above illustrates the total stock of household loans/debt (including non-profit organizations, which is small relative to the “household”) as a share of personal disposable income.

In the UK, household leverage peaked above that of the US at 161% of personal disposable income in Q1 2008, having fallen to 149% by Q1 2010. Furthermore, recent deleveraging by UK households has occurred through income gains, rather than paying down debt: spanning the period Q2 2009 to Q1 2010, the UK household stock of loans increased 1.2%, while disposable income grew 3.1% (you can download the data here).

Given the remaining leverage on balance, the divergence in household saving rates across the US and UK is probably not sustainable. The UK household saving rate is likely to increase, or at the very minimum, hold steady.

The problem is: that according to the sectoral balances approach, it’s impossible for the government and the private sector to increase saving simultaneously unless the UK is running epic current account surpluses (it’s not). Therefore, the £6.2billion in public “savings” may push UK households farther into the red. However, the more likely outcome is that UK public deficits rise amid shrinking aggregate demand (and with it, tax revenue) and the increasing household desire to save.

The punchline: the US household has something that the UK household does not: (still) expansionary fiscal policy ($26 billion in state aid and extending unemployment benefits, for example).

Rebecca Wilder

Tags: , , Comments (2) | |

Household saving rates in the US, UK, and Germany: (possibly) light at the end of the tunnel

Menzie Chinn at Econbrowser breaks down US import data by sector to argue the following (see entire article here):

What is clear is that consumer goods do not vary that much; now, part of auto and auto parts is going to satisfy consumer demand as well, and here we do have some evidence in support of the hypothesis of the consumer going back to his/her old ways of sucking in imports.

Consumption hardly seems resurgent, so attributing the increase in imports to consumers means that one is assuming a very high share of imports to incremental consumption — something I’m not sure makes sense. So, I think the book is still open on whether the consumer is going to drive the US back into a rapidly expanding trade deficit.

Another way to look at this is by comparing global household saving rates. Specifically, I look at the household saving rates across the US (the world’s largest economy in 2007, as measured in PPP dollars – download the data at the IMF World Economic Outlook database), UK (6th largest economy), Canada, and Germany (5th largest economy). The household saving ratio is calculated as gross household saving divided by personal disposable income, as reported in country National Accounts.

If the global economy is indeed “rebalancing”, then relative to disposable income the big spenders (US, UK) raise saving, while the big savers (Germany) increase spending. In contrast, if the global economy is returning to the pre-crisis “status quo”, then relative to disposable income household saving rate would:

  • fall in the US and UK
  • rise in Germany

(Using IMF data, here’s a chart that I put together last year of consumption shares across economies to illustrate the big spenders and big savers.)

The German household saving rate is rising, while the UK households saving rate is falling. In the US, we’re seeing the household saving rate stabilizing above pre-crisis levels, even increasing at the margin.

The table below lists average household savings rates for the pre- and post-crisis periods. Notably, the average US saving rate more than doubled to 4.8% since the previous 2005-2007 period, while that in the UK increased a much smaller 36% to 4.6%. Notably, German households increased average saving above an already elevated 10.6% average during the business cycle.

So generally, this simple analysis would suggest that Menzie Chinn’s skepticism of a “status quo” of US consumer imports is worthy. But with the status quo firmly in place in Germany, the household saving data paint a foreboding picture – certainly for the Eurozone, but possibly for the global economy as well.

I’m in no way “blaming” this on the Germans – the banking system there will eventually contend with the crappy Greek and Portuguese assets they hold on balance. But didn’t they learn their lesson? Relying on exports makes the economy highly susceptible to external demand shocks.

More on the UK vs US in my next post.

Rebecca Wilder

Note: Clearly, an analysis of this sort would require a much larger cross-section of household saving data. But differing measurement methodologies and data limitations make the comparison too arduous for a simple blog post. For example, Japan is not part of the analysis because only the expenditure approach to national income is available on a quarterly basis.

Tags: , Comments (12) | |

The compensationless recovery

New York Times David Leonhardt argues that real wages are rising, so those resilient workers that remain employed will benefit from the bounce-back in “effective pay”. The problem with this insight is twofold: first, the expansion phase of real hourly compensation, a broader measure of total earnings, is falling; and second, sitting atop a mountain of consumer and mortgage debt, the aggregate economy cannot afford a compensationless recovery.

From the NY Times:

But since this recent recession began in December 2007, real average hourly pay has risen nearly 5 percent. Some employers, especially state and local governments, have cut wages. But many more employers have continued to increase pay.

Something similar happened during the Great Depression, notes Bruce Judson of the Yale School of Management. Falling prices meant that workers who held their jobs received a surprisingly strong effective pay raise.

Rebecca: The referenced “real wages” are the real average hourly earnings figures for production and nonsupervisory workers, 80% of the total nonfarm payroll. The broader measure of total earnings is real hourly compensation (see Table A and get the data from the Fred database). Real hourly compensation measures compensation for all workers, including wages, 401k contributions, stock options, tips, and self-employed business owner compensation. (You can see a comparison of the earnings/compensation series in Exhibit 1 here.)

Since December 2007, real hourly compensation has increased just 1.3%. Furthermore, the index declined four consecutive quarters through Q2 2010, a first since 1979-1980. If the NBER dates the onset of the expansion at Q3 2009 (the first quarter of positive GDP growth in 2009), real hourly compensation will have dropped .7% through Q2 2010! That’s pathetic compared to the average 2.5% gain during the first 4 quarters of expansion spanning the previous 10 recessions.

The table lists the gain/loss of real hourly compensation, measured by the BLS, during the recession and early recovery for the business cycle as dated by the NBER.

Here’s how I see it: the problem is not that real hourly compensation is falling during the the recovery, per se, it’s that real hourly compensation is falling during the recovery of a balance sheet recession.

In the context of wage and compensation growth, the NY Times article was misleading in its comparison of the Great Depression to the ’07-’09 Great Recession. Mass default during the Great Depression wiped private-sector balance sheets clean, no debt. But not this time around. We’re going to need a lot of income growth (the BLS measure of real hourly compensation includes measures of income at the BEA) to increase saving enough to deleverage the aggregate household balance sheet.

I’ll say it again: we can’t afford a jobless recovery. Specifically, we can’t afford a compensationless recovery.

Rebecca Wilder
submit to reddit

Tags: , , Comments (22) | |

Flow of Funds Accounts: some are deleveraging, while others are not

by Rebecca

The Federal Reserve released its quarterly Flow of Funds Accounts, and the message is crystal clear: the private sector is dropping debt burden, while the public sector is growing it.

Quarterly private sector debt growth, households + nonfinancial business + finance, has been slowing or negative since the second half of 2007. In contrast, federal and state and local governments are selling debt like it’s going out of style, with 28.2% and 8.3% annualized debt growth in the second quarter of 2009.

It is no secret that the private sector is unwinding debt, but to what end? 100% of income? – 110%? – Or 65%?

According to Reuven Glick and Kevin J. Lansing at the San Francisco Fed, Japanese households dropped their debt burden to 95% of disposable income. If US households were to follow a similar path, then the debt cycle would be complete in 2018. An excerpt from the article:

After Japan’s bubbles burst, private nonfinancial firms undertook a massive deleveraging, reducing their collective debt-to-GDP ratio from 125% in 1991 to 95% in 2001. By reducing spending on investment, the firms changed from being net borrowers to net savers. If U.S. households were to undertake a similar deleveraging, their collective debt-to-income ratio would need to drop to around 100% by year-end 2018, returning to the level that prevailed in 2002.

There is deleveraging still left in the pipeline, but one cannot say that the Japanese experience foretells the path of US debt. The economic agents, their propensities to save, and underlying economic fundamentals are different: 100% debt to disposable income in Japan may not be the equilibrium level in the US. Unfortunately, though, nobody can tell you what the level is…just something less than 125%.

The path of saving (paying down debt)

The US economy has suffered a precipitous drop in consumer demand, as the marginal saving rate surged. Going forward, higher saving (the average saving rate) does not preclude income and economic growth per se, but increasing saving (the marginal saving effect) can.

As wealth effect ratios stabilize – the chart to the left features the wealth effect as household net worth/personal disposable income – I believe that household saving will stabilize and consumer spending will grow with income.

Admittedly, though, the lag structure of the recent anomalous wealth effect is not known, and the strong marginal effect on saving might continue (i.e., the saving rate grows, as in the San Francisco Fed paper). To be sure, the labor market has dropped wage growth to record lows (see Mark Thoma’s post here), and Q2 ’09 annual disposable income growth was negative (a first since 1951). Not good for contemporaneous saving and spending growth.

The next four quarters, or the early period of recovery, will be critical in setting the stage for income growth. The recovery is expected to be weak, with the consensus GDP growth forecast around 2.4% in Q4 2009. But given the precipitous decline in output, even a 5% annualized quarterly growth rate during the early recovery would be rather “weak”. There’s room for an upside surprise as financial and housing markets stabilize.

Rebecca Wilder (if you are interested, I listed additional Flow of Funds charts here)

Tags: , , , Comments (0) | |

Broken Clock Day

Yves Smith quotes Thomas Friedman accidentally telling the truth:

Since President Bush came to office, our national savings have gone from 6 percent of gross domestic product to 1 percent, and consumer debt has climbed from $8 trillion to $14 trillion.

Please explain this in the context of the “savings and investment boost” that was supposed to come from the 2001 and 2003 tax cuts.

That means you, Bob McTeer.

Bulldog to bulldog, as it were.

Tags: , , , Comments (0) | |

Assets and Debts of Younger Households

One thing we heard in the comments to Coberly’s bravura Social Security op-ed was an opinion that the younger generation should be allowed to opt out of the program so that they can save their own way to retirement riches. This had me thinking, what’s the data saying about recent behavior — after all, factors including the declining prevalence of private defined-benefit pensions would increase households’ need for private savings.

So are they actually saving early and often (a good idea, BTW)? Alas, the Fed’s Survey of Consumer Finances isn’t very fresh: we can expect 2007 survey results sometime next year. Still, the 2004 results give us a bit of stock bust and housing bubble, and the ability to make some long-range comparisons. Here’s the median net worth, holdings of financial and nonfinancial assets, and debt loads [*] for households headed by under-35s and 35-44 year olds, in thousands of 2004 dollars:

Ups and downs of the housing and stock markets being what they are, younger households in 2004 find themselves no more wealthy than their predecessors of nearly a generation ago, despite a a lot of risk transfer from institutions to individuals. This certainly doesn’t make much of a prima facie case for removing the retirement safety net.

It’ll be very interesting to see the non-financial asset and debt data for ’07.

[*] A quirk of the reporting is that the median asset holdings are reported for the subsets of households with assets. Given the prevalence of asset holdings for these aggregates, the asset and debt figures are roughly the 55-60th percentile figures.

Tags: , Comments (0) | |