January 09, 2013
Guest Contribution: "The Myth of 'Jobless Recoveries'"
(a.k.a. Okun’s Law is Alive and Well)
It is rare to call an economic relationship a “law.” But Okun’s Law has earned its name. In 1962, Arthur Okun found a relationship that has become enshrined in textbooks as Okun’s Law. The textbook version states when U.S. output dips one percent below its potential, unemployment rises above its natural rate by about half a percentage point.
Fifty years after Okun’s paper, we find that this relationship fits very well, including during the Great Recession. Deviations from Okun’s Law occur, but they are usually modest in size and short-lived. Our results imply that it is a fallacy to claim that the last three U.S. recoveries have been “jobless recoveries”. Rather, the recoveries have been slow, and the slow job growth that has resulted has been just about what one would expect based on Okun’s Law.
Our results matter for policy choices. Okun’s Law is a part of textbook models in which there are shocks to aggregate output, which lead firms to hire and fire workers. Paul Krugman for instance wrote in 2011: “Why is unemployment remaining high? Because growth is weak — period, full stop, end of story. Historically, low or negative growth has meant rising unemployment, fast growth falling unemployment (Okun’s Law) ... what we’ve been seeing lately is well within the normal range of noise.” Krugman’s policy prescription since the onset of the Great Recession has been for sufficient aggregate demand stimulus to boost growth and lower unemployment.
Skeptics of Okun’s Law question this prescription. They argue that Okun’s Law has broken down because of problems in the labor market, such as mismatch between workers and jobs. They stress labor market policies such as job training, not demand stimulus, as the key to reducing unemployment. For instance, Diana Furchtgott-Roth of the Manhattan Institute wrote in a recent op-ed that “Job creation isn't moving fast enough. It's time to try a different set of solutions.” Her proposed solutions, based on a study by Deloitte, are to boost the “talent competitiveness” of U.S. workers “through changes in education, immigration, occupational regulation, foreign direct investment, unemployment insurance and patent law.”
U.S. evidence: Fit & Stable
Using annual U.S. data from 1948 to 2011, we find that the Okun’s Law has a coefficient of –0.4 or –0.5, with an in the neighborhood of 0.8. Chart 1 illustrates the fit of the estimated Okun’s Law by plotting the unemployment gap (the gap between unemployment and the natural rate) against the output gap (output relative to potential). The relationship is very tight. No year is a major outlier in the graphs.
Chart 1. United States: Okun’s Law, 1948-2011 (Annual data) (Natural Rates Based on Hodrick-Prescott (HP) Filter with λ = 100)
Some economists, including Menzie Chinn and Ferrara and Mignon in Econbrowser posts, suggest that Okun’s Law is non-linear, with different unemployment effects of increases and decreases in output. At least for annual data, Chart 1 suggests that a linear Okun’s Law fits the data well. When we estimate separate coefficients for positive and negative output gaps, the estimated coefficients are –0.37 for positive output gaps and –0.39 for negative gaps.
Previous researchers also suggest that the coefficient in Okun’s Law varies over time. Once again, we find no evidence against our simple specification with a constant Okun coefficient. A test for a break in the Okun coefficient at an unknown date fails to reject the null of parameter stability for all three of our baseline specifications.
Our finding of a stable Okun’s Law is robust to various methods of measuring short-run movements in output and unemployment. We try alternatives to the Hodrick Prescott (HP) filter. We also estimate the relationship in “changes”, that is between the change in the unemployment rate and the change in (log) output, which does not require using the HP or any other filter. The relationship holds for quarterly as well as annual data. Chart 2 shows the tight fit between actual unemployment and the estimate based on Okun’s Law. Some residuals are evident during the early years of the Great Recession, for which Ferrara and Mignon provide some conjectures.
Chart 2. United States: Actual and Fitted Unemployment Rate, 1948Q2-2011Q4. Notes: Figure reports fitted unemployment rate from Okun specification estimated on quarterly data in levels with two lags and natural rates based on Hodrick-Prescott filter with λ = 100.
The Myth of Jobless Recoveries
It is often claimed that U.S. recoveries since the 1990s have been “jobless.” We find no evidence that Okun’s Law broke down during these episodes. Confusion on this issue has arisen, as Gali, Smets and Wouters (2012) have also noted, because output grew more slowly in recent recoveries than in earlier ones, causing high unemployment to linger.
To see why recent recoveries might appear jobless consider Chart 3, which shows the paths of output, unemployment, and the employment-population ratio from 2007 through 2011. We also present estimates of the long-run levels of the three variables based on their pre-recession behavior. Specifically, we estimate trends with the HP filter through 2007, and assume that the natural rate of unemployment, the growth of potential output, and the growth of the long-run employment-population ratio remain at their 2007 levels over 2008-2013.
Many interpret the experience shown in Chart 3 as a jobless recovery. For example, the website of National Public Radio (2011) presents similar graphs under the headline, “Output Came Back, Employment Didn’t.” These statements are true in the sense that the employment-population ratio has been steady at a low level, while the growth rate of output has returned to normal (the paths of actual and potential output are roughly parallel) and the level of output surpassed its pre-recession peak in 2011.
Yet, as the chart makes clear, Okun’s Law has not broken down. Since a large output gap has persisted, Okun’s Law predicts large deviations of employment and unemployment from their long-run levels. From 2009 through 2011, the output gap as measured in Chart 3 averaged -10.8 percent and the unemployment gap averaged 4.4 percentage points. In fact, the ratio of the two gaps, –0.41, is close to our estimate of the Okun coefficient.
Chart 3a. U.S. and the Great Recession: The Myth of a 'Jobless Recovery': Log of Real GDP
Chart 3b. U.S. and the Great Recession: The Myth of a 'Jobless Recovery': Unemployment Rate
Chart 3c. U.S. and the Great Recession: The Myth of a 'Jobless Recovery': Log of Employment-to-Population Ratio
Why have the last three recoveries been viewed as jobless, while previous recoveries were not? This is because the speed of recoveries has been slower than before. In the early 1990s and early 2000s, as well as after the Great Recession, slow growth meant that sizable output gaps persisted well into the recovery. In contrast, in most earlier recessions, the output trough was followed by a period of above-normal growth that pulled output back to its previous trend. As Okun’s Law predicts, unemployment also returned to normal, making the recoveries look job-full.
Chart 4 illustrates this point with data for the early 1980s. After the recession of 1981-82, output growth averaged nearly 6 percent over 1983-84, with the result that output, employment, and unemployment were all close to their previous trends by 1984. Based on experiences like this one, observers came to expect that the end of a recession would lead quickly to a full recovery of employment. They were surprised when this did not happen more recently, even though Okun’s Law has not changed.
Chart 4a. U.S. and the 1981 Recession: A Recovery that Looks ‘Job-full’: Log of Real GDP
Chart 4b. U.S. and the 1981 Recession: A Recovery that Looks ‘Job-full’: Unemployment Rate
Chart 4c. U.S. and the 1981 Recession: A Recovery that Looks ‘Job-full’: Log of Employment-to-Population Ratio
Does Okun’s Law hold outside the U.S.? In our NBER working paper, which was released this week and presented earlier at the IMF’s Annual Research Conference, we also report estimates of Okun’s Law for twenty advanced countries since 1980. While a stable Law fits the data for most countries, the coefficient in the relationship—the effect of a one-percent change in output on the unemployment rate—varies across countries. We estimate, for example, that the coefficient is –0.15 in Japan, –0.45 in the United States, and –0.85 in Spain. These differences reflect special features of national labor markets, such as Japan’s tradition of lifetime employment and the prevalence of temporary employment contracts in Spain.
The bottom-line: Okun’s Law is not as universal as the law of gravity (which has the same parameters in all advanced economies), but it is strong and stable by the standards of macroeconomics.
This post written by Laurence Ball, Daniel Leigh and Prakash Loungani.
Posted by Menzie Chinn at January 9, 2013 11:45 PMdigg this | reddit
Why was the data set for the research started with 1948? The BLS has data going back to 1884. My recollection is that Okun himself only used the data from 1945 to 1960 in his original work - which has also puzzled me.
I don't quite get why the data prior to the post-WW2 period is irrelevant.
Posted by: SecondLook at January 9, 2013 11:57 PM
Employment is slow to recover because the US economy is losing productive capacity. The economy has fallen into a sup-optimal trap since 2009.
The problem now is that the natural rate of unemployment has risen 3%. (graphs #4 and #5 in the link below)
My updated research has a new graph (#22) showing a new growth model and how the US economy did not grow in the wisest way.
My research shows there are deeper factors underlying why the coefficient would be different between different economies. The differences between economies may include being closer to maturity, being in a sub-optimal trap, having a higher effective labor share of income and a different position from the "balanced growth" point of capacity utilization.
More importantly, the US economy is now hitting a wall of peaking profit rates and constrained capacity utilization which affect the observations of Okun's law. Unemployment is not coming down below 7.0%. and capacity utilization is not going to rise above 79.5%. There is your "full stop" as Krugman puts it.
Posted by: Edward Lambert at January 10, 2013 01:01 AM
This is well explained and easy to understand. I noticed you did not get into why the current recovery is slower than previous ones. Maybe that would have cluttered the main thrust of this post...
Posted by: Anonymous at January 10, 2013 04:39 AM
We need to move towards more pronounced qualitative treatments that delve into the aspect of labor demand and supply issues that go beyond quantitative analysis; the fact that job crunch has led to an exponential growth in student loans does not signify that graduates have a lien on jobs once they pass out. The fundamental problem is that there is no mechanism by which demand for jobs would allow students to fall in queue or the other way around (supply of students that would make jobs possible), as there are a whole lot of other things in the fray that works against it. If growth is left unsupported by the supply of labor, we have finally an unsustainable growth, something that is only transitory in nature.
Posted by: Procyon Mukherjee at January 10, 2013 04:43 AM
With all due respect, just because Okun's Law did not break down one cannot say that the recoveries were not jobless. The reason they were called jobless recoveries was that jobs were not available. Full stop. There cannot be another definition.
Posted by: Luke Ignaczak at January 10, 2013 05:58 AM
I haven't pulled down their NBER working paper yet, but based on their summary a few things kind of jumped out at me. These are kind of obvious, so hopefully they will address them in their paper.
First, there isn't just one Okun's Law, but several different versions. The authors are using the version of Okun's Law that compares the unemployment rate to the gap between potential GDP and actual GDP. My take on Menzie's post was that Menzie was using the version of quarterly differences in real GDP and unemployment. So I think they are talking about two different things.
Second, potential GDP is something that is guesstimated and not known in the way that real GDP is known. And of course there are different ways to estimate potential GDP. Very long and very slow recoveries tend to undercut the confidence in measures of potential GDP. And potential GDP measures do not typically include expansions in the underground economy that tends to happen during long recessions.
Third, they don't really discuss the possibility that the unemployment rate itself might have it's own law. In other words, human nature may have a tendency to just drop out of the labor force after prolonged unemployment, and the fall in the observed unemployment rate might just be picking up a law of human nature. It would be interesting to see if their results hold up under other measures that align with a potential output gap; e.g., median unemployment duration, discouraged workers, changes in long term unemployed, etc.
Fourth, one implication of the finding is that productivity has also been constant during the period they studied (1949 to 2011).
Fifth, a lot of the growth in GDP since the recovery started has gone to those at the very top. It's a little hard to reconcile extremely uneven income growth with a stable output/unemployment relationship. And I think this is really at the heart of the problem. When people talk about the "jobless recovery" they are usually referring to the paucity of "real" jobs that don't involve flipping burgers and scrubbing rich folks' floors.
Finally, I completely agree with their statement: "...in most earlier recessions, the output trough was followed by a period of above-normal growth that pulled output back to its previous trend." One of my biggest complaints with Obama's economic team was that they wanted to follow a return to full employment from below the trend line...one might say asymptotically from below. That was a big mistake.
Posted by: 2slugbaits at January 10, 2013 06:05 AM
So, a "jobless" recovery, or as the guest posters put it, a "slow growth" recovery, what's the difference?
We aren't trying to categorize the recovery based on semantics, or define and evaluate recoveries by whether or not they follow Okun's law. The folks who are calling this a jobless recovery are not saying Okun's law isn't holding (unless I am very much mistaken - please link if you can find it, someone).
The point is unemployment is still well over NAIRU and thus this is a jobless recovery however you look at it because there are still millions of jobless workers!
Posted by: Nick at January 10, 2013 06:13 AM
That is all fine, for once let the economists be simple,move the GDPs up.
Posted by: ppcm at January 10, 2013 06:19 AM
"A study made by the U.S. Department of Labor itself found that "an increase in unemployment benefits leads to an increase in the duration of unemployment." This country can have as much unemployment as it wants to pay for."
Posted by: Ricardo at January 10, 2013 08:16 AM
“Jobless recovery” is the term that entered the language in the wake of the 1990 recession to describe the phenomenon of employment taking longer than normal to get back to its pre-recession peak. Like stagflation in the 70s, joblessness in this sense had not been seen since the Great Depression. Perhaps another choice of words would have been better, but the phenomenon is clear-cut, it is important, and jobless recovery is surely descriptive enough. Joblessness is one of the two or three most important economic concepts to the man on the street. Wages or income or standard of living being the other concepts. All else is derivative and of second order importance to the average person. Jobs are primary.
Jobless recovery is a short-term phenomenon. It’s specifically related to recoveries and some indeterminate period beyond if the joblessness is bad enough. Recoveries are from the trough of the recession until real GDP regains its prior peak. Some, though not a lot, of attention was paid to this phenomenon in the 1990 recovery. Much more was paid to it following the 2000 recession as that recession was more jobless. But a search for this term shows economists all over the map in their explanations. Joblessness has sequentially worsened from recession to recession over the past 3 decades. This sequential worsening is what needs explication. The authors of this paper advance our understanding of this phenomenon – one that ranks right up there with stagflation – not one jot or tittle.
Of course Krugman and a thousand other economists are correct in claiming that slow growth is the predominant causal factor of the extent of joblessness. And because the focus of jobless recovery is on the recovery period and its immediate aftermath, we are dealing with a short-term or cyclical concept. In the short-term, insufficient aggregate demand is always the answer. However it is not the full answer as recessions are well understood to be processes that purge prior excesses that have built up. Notably (as just one example), excess labor in construction and finance that appeared once the housing bubble burst are being purged in the wake of this latest recession. This rasping and scraping is a necessary, healthy though painful phenomenon in an ever evolving modern economy. It has been with us all recessions before. What has not always been with us is the sequential lengthening of joblessness we now see.
Saying this is due to a shortfall of aggregate demand is near tautological. It explains little. The real question is: Why have these episodes of demand shortfall resulted in sequentially longer intervals from the time GDP reattains its prior peak to when employment regains it prior peak. The sequence from 1975 to present expressed in quarters looks like this: 1, 2, 5, 7, and this time 15 (at current rate of employment growth). Of course since the output gap approaches zero around the time payroll employment reattains its prior peak, explaining employment via its relationship with the output gap is as vacuous of explanatory power as is the aggregate demand explanation.
Here let us leave this paper and strike out into the territory of real explanation. The data is in front of us. How to explain it? Viewing the situation from the standpoint of productivity doesn’t get us very far, though admittedly it may open some doors. For if we say full employment arises from adequate growth of manhours and productivity, which by definition equals GDP growth, we have hardly come any further. The question might then devolve onto productivity. Why has productivity rather than jobs sequentially been filling more of the recession induced gaps since 1975?
Joblessness is a pressing topic, both in importance to households and when measured in millions of jobs not created except over evermore frighteningly long spans. For an effect this big, there must be a proportionately large cause. This is not to say there are not other causes, but it is surely how the greatest of scientists throughout history approached their problem. There is only one variable of which I am aware that meets this condition – the trade deficit. What’s happening in the eurozone powerfully confirms this beyond the shores of the US. As the current account deficit must algebraically conform to a precisely equal lack of savings in the deficit country, we are led to the nation’s surplus. The surplus arises jointly from businesses, households, and government. The surplus as a percent of national income was stable in the 50s and 60s at around 10%, and then in the mid-70s began to steadily decline, with decline accelerating this past 15 years. The predominate component driving the decline being the government deficit.
Trillion dollar government deficits (and the cumulation of deficits gone before) are with first order effect causing the current record jobless recovery. One of the main transmission mechanisms has been and still is offshoring. Not only does the public not understand this, by and large not even economists understand it. It cuts a gaping hole in the soft theoretical underbelly of Keynesian deficit spending.
Moreover we must at some point transition over from the short-run to the long-run. The cumulation of all these critical variables has pushed the US economy into the Reinhart Rogoff region. Specifically, the cumulation of all the annual fiscal deficits has taken the federal debt to unprecedented highs. The debt now lays heavy on potential growth, squashing it. A good surgeon would go in and remove the cancerous growth of too much government spending. A lesser surgeon would opt to go in and raise taxes. We know, however, from the empirical record that the latter procedure only validates even bigger government. As government is wholly supported by private sector production, and as taxes extract vitality from the private sector, and as the deficit and now the debt must be addressed to remedy our original problem – the one the public is instinctively and correctly most concerned with, namely joblessness – it is past time for the good surgeon to go to work and for those recommending more bloodletting of the private sector via higher taxes to be put out to pasture.
Alas, it will not be. In a democracy where enough voters are ignorant of what’s laid out here, where they then in their own narrow self-interest vote more government perquisites to themselves, where the representatives they vote for are more self-interested in staying in power, and where by a 70-to-30 ratio economists are ideologically driven by bastardized Keynesianism and themselves being either directly or indirectly in the employ of one or another government including grant money where one hand washes the other … all this is why the 6-decade long slide in real growth is going to continue apace. The end point – and this in the absence of one and only one thing – a grassroots movement informed by a real understanding of these goings-on – that endpoint some decades from now is a grinding toward zero of the rate of real per capita growth. Recoveries in the future will be even more jobless. And at some point a threshold will be crossed beyond which lies major civil strife.
Posted by: JBH at January 10, 2013 09:16 AM
reply to 2slugbaits...
Your perspectives on the issue agree with my research...
In regards to your first point about the difference between potential real GDP and actual real GDP, my research matches this to the difference between capacity utilization and a labor share of income constraint on capacity utilization. This difference is bounded by the unemployment rate. This approach also reflects your fifth point of uneven income growth and output. I agree this is at the heart of the problem.
And in regards to Menzie Chinn's approach using the difference between real GDP and unemployment, my research shows that unemployment and real GDP both obey the same constraint of total unused available capacity for factor utilzation.
So the distinct angles of approach to Okun's law are unified in the equations of my research. (link in my above comment. Sorry that link doesn't have the updated bibliography.)
Posted by: Edward Lambert at January 10, 2013 09:52 AM
Excellent work, Edward.
Examining the reported labor productivity, profit rate and to GDP, and private investment and depreciation rates, it appears that employment peaked in Q1-Q2 '12 and has been overstated since, confirming your conclusions.
Profits have peaked with employment and investment, implying another cyclical reduction in costs, employment, investment, and production is ahead to conform to the emerging trend rate of real GDP around 0-0.6% and negative per capita.
Financial profits ("rentier taxes" of $2,500 per capita) are again back to around 50% of total profits and at a record high as a share of GDP at 5%. This is like the Mafia firebombing your house and executing your family, the gov't paying them for the fuel and bullets the gangsters used to do the job, and then giving the Mafia's contractors the job of rebuilding your house to sell to someone else at a state-sponsored profit.
If the banksters are doing G-d's work, it's time for a new god, or else the god will destroy us.
Regarding labor, the regressive 14% payroll tax per worker, equivalent $8,700/year compounding interest claim, i.e., "rentier tax", per employed American, and debilitating equivalent $19,400 per employment person for total US spending for "health care", i.e., financialized private insurance premia, copayments, and coinsurance, is resulting in prohibitive costs for labor ("destruction of labor product") and a primary factor resulting in no employment growth per capita, offshoring, and accelerating automation of paid employment and loss of purchasing power.
Reiterating, in terms of equivalency, it costs the US 73% of an employed person's average labor product for payroll taxes (including what the employer pays), "health care" (insurance), and household debt service ("rentier tax").
IOW, before any net labor product is produced or realized by a worker, an average per worker cost of $35,000 for payroll taxes, "health care", and household debt service is required.
If we had wanted to destroy labor product and purchasing power, we could not have designed a more optimal means for doing so.
No doubt then why the US has not created a net new full-time private sector job per capita since the mid- to late 1970s to early to mid-1980s (when US crude oil production peaked and began to decline and US deindustrialization and financialization commenced in earnest).
Jobless "recovery"? How about joblessness and loss of income and purchasing power for tens of millions of Americans hereafter as the "new norm"?
If only eCONomists working for those doing G-d's work were to be included in the many tens of millions losing their jobs and purchasing power, a more realistic examination of the current once-in-history phenomenon might occur and result in proposing alternatives to paid labor for purchasing power and labor and rentier taxes on labor. One can hope . . .
Posted by: Bruce Carman at January 10, 2013 10:06 AM
JBH, also consider the critical event that occurred in 1970-85, which few eCONomists, politicians, or financial media pundits discuss: The US reached peak domestic production of crude oil after which production has fallen respectively per capita 60% and 50% since 1970 and 1985. By the mid-1980s, the US could no longer afford an industrial economy, which resulted in desindustrialization and financialization.
Since then, total credit market debt owed has increased by $42 trillion, whereas debt service and payroll taxes have increased per employed American along with total local, state, and federal gov't spending rising as a share of GDP to compensate for the loss of crude oil and goods production and increasing returns to labor's share of GDP.
The once-in-a-lifetime positive supply-side reflationary effects to asset prices, profits, and consumption from falling nominal interest rates of the Long Wave Downwave are over. Increasing debt to income and GDP, even at low nominal rates, will no longer result in reflationary growth real GDP per capita, profits, production, GDP, and gov't receipts. A demand-side regime is now required, including increasing returns to labor's share of GDP vs. financial capital's share.
But the high price of oil, debt service to incomes, payroll taxes, accelerating automation of labor, and medical insurance costs are destroying the purchasing power of labor at the same time Boomer demographic drag effects are bearing down on the rate of consumer spending and increasing the rate of drawdown of elder transfers.
There are no conventional supply-side, Keynesian, monetarist, communist, socialist, corporate-statist textbook solutions to the convergent cumulative effects of the aforementioned factors.
A radical reset and redefinition of the labor division, work, money, taxes, ownership of the means of production, income distribution, immigration, and role and function of gov't is required to avoid what you correctly suggest is the likely undesirable outcome.
Posted by: Bruce Carman at January 10, 2013 11:03 AM
That was a thoughtful comment, Slugs.
Posted by: Steven Kopits at January 10, 2013 11:22 AM
i agree with most here. Its correct that recent recoveries have been weaker. I looked in detail at this about 10 years ago concluded for the recent recoveries there is
1 Less (no) unintended inventory accumulation cycle less bounceback, and a muted industrial production cycle.
2 prompter countercyclical respnse to inflation less inflation cycle.
However while the latest recovery rate is in line with the two prior recoveries, this does not account for severity of downturn and excess capacity. As % of downturn speed and depth this recovery looks real weak, compared to last two.
reflects the standard pattern of weak recovery after banking crisis,
also both the corporate and housing capital stock remained with huge excess capacity, mostly not worked of yet
these two can be standard broadly accepted modelling drivers. where they will end, the post banking malaise is said by But i think also an even more broken unreformed unreformed financial system, and inadequate total supply including the collapse of the asset backed fixed income market with only small signs of recovery.
Posted by: AWH at January 10, 2013 12:09 PM
Basically what happens is, it takes longer for the "above average" growth to get rolling than in the past.
Considering this was a bigger recession than the last 2, it is taking longer to get the ball rolling......little surprise. IMO, it is because the system is credit based rather than wage or earnings based.
Once debt servicing falls far enough that enough credit can be produced, all confidence will surge and above trend growth will take place. We are seeing it already with increased borrowing, yet still to far low for decisive recovery.......yet. Watch the next 18 months however.........
Capital likes this system because it holds down labor and gives them easy access to public funds through easy new debt creation in times of crisis. Labor can tolerate it because no deep recession and a social insurance structure that supports them in times of crisis. Finance loves it because they run it as much as they did the old gold standard system before its collapse and the tough restrictions during the BW's era.
The problem comes is the government cannot issue "new" debt. Then the whole thing dies. No way to bail it out, it just collapses. This imo, is where alot of free market liberals get confused. It isn't existing debt that matters. I could care less if the US has 16 trillion or 10000 trillion of debt. That is all the price of doing business. It is new debt creation that counts. It is the fireman.
Posted by: The Rage at January 10, 2013 12:37 PM
Bruce: I concur with your point about the impact of peak oil. In particular as oil price controls were in place until Reagan, the growing dependence on foreign oil set in motion one strain of the long rise in the current account deficit. Indeed, policy could operate directly on domestic oil and gas production today to shrink the trade deficit and at the same time create jobs. There is something of the cat having ahold of its own tail here as you point out. The decline of domestic oil production shaved the growth of national income resulting in the personal saving rate double-peaking across 1974 and 1982, and then plunging. Households acted the only way they could to maintain consumption standards – buying more on credit. At the same time, as the nation’s belt tightened there was an impulse from voters for government to rescue them. Debt rose from this source as well.
You next mention the need to increase labor’s share with which I also concur. Here Kalecki’s insight from rearranging the NIPA accounting identity comes into play. Profits must equal investment plus the government deficit minus the trade deficit (saving of foreigners) minus household savings. The share of profits is the inverse compliment of labor’s share. The profit share is at an all-time high. Labor’s share, whatever it may be, must be in direct (and substantial) relation to wages. A couple months ago, year-over-year wage increases made an historic low. Wages are the largest component of disposable income. As income goes so goes consumer spending (and therefore GDP), because the long buildup of household debt has finally constrained household ability to augment income with credit as never before.
What stands out, other things equal, is that reducing the government deficit will lower profit’s share and therefore raise labor’s share. We’re looking at a trillion dollars a year here. Deficit reduction is thus doubly commendable; wages will rise and society’s resources will reallocate toward more sustainable market-determined output in the private sector. As the pool of saving out of income grows, this will have a causal effect on the need for foreign savings. In others words, the trade deficit will begin to shrink and that will inhibit the offshore hemorrhaging of jobs and create jobs domestically in both export and import-competing industries. I carefully stayed away from actions to reverse Chinese currency manipulation in my previous post, but this is another policy lever that could (and should) be activated to hasten the US job creation process along.
“There are no conventional supply-side, Keynesian, monetarist, communist, socialist, corporate-statist textbook solutions to the convergent cumulative effects of the aforementioned factors.” I am a mere tyro in my understanding of the Austrian school. Yet its proponents have the most sensible answer, just as all the schools you mention assuredly do not. Cut deficits, cut debt, cut government spending, cut taxes, cut regulations, free up domestic petroleum production, quit bashing business, and move toward the position of a healthy national surplus. The reigning paradigm of economics, lack of understanding on the part of the grass roots, and lack of political will are the deeper-seated obstacles in the way. And let me not leave out belief in the ideological goal of redistribution that trumps the economics of growth.
Posted by: JBH at January 10, 2013 01:43 PM
"I noticed you did not get into why the current recovery is slower than previous ones"
Because at the lower end of the wage scale, it makes more sense to get handouts than work, which ultimately lowers our national productive growth capacity. More handouts further reduce the incentive to work and as such our economy continues to grow suboptimally.
Posted by: Anonymous at January 10, 2013 02:42 PM
JBH: "The debt now lays heavy on potential growth, squashing it."
Whatever are you talking about? Interest on the debt is at the lowest level since WWII. How is is possible that the lowest debt burden in 50 years is crushing growth now?
Posted by: Joseph at January 10, 2013 03:01 PM
Luke, it's not that they were conforming to Okun's law by being recoveries and not jobless; they were jobless. It's that they were conforming to Okun's law by not being recoveries, or not much of a recovery, anyway.
Posted by: J Sterling at January 10, 2013 03:06 PM
excuse me. I omitted part of sentence above. The post crises banking malaises "are said by Rogoff and Reinhart to typically last only about five years. we are now 4 years into this one?"
Posted by: AWH at January 10, 2013 03:15 PM
One can, with moderate accuracy, measure the GDP and the unemployment rate, but not the potential GDP. Okun's Law seems to be less of a law and more of a definition of potential GDP.
Posted by: Kaleberg at January 10, 2013 06:02 PM
- Most of the net jobs created since the tech crash have been in health care and government, financed by rapid growth in government debt. Health care is headed north of 20% of GDP. Sustainable?
- Does the law hold using broader measures of unemployment? (discouraged workers? "disabled" workers?)
Posted by: Larry at January 10, 2013 08:39 PM
I am asking who might be interested in working with me to publish my independent research. I talked with an acquaintance of mine today from the Associated Press. He says it must be academically published first before they will print about it. The equations do need to be published just the same.
Is anyone interested? My email is on this research document...
Posted by: Edward Lambert at January 11, 2013 12:16 AM
Joseph: Here's the starting point to help you answer your question:
Posted by: JBH at January 11, 2013 03:41 AM
Let's assume that your are correct, that off-shoring is causing large government deficits, that are causing slow recoveries. Why do you suggest the cure is to cut the deficit? It seems most logical to suggest policy that addresses off-shoring and tie fiscal policy to the actual result of off-shoring policy. By your own logic, simply forcing surplus would be a disaster.
Posted by: Craig Jackson at January 11, 2013 06:11 AM
The link provided to your research and email address is not operative.
Neither is mine
Posted by: ppcm at January 11, 2013 07:30 AM
JBH, that still doesn't explain your statement "The debt now lays heavy on potential growth, squashing it."
Really, you think that the burden of debt is responsible for slow growth now, at a time when the burden is as low as it has been for 50 years! Please, enlighten us with some mechanism for your statement in which the lowest debt burden in decades is slowing growth. Oh, and also explain how your mechanism includes the Reagan years when the burden of the debt was 200% of what it is now.
Posted by: Joseph at January 11, 2013 07:49 AM
"Unemployment is not coming down below 7.0%. and capacity utilization is not going to rise above 79.5%. There is your "full stop" as Krugman puts it."
How confident do you really think you can be in that 7%?
From a look at your paper, it seems you are basing this on fitting the data to what you say is a new trendline....since 2010. So you are basing this on 3 years data? When the previous 42 years of data suggested a natural rate of 4.3%?
I'm not buying it. Maybe something fundamental really changed in 2009-2010. At this point I think that's an interesting theory. But even if it did, your error bars in estimating that new limit would have to be huge.
And It seems to me the main thing that has changed is likely asset prices. Prices on housing and equities especially tend to have a strong demand feedback. At the same time, high levels of household debt also limited demand growth. But household balance sheets have already recovered quite a bit, and asset prices have plenty of room to recover as well. Demand growth may be slow here, but I doubt there is any wall there at 7%, or even 6% unemployment.
Posted by: acerimusdux at January 11, 2013 01:09 PM
Craig Jackson: Large government deficits are absorbing far too much national savings. In the 3rd quarter, gross saving (a flow) was $1,975 billion at an annual rate: businesses $2,099 billion, households $740 billion, and government dissaving of $864 billion. Gross domestic investment including government and housing was $2,555 billion. The shortfall was made up by rest-of-the world savings lent to us. The current account deficit was $478 billion (derived from a different source). By definition, the current account deficit (nearly all in trade) must equal the flow of foreign saving to the US.
From the top down, the nation’s surplus, its investment, and the new technology and innovations that get incorporated into the stock of capital are the main determinants of economic expansion. Without exception, from the dawn of history all civilizations and all empires fell when their surplus went negative. The difference between gross and net is what’s necessary to maintain the current stock of capital. Consumption of fixed capital – the depreciation that must be replaced on an ongoing basis to maintain the stock – was $2,020 billion. If there’s no addition over and above this (that is, positive net investment), growth ceases and history tells us what happens next. On a net basis the surplus (net national saving) went negative Obama’s first year and has been negative ever since. It is an iron-clad fact that unless the lack of surplus is addressed, some decades hence the US economy will collapse and with it our nation. Even if the entire $864 billion of government dissaving were eliminated today by budget balance, ignoring second order effects, net national saving would be only half the percentage of national income it was during the heydays of the 50s and 60s. And real growth would be hard pressed to reach even 3%. Currently the CBO projects a federal deficit on the order of $700 billion in 2022, and rising. This with real growth projected at 2.8%, which we’ll be lucky to see half to two-thirds of. If nothing else, I trust this will help you gain respect for the surplus. Along with the price system and productivity, the surplus is one of the most important concepts in economics.
Offshoring is not causing our budget deficits in a primary sense. Causality is predominately the other way around. Algebraically, if households and businesses maintain status quo with their saving, the government deficit drives the current account deficit (again, mostly trade). Hence it is the government deficit that must be operated on. Shrink government and production will return to our shores. This does not say offshoring should not also be directly addressed. I do not know the elasticity of offshoring, but clearly the sign is negative. Any change that keeps more production on our shores raises economic growth, and flows through to higher tax revenues and shrinkage of the budget deficit (therefore less need for foreign saving). Moreover we are now in uncharted waters with the federal debt and with our debt to foreigners. Both must be reduced as a % of GDP for prosperity to return. Work through all this and you will conclude that rather than “disaster,” moving toward surplus is salubrious and absolutely vital.
Suppose Obama forces more deficit spending like Krugman and others want. Many changes will take place as a consequence of the additional dissaving. Think in terms of force vectors. Businesses will be pressured to save more. Households as well. There will be need for more foreign saving. The trade deficit will worsen. The tendency will be to offshore more. The rate of investment will be under pressure to contract. The federal debt will rise along with expectations for such in the out years. The interest cost on the debt will go up, further widening the deficit. The timing of the next sovereign debt downgrade will move forward. Additional uncertainty will be created. And most importantly the national surplus will go further into the hole. Of course the stimulus will have an offsetting impact on these vectors. But right off the top some stimulus will dissipate via imports. What remains will have to have enough of what we might call joint positive consequences on GDP to offset the host of countervailing forces just described. Even then the near-term positive impact would quickly (within quarters) dissipate. After which the negatives would come back into their own and have an even greater dampening effect than they have now.
Posted by: JBH at January 11, 2013 01:46 PM
Coincidence Alert: In response to Menzie's previous post several posters suggested that economists associated with non-Ph.D granting institutions such as Brookings and the Urban Institute were second rate economists. By a happy coincidence Menzie's subsequent post was all about Okun's Law. For what it's worth, Arthur Okun was associated with Brookings.
JBH: What stands out, other things equal, is that reducing the government deficit will lower profit’s share and therefore raise labor’s share. We’re looking at a trillion dollars a year here. Deficit reduction is thus doubly commendable; wages will rise and society’s resources will reallocate toward more sustainable market-determined output in the private sector.
This is crackpot economics. Deficit reduction when the economy is at full employment would be a good thing. Deficit reduction when there is a persistent output gap simply reduces demand for labor and drives down wages...or to the extent that wages are downwardly rigid deficit reduction increases unemployment.
As the pool of saving out of income grows, this will have a causal effect on the need for foreign savings.
More crackpot economics. The pool of private saving is already quite large. That saving is looking for a place to land. Given weak private sector investment demand, the borrower of last resort must be the government. The private sector has been increasing private saving as the private sector seeks to repair balance sheets. But the private sector cannot increase saving unless some other agent is willing and able to absorb that higher saving.
I am a mere tyro in my understanding of the Austrian school. Yet its proponents have the most sensible answer, just as all the schools you mention assuredly do not. Cut deficits, cut debt, cut government spending, cut taxes, cut regulations, free up domestic petroleum production, quit bashing business, and move toward the position of a healthy national surplus.
Here we go from merely crackpot economics all the way to Austrian voodoo mumbo-jumbo. There's a reason why Austrian economics is fringe.
Posted by: 2slugbaits at January 11, 2013 02:23 PM
Joseph, I specifically said debt not burden of the debt. By burden I take it you mean interest cost. That at the present time is not a significant factor for the US, nor was it in the Reagan era. However it is a nice comparison. The interest rate dynamic then versus now was quite different. The market knew by 1983 that rates were plunging from the artificial highs engineered by Volcker to break inflation, and that they would go much lower. It could thus look into the future with little trepidation. Today things are the exact opposite. Rates are at historic lows and they are artificial. By 2022 the interest cost on the much larger debt (as a percent of GDP) will approach $1 trillion. It only takes the yield curve moving a couple points higher, and the much higher debt we’ll have in 2022 ($16 trillion today going to $20 to $25 trillion).
Having cleared up this misunderstanding, I’d simply restate that the debt lays heavy on potential growth already now as well as in the future. Again, read the Rogoff papers. Along the way collect the various transmission mechanisms (from debt-beyond-the-danger-threshold to GDP growth) they mention and you will have the answer you seek.
Posted by: JBH at January 11, 2013 03:14 PM
JBH: "I’d simply restate that the debt lays heavy on potential growth already now as well as in the future. Again, read the Rogoff papers."
I am familiar with the Rogoff papers. He says that, with very high debt levels, maybe, someday, in the future that the bond vigilantes might come charging over the hill and we're all gonna die. (Although, I doubt Rogoff ever imagined that leading the charge would be irresponsible idiot Republicans deliberately defaulting on 200 years of the full faith and credit of the U.S. Government.) However, that time is not "now," contrary to your claim, and there are still no signs on the horizon. Meanwhile, our debt burden in near record lows.
Although you linked to Rogoff's page, you apparently didn't read or don't understand his papers. Here is a link to an article written by Reinhart and Rogoff at Bloomberg that explains it in layman's terms so that even you might be able to understand.
Rogoff takes folks like Hassett, Hubbard and Taylor to the woodshed for deliberately "grossly misinterpreting" the Reinhard/Rogoff papers for disingenuous political purposes. He states that the current recovery is typical of systemic financial crises around the world that they studied. He says the typical recovery from this type of crisis takes 4 to 10 years and is more U or L shaped than V shaped. He further says the the current recovery is "positively brisk" compared to the typical U.S. recovery from a systemic financial crisis.
Sorry, but your entire premise about current debt slowing growth is reputed by the very sources you cite.
Posted by: Joseph at January 11, 2013 08:41 PM
You are looking at all the right variables and relationships, but what you seem to be missing is that investment is demand driven. Eliminating $864 billion in government dissaving would not increase our net national investment, it would lower it.
The only way that investment could benefit from removing government demand for savings is that this would tend to lower interest rates. But how much lower do you want them to go? Clearly, they are as low as they are because investment is not profitable. This is a symptom of a global shortfall in demand. Lower interest rates at this point are not going to drive investment. The truth is, historically, strong investment tends to be strongly correlated with higher, not lower nominal interest rates.
The biggest mistake Obama made was to focus too much on short term stimulus, such as transfer payments and tax cuts, and not enough on public investment. The stimulus package only included about $100B in infrastructure investments. We could use $1T to $2T.
That type of spending tends to attract, not crowd out, private investment. It also maintains demand in the meantime, and sustains a well paid, well trained, productive workforce which makes private hiring more attractive.
There are only two ways to improve our trade balance. One is to devalue the currency, improving exchange rates. The other is to make this country a more productive and profitable place for companies to invest. Simply cutting government deficit spending might tend to send both of those variables in the wrong direction.
Posted by: acerimusdux at January 12, 2013 02:03 AM
JBH The debt held by the public is $11.6T. You need to quit citing this $16T figure, which is irrelevant to the discussion.
As Joseph says, you have badly misunderstood the R&R paper. While Rogoff is not particularly keen on any new fiscal stimulus, he has made it quite clear that he does not support contractionary policies either. Rogoff's views are a lot closer to Krugman's than yours.
Now you're right that the debt we have already incurred will be a drag on future growth once we fully emerge from the recession. That's just a fact of life. It's also one of the reasons why many of us argued against the Bush tax cuts in the early 2000s. Bush ran structural deficits at full employment, which ate into any headspace we would need in the event of an economic downturn. But what's done is done. The relevant question is where do we go from here? The first priority has to be getting the economy back to full employment. Austerity doesn't get us there. In fact, austerity just drags things out and ultimately increases the debt. Once we have returned to full employment, then we need to raise taxes. If you want to cut back spending, then fine; but tell us what you want to cut back on. Defense? Veterans affairs? Big Ag? Those are the big discretionay programs. They are also ones that are near-and-dear to the hearts of Red State politicians. Or are you going to repeat your earlier claim that the way to reduce the debt is to cut taxes?
But since you're so worried about interest payments, does this mean you will be writing your congress critter demanding that he or she unconditionally vote to raise the debt ceiling? Right now the biggest threat to interest rates isn't the debt; it's whacked out Tea Party politicians making irresponsible noises about letting the US govt default on bond payments and Treasury disbursements.
Posted by: 2slugbaits at January 12, 2013 07:48 AM
I don't think anybody is talking about default. The point is that there is enough revenue to pay obligations to creditors. Social Security and Medicare/caid can't file a default suit against the Federal government for refusing make a payment to the federal entitlement fund.
I have no idea why so many on the left want to take the easy way out and raise tax rates on everyone and every business. Raising tax rates should always be the last resort, not the first.
Cap deductions for everybody at the same levels they are capped for the "rich" in The American Taxpayer Relief Act. The result would be that some of those who avoid federal income tax (see 47%) through the use of extraordinary claims will have to pay some federal income taxe, thereby broadening the base. An alternative would be to set a single cap for all taxpayers and let them claim whatever they want. I like that approach better because it preserves mortgage, charity, municipal bond, etc deductions.
Create a bipartisan commission to eliminate wasteful spending and redundant programs. Coburn has estimated this can save $100's of billions a year.
Raise the 'retirement' age 2 years for full benefits, add means testing (entitlements are not a pension account, it's a transfer). Align and flatten the coroporate and small business tax rates, then cap deductions and/or eliminate others. There are other simple steps floating around that can in total reduce the projected increase in entitlement spending by a few hundred billion per year.
The steps above will go a long way toward reducing the deficit to the point where we will have a surplus during expansion years with ~ 2.5% and above GDP growth.
Posted by: tj at January 12, 2013 10:03 AM
tj: "I don't think anybody is talking about default. The point is that there is enough revenue to pay obligations to creditors."
Apparently you are unfamiliar with the financial industry's practice of "universal default." If you don't pay your electric bill, the bank that issues your credit card jacks up your interest rate. This practice was recently limited by law for credit cards, but it certainly isn't for the bond market.
Just because you can pay interest on bonds does not mean that interest rates won't skyrocket as investors decide that not paying your other bills makes you an unreliable credit risk.
Posted by: Joseph at January 12, 2013 11:42 AM
Joseph: OK, suppose Reinhart and Rogoff were wrong. What would you do? You have only one answer. Fiscal stimulus in the form of more spending. So say last year you ramped up spending by $1 trillion. With a generous impact multiplier of 1, and with the stimulus passing through in its entirety to nominal GDP last year, GDP would now be $16,776 billion instead of $15,776. And the deficit would be on the order of $2,089 billion instead of $1,089, with no more incremental stimulus to follow as only changes in the deficit can stimulate. Now you’ve not only taken the debt higher, in addition the $2 trillion deficits incrementally add another $1 trillion to the debt annually from now on. With no further deficit reduction in the out years, and assuming nominal growth of 4%, the debt rises to $37 trillion by 2022 (little different than the $25 trillion you can calculate from the CBO’s latest numbers plus the additional $10 trillion) taking the debt-to-GDP ratio to 149%! Beyond Greece you might say.
Between now and then two things would happen. The US credit rating would be downgraded many notches. And the interest cost of the debt would soar (as already it is going to). Conservatively, interest rates across the yield curve will rise a minimum of 2 points well before 2022. This is less than what the CBO currently projects in the current circumstances. The interest rate on the debt today is 2.2%, and will go to 4.2% at a minimum. 4.2% on $37 trillion is $1.5 trillion, which means the deficit in fact would really be more on the order of $3 trillion in 2022. In other words, your Keynesian stimulus proposal is explosive. Nor would deficit reduction from the assumed $2 trillion deficit last year alter things much. With annual deficit reduction of $100 billion per year – and ignoring the interest cost part of the equation which as a separate matter will still grow to be huge – the debt-to-GDP ratio would be around 130% in 2022. Moreover, assuming a multiplier of ½ and constant 2% inflation, real growth would fall to steady state 1 3/4ths percent. All of this is precisely Reinhart Rogoff’s point.
At the 13th quarter out, this is the weakest recovery in a century. Real GDP has mustered only 2.2% growth, and that in the historically strongest part of the expansion. The inventory rebound is now over. Moreover, this recovery has been far weaker relative to other recoveries if adjusted for depth of recession, as its trough was the deepest since the Depression. Comparable recovery growth rates after 1974 and 1982 were 5.4 and 5.7%. Furthermore, this pitiful 2.2% was accomplished only with the aid of historically record monetary and fiscal stimulus. There is now no more ammunition left without creating an even more debilitating debt explosion. Reinhart and Rogoff’s results have already been validated by the 3 years of empirical data already in. That this was the biggest financial crisis since the Depression is a fact. The conjunction of now record household and sovereign debt is a fact. Reinhart Rogoff found that in past circumstances like this debt constrained growth. That is an empirical observation. Few of those psychologically disposed to swim in the school of liberal groupthink are able to step outside and view this hard empirical evidence critically and with an open and independent mind.
Posted by: JBH at January 13, 2013 07:20 AM
JBH: Two observations. (1) To my knowledge, there is no quarterly GDP series that extends back a century; how do you know that this is the weakest expansion in GDP terms (easily falsifiable for other indicators)? (2) you do know that Japanese gross debt-to-GDP is something in excess of 230%?
Posted by: Menzie Chinn at January 13, 2013 09:10 AM
Acerimusdux: I well understand that investment is demand driven. Expectations about future economic growth affect the willingness of businesses to ramp up net investment. I do not for a moment propose eliminating $864 billion of government dissaving in one fell swoop. That would result in a recession worse than the last one. Rather, I view the cuts spread over time at the rate of $100 billion or more annually. This would shave on the order of 3/10ths of a percent off real growth until gross debt fell back below 90% of GDP. This is the price that must be paid for letting debt get too high.
Low interest rates are indeed a symptom of the global shortfall in demand. The Fed has pinned the short end of the yield curve to zero, and along with the flow from the eurozone the entire curve is compacted down. For the next couple years interest rates can effectively be treated as a constant, and thus neutral in regard to incremental business investment decisions of a short-term nature. However, the average age of business capital is around 9 years. Hence businesses surely look far into the future to anticipate whether or not today’s project reasonably promises to pay off. Higher uncertainty about the future foreshortens the life of projects businesses would be willing to go forward with today. This is so regardless of today’s low cost of debt financing, except for the concern that today’s artificially low rates will have to rise in the out years to combat inflation, and therefore slow aggregate demand in those out years. So policy has set up a condition where uncertainty is high, future demand is going to be constrained by rates that have nowhere to go but up, and any further additions to deficit spending will find the US even more reliant on foreign savings and suffering the many consequences of an even greater trade imbalance. Unless both the deficit-to-GDP and debt-to-GDP ratios stabilize or fall – amounting to the perception by rating agencies that the US is getting its fiscal house in order – the risk premium on “risk free” US Treasuries will rise. The US is not going to default on its debt even if Congress refuses to cut spending. A different consequence will prevail.
In an important BIS paper in 2010, The Future of Public Debt, Cecchetti et al plot out the dismal path ahead. “Unless the stance of fiscal policy changes, or age-related spending is cut, by 2020 the primary deficit-to-GDP ratio will rise to 13% in Ireland; 8–10% in Japan, Spain, the United Kingdom and the United States; and 3–7% in Austria, Belgium, Germany, Greece, the Netherlands and Portugal.”
Their graph 4 is the visual heart of it. In the baseline case, they assume that government total revenue and non-age related primary spending remain a constant percentage of GDP at the 2011 level (OECD projected numbers). In this baseline case, the US debt-to-GDP ratio exceeds 150% in the coming decade. “Without a change in policy, the path is unstable. … Seeing that the status quo is untenable, countries are embarking on fiscal consolidation plans. In the United States the aim is to bring the total federal budget deficit down from 11% to 4% of GDP by 2015. … To examine the long-run implications of a gradual fiscal adjustment similar to the ones being proposed, we project the debt ratio assuming that the primary balance improves by 1 percentage point of GDP in each year for five years starting in 2012.” This scenario 2 translates into cuts on the order of $170 billion this year, and larger cuts each year beyond. Still with annual deficit reductions like this, the debt ratio (Cecchetti uses gross debt as do Reinhart and Rogoff) will reach 130% in 2020, and then rise exponentially. The reason for this exponential rise is interest costs.
There’s a host of factors on the horizon even if as-yet-not-evident foresight and political will start right away cutting the deficit by 1% of GDP. Amongst these: the probability that a future shock will trigger unstable debt dynamics; higher risk premia on debt; potentially lower long-term growth as a larger share of society’s resources is permanently spent servicing the debt; larger borrowing from abroad which means domestic income is reduced by interest paid to foreigners; reduction in both the size and effectiveness of any future fiscal response to an adverse shock; the potential triggering of a sudden increase in long-term inflation expectations; and in the absence of fiscal tightening, that monetary policy may ultimately become impotent to control inflation, as policy rate increases lead to higher interest on the public debt and to a cascade of higher debt.
All this and more must be weighed against the opposing desire to put off fiscal consolidation until the economy is back close to full-employment. Without a doubt inflation will start to rise as the economy approaches full-employment. The timing of course is in question. But by 2016, I expect the Fed’s now-higher inflation target to be surpassed. Moreover, I expect them to move the goalposts above 3%. Global investors will not like this. The consensus already expects higher rates 12 months from now, with a steady rise thereafter. The great debate in economics today is between those whose theories were so deficient they could not predict the greatest crisis since the Depression, and those who recognize that debt matters tremendously, and that on current trajectory we have already passed a tipping point.
Posted by: JBH at January 13, 2013 01:17 PM
JBH The conjunction of now record household and sovereign debt is a fact.
No, this is a Fox News Faux Fact. Private debt has been coming down as households and businesses deleverage. Indeed, one of the arguments for expanding government debt is to provide households and businesses with the financial assets they demand in order to continue deleveraging. If no one wants to borrow their desired savings, then the money just sits in a vault or under a mattress. In either event the saving leaks out of the income flow...and since saving is a flow variable the loss is permanent and nonrecoverable. You seem to have a very difficult time understanding the simple intuition that Person A cannot save unless Person B is willing to borrow. I suspect it's because your understanding of macro is that it's just an extension of your personal household experience.
And the deficit would be on the order of $2,089 billion instead of $1,089, with no more incremental stimulus to follow as only changes in the deficit can stimulate.
Huh? This is complete nonsense. You wouldn't need stimulus once the economy got back to something like full employment. No one is arguing for permanent deficit financing...except perhaps the Tea Party nuts who don't understand that austerity policies coupled with tax cuts increase both the cyclical and structural deficits. The objective is to unwind the long run deficit, and you do that by bringing the economy up to full employment by shocking aggregate demand. And once the economy is at equilibrium output you start running (at least) primary surpluses.
With no further deficit reduction in the out years, and assuming nominal growth of 4%, the debt rises to $37 trillion by 2022
There is so much wrong with this statement that I hardly know where to start. First, debt does not grow with nominal GDP growth. Debt reflects an obligation in the past. So your math is just goofy. Second, $37T in debt also represents $37T in assets. What you worry about is the interest rate, and short of default by not extending the debt ceiling, there is no plausible way that interest rates go to a minimum of 4.2% given a nominal growth rate of only 4%. If you are working off of a macroeconomic model that gives you that kind of a result, then you better rethink your macro model. Third, why do you assume no further deficit reduction in the outyears? Fiscal stimulus is not supposed to be permanent. Your whole scenario is a strawman. Try arguing against a real argument.
Furthermore, this pitiful 2.2% was accomplished only with the aid of historically record monetary and fiscal stimulus.
More Fox News Faux Facts. This recession started with interest already very near the zero lower bound. There was very little room for further interest rate easing. The fact that interest rates are lower than they were in 1982 does not mean monetary policy was as potent as it was in 1982. Going back to the old IS-LM model, back in 1982 the LM curve was relatively vertical and the investment response factor was fairly strong. In the Great Recession the LM curve was flat (the old liquidity trap) and the investment response factor was very weak. And you might want to double-check the econ literature with regard to your views on the depth of a recession and the speed of recovery.
Moreover, assuming a multiplier of ½ and constant 2% inflation, real growth would fall to steady state 1 3/4ths percent. All of this is precisely Reinhart Rogoff’s point.
Here's another interpretation. In the wake of financial recessions, households and businesses try to deleverage debt. As long as inflation is low, households and businesses find it hard to deleverage. As long as governments follow fiscal austerity, saving leaks from the income flow, which then pushes down aggregate demand. If you're a small country that doesn't control its own currency, then the fiscal response in R&R matches what you would expect under the old Mundell-Fleming IS-LM-BOP model with fixed exchange rates. Do you think that model applies to the US economy today?
Posted by: 2slugbaits at January 13, 2013 01:30 PM
Menzie: The relevant quarterly data can be found at: http://faculty-web.at.northwestern.edu/economics/gordon/researchhome.html. This is the Gordon-Krenn data set for 1913-54. We make do with the best we can find. I have longstanding and great respect for Robert Gordon’s work. This data set permits me to make my statement looking a century back. I know the postwar period like the back of my hand. Recovery out of the trough of the Depression and the ‘37 recession are no-brainers. That left only the 1913 and 1918 recessions. These too at the horizon of 13 quarters from the trough – the span of the current recovery through Q3 – had faster recoveries. What is important about comparing the full span of this recovery with the same interval in prior recessions is testing whether Reinhart Rogoff are right or not. They are not talking short-term stuff. Each quarter that goes by this recovery will suffer more in comparison to all previous.
Yep, I’m keenly aware of the Japanese gross debt-to-GDP ratio. Japan’s rose above 90% in 1995. Since base year 1994, Japanese real growth has averaged 0.8%. Their fiscal crisis was horrendous. Their debt ratio is now the world’s largest. Fortunately for them, rest-of-the-world growth has kept them afloat. Such will not be the case going forward for the US. This you might say is exhibit one in real time in Reinhart Rogoff’s case.
On a separate note, I’d like to thank you and Jim Hamilton for all the work you put into this site. You always post my comments promptly even though your quick perusal often finds them diametrically opposed to your thoughts. This is honorable, it is how science advances, and I for one want to commend you for it.
Posted by: JBH at January 13, 2013 02:57 PM
From your last post, we don't seem to be too much in disagreement as to the optimal path of policy in the long run. It is true for example, that an additional $1.4T in deficit reduction over the next decade would put us on a path to have the debt/GDP ratio under 75% by 2022. So maybe $140B a year. That would amount to about 3.5% of the current budget, about 0.9% of current GDP, and probably under 0.5% of annual GDP for the decade.
But what are the consequences really of not doing this? To me the consequence of an incrementally suboptimal policy will be an incrementally suboptimal result. I see little risk there of any great crisis.
Moreover, you seem to agree that any consequences will come in the future. So why shouldn't we wait to make the needed cuts until they are actually needed? You suggest that doing so now will increase confidence in the future. But you seem to ignore the risk on the other side, that cutting too much now will hurt confidence in the future. If one judges by the reaction of the markets, the fiscal cliff deal which caused this problem was quite popular there. Remember, under current law entering the year, there was no long term debt problem. Simply allowing the fiscal cliff and sequester cuts to occur at that point had us on a course for sustainable deficits without any additional action needed.
I would also point out, as far as the concern about record household and sovereign debt, that net debt in the economy must add up to zero. So lump foreign held debt in with private, and for there to be large public deficits, there must be a large private surplus. If there is an excess of financial debts, there must also be an excess of financial assets.
If it's gross debt levels that are the issue, well they tend to increase with economic activity. The only way to have more economic activity without debt is to increase the amount of currency available. This is also the solution if the issue is foreign holdings, or the trade deficit, being too high. Currency devaluation is achieved by increasing the supply of currency, also making economic activity possible with less reliance on debt expansion.
I have a hard time worrying too much about an uncertain future loss of faith in the currency when at present the currency, if anything, seems to be overly strong.
Posted by: acerimusdux at January 14, 2013 02:27 PM
reply to acerimusdux...
You are not buying that unemployment is stuck above 7%.
I would suggest that you go deeper into a study of growth models and steady state dynamics. They do reach a point where factor utilization is increasingly limited. Are you just not ready to accept that we arrived there so quickly due to the dynamics of the crisis?
Posted by: Edward Lambert at January 14, 2013 02:41 PM
I believe you are missing a few words in the top of this.
Using annual U.S. data from 1948 to 2011, we find that the Okun’s Law has a coefficient of –0.4 or –0.5, with an in the neighborhood of 0.8.
What is the R squared of the regression? Are you trying to say it is .8?
If you ran a regression with major economies put together, what would happen to the R squared (ie how much of Okun's law is country specific, and how much is not)?
Posted by: fladem at January 15, 2013 09:21 AM
No I can accept that it's possible, and I think it's an interesting theory. But I think it ends up being very difficult to say exactly what the current steady state might be (if there even is one).
Here's where I think you are maybe on the right track. I think that:
1. Record levels of wealth and income inequity may make it harder to sustain, in the absense of fiscal stimulus, sufficient aggregate demand needed to support previous levels of capacity utilization.
2. Demand was sustained for awhile in part due to asset bubbles in equities and then housing. It becomes difficult to reflate these bubbles once they've burst.
That said, I can't say I know how things will play out going forward. With unemployment already down to 7.8%, we could easily be under 7% within a year. I THINK we could have a down year in the stock market, after a good 3-year run, especially as the S&P 500 at around 1470 now could hit some resistance if it reaches the 2007 high around 1560. That plus what remains of the fiscal bluff could lead things to stall by the second half.
But there's also a reason they say "don't fight the Fed." If the Fed keeps up it's current pace of asset purchases, that would amount to $1T of monetary stimulus for the year. I can't say for sure they won't be able to reflate those bubbles some, which could provide enough fuel to demand.
And if I could reliably predict which direction the stock market will go I'd be a lot wealthier than I am. Predicting at exactly what unemployment level inflation might resurface is maybe only ten times as difficult.
Posted by: acerimusdux at January 15, 2013 07:44 PM
reply to acerimusdux...
If you could know with certainty that capacity utilization is not going above 79.4% and unemployment is not going below 7%, wouldn't that help you make money right now? ... Yes,
That is what my equations are saying.
The new capacity utilization number came out today and once again we see it stalling below 79%. The economy cannot improve because we are at the UT zero lower bound.
The economy sat at the UT zero lower bound from 2005 through 2007 before the crisis. Remember how capacity utilization sat at 80% for years. That was the UT zero lower bound.
Posted by: Edward Lambert at January 16, 2013 12:03 PM
Edward.... I think you may be falling victim to "false precision".
Your equations are using a regression analysis, a technique which doesn't ever tell anything with certainty, and which can at best only ever give an estimate based on how well some real world data best fits a model. In this case, you are looking at an estimate based on only a few years of real world data, and fitting that to a model which also may not exactly describe how the world works. The real world relationship may even not be linear. And some of the variables seem to be interdependent (like unemployment and unused capacity).
As for 2005-2007, the civilian unemployment rate at that time was below 5% (from 6/2005 to 12/2007 it was 5% or less). I would expect that the constraint at that time was labor, not capital or usable capacity.
In addition, some structural adjustment has already occurred since 2007 as well, as trade deficits are down about $150B a year from what they were then. That also might increase the level of capacity utilization which could be reached.
By the way, what was the standard error in the regression equation which estimated that 7.1% limit (graph 5)?
Posted by: acerimusdux at January 16, 2013 02:33 PM
reply to acerimusdux...
It is true. Regression analysis is weak in many things. A sine function gives an R2 close to 0. Yet, it has a definite pattern.
The long-run relationship of cu - cu(ls) shows a sine pattern. Even unemployment (u) shows a sine pattern. Yet, it is interesting that these two sine patterns reflect each other in the data. This sine reflection is what got me to develop the original UT equation... (UT not squared)
UT = u - (cu - cu(ls))
UT is non-negative in dynamics.
And you are correct about the data from 2005 to 2007. That low unemployment value was the constraint. Yet, the equation still kept UT positive. The UT values from 2005 to 2007 went like this (quarterly starting at 1Q-2005 to 1Q-2008) 18-16-17-14-14-9-11-16-17-10-10-12-20. These values sat near the UT zero lower bound during that whole time. Capacity utilization didn't budge up, and unemployment didn't budge down. Even the effective labor share constraint cu(ls) had a range during that time from 76.6% to 78.3%. All these terms "obeyed" the constraints for 3 years to keep UT a positive value.
Trade deficits won't effect the UT zero lower bound constraint. The UT ZLB held steady when the trade deficit was a lot lower in relationship to GDP. Imports still create jobs through "value added" employment. Capacity utilization is a function of internal capacity dynamics. Now, real GDP can still rise at the UT ZLB. The data shows that. However, the profit rates will peak. (see graph #8)
The standard deviation for UT data from 1967 to 2008 is 13%. From 2010 to 2012 it is 7.6%.
The standard deviation for unemployment data from 1967 to 2008 is 1.5%. From 1Q-2010 to 3Q-2012 it is 0.74%.
here is a link to the data. (The full data didn't upload, but the core values are there.)
Posted by: Edward Lambert at January 17, 2013 09:53 AM
Your equation for the labor share constraint especially makes no sense to me.
1. I don't think it makes sense to run the regression on smoothed data.
2. It makes no sense to to me to have a negative y-intercept. Capacity utilization can't fall below zero. It would be better to force the zero intercept.
3. You aren't reporting the 95% confidence interval, which I think is going to be quite large if you aren't forcing the y-intercept to zero.
4. Even if you do it right, keep in mind you are still estimating something which may not actually exist. You can only see if it works as a predictive model going forward.
Posted by: acerimusdux at January 18, 2013 03:25 PM
reply to acerimusdux...
1. why not?
2. It only has a negative y-intercept because labor share is taken from 2005=100. That number which itself can go over 100 (not possible in real terms) must be calibrated down. The -22.6 is just a calibration to bring the 2005=100 down to a number which can only exist between 0 and 1.
3. If you look at graphs 22 and 23, there is a sloping line from (0,0) to (1,1). This turns out to be the real estimated line that you see in graph 1. It's equation in graph 1 is...
cu = ls - 22.6.
It's equation in graphs 22 and 23 is...
cu = ls
In graphs 22 and 23, we finally see that there is no negative y-intercept.
4. If the data support a zero lower bound, there is evidence that it exists. If you look at the data over the years and understand the distortions in the governments numbers, the data supports that there is a constraint when UT hits zero.
As of right now, capacity utilization has hit the zero lower bound for the last year. Here are the quarterly numbers for cap util since 2Q-2011...
76.2, 77.1, 77.9, 78.7, 78.9, 78.6, 78.5
The numbers reflect a hitting against a constraint because they slow down and stop. and it just so happens that they did this at the UT zero lower bound.
Posted by: Edward Lambert at January 18, 2013 04:04 PM
1. Smoothing loses data. I'd rather feed actual data to the regression, and let the regression do the fitting or smoothing.
2. Yes I get that the -22.6 is a calibration, it's just one that's more inaccurate than it should be. You have a difference in scale there, so the calibration should be multiplying, not adding. The relationship is more like y = .78x.
When I run a regression with a forced zero intercept on the quarterly data since 1967, I get that equation with an R squared of .9975, a standard error of .003, and a 95% confidence that the coefficient is between .773 and .784. We KNOW that when LS is zero, CU will be zero. So nothing wrong with using that if it's that much more accurate.
Just using that to modify your UT equation, we have LS at 94.2, and TCU at 78.8, and unemployment at 7.8, so basically 7.8-78.8+(94.2*.78)=2.5, so current UT would be the square root of that, only about 1.6%.
Also, looking at the residuals from that regression, since 1967 TCU less predicted TCU has never been more than +/- 11 points, and it's never been above predicted by more than 7, so it would be a unique 50-year event if capacity utilization were to rise above 85% (and unlikely to rise even above 81%) in the next few years without labor share also climbing.
Again, I think you are on the right track here in general. I just think the way this is derived it seems more a rule of thumb than a precise or unchanging law.
Posted by: acerimusdux at January 19, 2013 05:28 AM
First, thank you so much for taking the time to evaluate the equation. It is greatly appreciated and you are a true professional by nature...
I like your correction to the equation... cu(ls)=0.78*ls
You make a mistake in using the equation, but it only a matter of getting the numbers right. You have this...
square root of 2.5 = 1.6%
I get this...
square root of 0.0288 = 0.1696
I compared your equation UT2=u-cu+(ls*0.78) and my equation UT2=u-cu+(ls-22.6) and the plots were very similar up until a few years ago when the labor share really dropped. Your equation shows that the economy has more room to maneuver at the moment than I thought. I was seeing UT=11%. Now I am seeing UT=18%.
I like your correction of the equation. Now I am going to rerun all the graphs using actual data using your equation. The natural rate of unemployment graph will be interesting.
I'll post the graphs for you to look at. If for some reason we lose contact here... my email is firstname.lastname@example.org.
Thanks again for your great constructive feedback.
Posted by: Edward Lambert at January 19, 2013 11:29 AM
reply to acerimusdux...
I re-ran the graphs using your equation...
cu(ls) = 0.78 * ls (ls=2005=100)
Here is the link to view the new graphs online...
Each graph highlighted with yellow just below the graph was changed.
Not many changes really, but there is more room for the economy than I thought before.
The natural rate of unemployment went up only 2.4% instead of 2.8%.
The profit rate came down and there is more room for the profit rate to increase.
For the graphs that weren't changed, you can just view UT = 18%, instead of 11%.
The last graph for fed funds rate didn't change much. I also changed it to a 2.5% inflation target, from 3.5%.
It's interesting that graphs #20 and #21 look almost identical.
It's also interesting to see that in graph #21, the equilibrium level for cu* actually rose from 87.6% to 87.9%. That means the economy is farther from equilibrium than before, but there is more room to maneuver to get back.
One important thing... You will notice that in graph #23, the yellow line does not end perfectly in the center of the dark blue line. It is very close though. This means that your 0.78 multiplier needs just a slight tweaking.
Now there are two equations to evaluate.
Posted by: Edward Lambert at January 19, 2013 03:04 PM
OK I misunderstood how you were using that, I thought you were plugging whole numbers for the percentages into that equation, rather than the decimal. Not sure what that is supposed to be measuring really anyway, is it derived from anywhere?
The rest of it makes a little more sense. Basically, CU for this period hasn't strayed more than about 8 points from your estimated mean based on LS, and unemployment hasn't gone more than 8 points over it's low. And they will obviously tend to move in opposite directions. So there will be a pretty strong negative correlation there, and I guess you can add the terms together for a sort of metaphor for total unused capacity.
Looking at the actual relationship between CU and unemployment, if I run a regression on (1-CU) as a predictor for the unemployment rate (zero intercept), I get a coefficient of .32 with a 95% confidence range of .31 to .33. And actual unemployment is rarely more than 2 points away from predicted (only happened in 2000-2001). So if we theorize a lower limit of around .32(1-CU)-.02, and then instead of actual CU use estimated CU according to the labor share, we could have a rough estimate of the lower limit on unemployment of .32(1-(.78*LS))-.02.
That would put the current limit at around 6.5%, and it seems to give reasonable limits historically, as it was as low as 3% in 1969-1970, and still around 4% as recently as 2001 before the recent drop in labor share.
So if you were to say unemployment in the current cycle won't drop below 6%, that's a horse I might bet on. Without incomes rising, the Fed will have to inflate asset bubbles again to break that limit, and I think recent experiences will cause those markets to get jittery if they rise too quickly this time.
Posted by: acerimusdux at January 20, 2013 09:57 AM
reply to acerimusdux...
This theory for UT would need professional math like yours in order to be published well.
I understand the concepts and only use a raw math to bring it out.
In graph #5 showing the natural rate of unemployment, the regressed polynomial line is pointing toward 6.7%, but as UT drops, that line should start to bend upwards. The y-intercept will increase in theory signaling an even higher natural rate of unemployment.
The difference between the two regressed lines is around 2.7 or 2.8%, which is larger than the difference in their y-intercepts.
But when I say natural rate of unemployment, I am not saying lower limit like you are calculating. The unemployment rate can be below or above the natural rate.
Would it then be bold yet better to say there is a natural rate of 7.1% with a somewhat hard lower limit of 6.4%?
I am using your version of cu(ls) = 0.78 * ls...
Posted by: Edward Lambert at January 20, 2013 01:12 PM
Not unreasonable if you think there is a strong labor share constraint. If you are looking at the mean of the trendline which we are fluctuating around, then my estimate of the current rate there would already be at 8.5% given the current labor share.
One problem with this estimate may be that the recent sharp drop in labor share is likely a temporary result of the early stages of recovery, and not the establishment of a new trend. There does seem to be a more gradual longterm downward trend, but I'm not convinced the game changed so much in 2009-2010. See for example:
"Our model indicates that the labor share is currently 1 to 1.5 percentage points below its long-run trend level. Part of the decline in the labor share in the past five years was temporary, and it will be reversed as the recovery continues."
Still, even if I plug .96 into my equation as the long run labor share trend, instead of the current .942, we would still hit my estimated limit there at around 6% unemployment. But while we only broke that limit once since 1967, that once was the most recent recovery in 2001. So possibly there is no hard constraint there either, and any relationship there is dynamic as well, and we could at least temporarily get to 5.5% or less.
Mainly, I think any actual relationship must be based on the importance of labor income in influencing demand. But there are so many things that can influence demand that I have trouble believing in a truly hard constraint. Between the possibilities of a new bull market in stocks, a strong housing recovery, stronger fiscal policy (unlikely) and aggressive monetary policy (maybe with the Fed currently on a pace for $1T a year in purchases), I can't say we'll see a hard limit.
I also think we could finally see wages begin to recover. For example, the Fed for a change could avoid choking off a recovery prematurely and lets wages actually rise, even as inflation creeps up over 4% temporarily. But perhaps that's wishful thinking.
Posted by: acerimusdux at January 22, 2013 05:07 PM
No, interest rate chanegs don't affect aggregate supply. They affect aggregate demand through chanegs in consumption (goes down because consumers save more), investment (goes down because the return on investment has relatively fallen compared to saving), and net exports (goes down because of capital inflows attracted by the higher rate of return on investments in Canada). In the long run, interest rate chanegs would affect AS, through the effect the change had on investment (lower interest rates > increased investment > increased capital stock > increased productive capacity > increased output) but that would take a long time.
Posted by: Jaroslav at February 18, 2013 01:20 PM
CBO models the output gap on the unemployment gap. It would be astounding if there WAS a "major outlier" in Chart 1. It would be astounding if Chart 2 did NOT show a tight fit.
Ball Leigh and Loungani consider their charts "evidence". Sadly, the charts are not evidence.
Posted by: The Arthurian at March 14, 2013 02:03 AM
I was wrong. In my previous comment I said "CBO models the output gap on the unemployment gap." In an exchange of emails, Laurence Ball agreed that "using CBO output gaps is circular" but pointed out that for the post he "calculate[d] potential output and output gaps separately from unemployment gaps and the natural rate, using the Hodrick-Prescott filter..."
Using Kurt Annen's H-P code to determine trends, I was able to duplicate quite satisfactorily the first two charts from the post: the linear relation between gaps, and the similarity of estimated to actual unemployment. The latter is quite amazing. Take the difference between real GDP and its HP trend, halve that difference, and add that to the HP trend of unemployment. The result is an extremely close match to FRED's UNRATE. Crude as the calc is.
In my version of their work, my "output gap" is only the difference between RGDP and its HP trend. To me a trend of realized RGDP is substantially different from the path of Potential RGDP. But this is a separate matter and does not detract from my conclusion:
Ball Leigh and Loungani's charts ARE evidence, and good evidence at that.
Posted by: The Arthurian at April 2, 2013 12:13 AM