Sentences to ponder

This first figure shows that aggregate demand growth has not been affected by a tightening of fiscal policy since 2010.  Specifically, it shows that nominal GDP (NGDP) growth has been remarkably stable since about mid-2010 despite a contraction in federal government expenditures.

That is from Macro and Other Market Musings, and there is another good picture at the link.  I understand full well that this is “unadjusted” and one may well argue “growth could have been stronger.”  I’ll simply note that I’d like to see discussions of fiscal policy accompanied by this picture as a starting point for analysis.

I would not, by the way, endorse the author’s conclusion about crashing into the fiscal cliff;; for one thing uncertainty and sectoral effects would be significant and would interact negatively with AD effects, even under perfect Fed policy.

For the pointer I thank David Levey.


The so-called fiscal cliff is just the drunk waking up from his hangover and realizing (dimly) that he shouldn't crack open another beer. Thanks to Reason magazine for the metaphor (

Many drunks are want to crack that beer right open.

Like our Drunk-in-Chief, e.g.

I feel like that graph would be more insightful if it included 2008 data. One really, really may well argue that growth could have been stronger, and including the 2008 data helps to put that view in perspective as well.

Well that's what the first poster said, but his comment disappeared. We need a better board.

It was my comment that Tyler removed, making the timing point (should start the chart at 2001 rather than in the middle of the recession); that it doesn't measure against GDP potential; nor does it fit with the trope that the stimulus didn't help, since expenditures would have been even lower had there not been the stimulus.

I may have forgotten other points, so I ask Tyler to post my comment he decided to remove unless he wants to be considered a censor.

Tyler, please post my original comment below, along with an explanation of why it was removed unless you want to be considered a censor.

I don't think TC is a censor, I think this software is the problem, speaking as a programmer. This is one of the few if not only places where you can rant and rave pretty much to your heart's content on economics. And TC even will answer your email on occasion. Imagine Mankiw or Krugman or DeLong doing that. Let's face it: economics is fashion and we should be happy that TC puts up with the penny ante crowd here, as it does dilute his trademark somewhat to have any fool post stuff on his blog, pretty much unmoderated.

Click on the picture. You may have posted in the wrong place. There are comments that seem to be associated with the picture instead of the blog post.

You are correct. Thank you very much for pointing this out and restoring my faith in humanity.

Bill complain about a graph? That's unpossible!

If you follow his source, you can construct your own graph based on Fed data.

The intersection of the two graph-series in mid-2010 is merely a plotting artifact right?

Good catch. Not only are the series already expressed as relative changes instead of absolute quantities, they are plotted on different axes. The dotted line that highlights the intercept is pure visual deception. That point has no significance whatsoever.

The chart should probably also include state and local government spending as well.

Good point, JFA. Of course the main declining portion of the US economy over the last two years of so has been the state and local sector, with employment rising in others. This has been where the austerity in fed spending has hit with the removal of subsidies for those levels of government, although they were facing declines from their declining revenues as well, with the property tax assessment downgrades lagging the declines in the real estate markets.

As it is, another argument for "could have been worse" (or better) is the comparison with Europe, where we have seen much more severe austerity and a return to recession.

I'm confused. Why wouldn't Krugman like this figure? Before mid 2010 federal expenditures as a percentage of NGDP are increasing or constant and the rate of NGDP growth is accelerating. After mid-2010 nominal federal spending is held, more or less, constant (calculate for yourself what the fed expenditures/NGDP ratio would be with 4% NGDP growth starting in 2010 and constant expenditures) and NGDP growth stagnates. The chart seems to show that (under the prevailing economic conditions of the last few years) having nominal federal expenditures grow at the same rate as NGDP is associated with increasing rates of NGDP growth. NGDP growth slows when federal spending doesn't keep pace.

But NGDP growth did not slow, it continued at 4%. If you believe what you are saying, then all we have to do is have spending shoot up one year, NGDP growth will go up to like 10%/yr, then lower spending to far below what is was to being with and NGDP growth will continue at 10% indefinitely. See what I'm saying?

Thanks Cliff!
I must not have been as clear as I had hoped. During mid-2009 to mid-20010, the fed expenditures/NGDP ratio held fairly constant at about 27% while NGDP growth rates went from -3% to +4%. From mid-2010 to mid-2012, fed expenditures/NGDP dropped while NGDP growth rates held constant at +4%.

The annual change in NGDP growth went from +7% when fed spending grew at the same rate as NGDP to +0% when fed spending started to grow more slowly than total NGDP, or perhaps I am misinterpreting the figure. Also, I did not mean to imply that any particular rate of federal spending would produce spectacular growth forever. Economic conditions change, and optimal policy probably too. I was only trying to interpret the historical relationships described by the figure.

The claim "demand growth has not been affected by a tightening of fiscal policy" would be stronger if the indicator of "fiscal policy" included revenue and transfers.

"This first figure shows that aggregate demand growth has not been affected by a tightening of fiscal policy since 2010."

The slope of the blue line changes, and then the slope of the red line changes. In 2010.

It will take more than this figure to show the things the author claims that this figure shows.

Precisely! This should be the headline of the post, not a burried comment.

And with tight money at the same time.

Krugman and Sumner just fell past my window with linked arms.

Actually, I think NGDP is about to stumble. I don't expect another recession, but most of the indicators are showing a slowdown in 2H12.

I think the graph is confusing because it plots a rate of change in a variable versus a variable. A flat red line represents a steady growth rate. The descending blue line shows Fed government expenditures declining as a share of nominal GDP.


Good point, but i checked it out and the results don't change using same units of measurement.

My response to Beckworth:

Shouldn't you be looking a net federal debt or balance rather than just expenditures to see the full position of fiscal policy. Just expenditures does not measure total fiscal policy.

Or we could be in a position like the Japanese were fiscal stimulus appears to keep the economy from contracting but it can not generate growth.

"This first figure shows that aggregate demand growth has not been affected by a tightening of fiscal policy since 2010."

As Mr. Cowen mentions, everyone in the room knows better than to make causal claims based on the short-term trend lines in these graphs.
But if one were to do so, shouldn't it come up that the slopes of these two lines are negatively correlated?

Eli, If you are accelerating into a recession, and you increased spending to arrest it, and didn't correct for lags in the chart, of course it would be negatively correlated If you want to look for an unlagged correlation, ask Hoover over a three year time period, and you'll find a positive correlation. You're argument is that the negative correlation--federal spending increased--caused a decline in GDP in 2009 is, well, interesting.

I also LOVE charts, like this one, where the items identified on the left axis is different than the items on the right axis and the percentages on each axis have different spacing for percentage changes.

My mistake; I meant *Possitively* correlated!
My point was that growth rates seem to go up shortly after federal spending goes up, and then levels off shortly after federal spending goes down. I don't mean to make a causal connection of any kind, but pretty certainly it isn't sufficient to show Mr. Beckworth's claim that austerity doesn't affect growth rates.

What I get out of this is Obama was right "the economy is doing fine" and all the conservatives were wrong when they ridiculed Obama for being out of touch because the economy is in steep decline.

I guess that's why Boehner proposed zero changes on the fiscal front for years: the economy is doing fine and the current debt and deficit are not a problem because the Fed can just continue to print money.

From Beckworth's pos:
"P.S. Though the Fed has been doing a remarkable job keeping NGDP growth stable since mid-2010, it has yet to allow a period of catch-up nominal spending growth that would return NGDP to its pre-crisis trend. So the Fed's work is not complete. "

This chart pretty clearly shows the Sumner Critique: The Fed moves last.

What is knowledge worth? An austerity slope might be worth it, at least for a time.

All this graph provides is another example of what can be done with carefully selective scaling and framing. How about some econometrics?

I know that true Austrians eschew econometrics, but graphs are no more than weak, intuitive econometrics. If you can't trust statistics, why would you place greater faith in poorer statistical methods?



We are all familiar with tricks played with graphs, and you don't lose sight of the underlying point if you spent more than five seconds looking at it.

Your better critique is the rudimentary association between only two variables and no consideration of lags.

Alas, in 2D space, charts are a bit constrained as didactic tools no matter how good your model.

Good point about Austrians.

"Specifically, it shows that nominal GDP (NGDP) growth has been remarkably stable since about mid-2010 INSPITE OF a contraction in federal government expenditures."

fixed that for you.

x firms made all the increase in growth (they were in NY City) and everyone else went to hell!
The graph and the info tells us nothing about how we the whole of the humans in the Country did.
What mis-direction.

Unadjusted? But what would the null hypothesis be? Are you saying that if billions of fiscal austerity had not been implemented, then at exactly that point monetary policy (the only other thing that really affects AD/NGDP, barring exogenous financial panics) would have at that very moment deviated from its steady trend and boosted NGDP by that many billions of dollars? Fiscal contraction was offsetting monetary expansion, instead of the converse? Seriously.

Whoops, sorry, NGDP growth had an upward trend, didn't it? Which may have continued. So that's not a very good argument then, I guess.

Government expenditure as a % of GDP is in essence a 'real' variable, as it's deflated by the implicit price level of NGDP. NGDP is of course a nominal variable. If prices in a sector of the economy are booming after 2009 (say: oil) this causes a rise in nominal GDP which will also show up as a decline of the % of nominal government expenditure (or any other non-oil sector) in total nominal production. In other words: the whole thing can be influenced by terms of trade effects. The right thing to do is to compare government spending with the other kinds of spending, as Brad deLong does on his blog . Which shows that exports and business equipment expenditure saved part of the day (consumer spending is not shown by Brad). These shares of course always add to 100% and you still will have to analyse the level of the aggregate - but that's exactly the problem with the graph above.

GDP before 2009:
Government Expenditures:[1][id]=GCEC96


Thanks for posting the link. My first instinct after reading Beckworth's post was to create similar graphs for that time period displaying monetary policy. As expected, there is very little correlation between the MB or M2 and NGDP over the same time period. Even if the initial logic were correct, it seems the same logic would suggest monetary policy is ineffective. We must be careful of what we infer from these types of graphs.

But Sumner would say that the MB and M2 are not good indicators of monetary policy. The right measure is expected NGDP growth.

Comments for this post are closed