Stop predicting recessions!
Last year, lots of people said the Australian miracle would finally come to an end. I had commenters mocking me for arguing that Australia wasn’t a bubble. They insisted that the bubble was already bursting. (I wish you guys would come back to my comment section—I miss you.)
Now we are well into 2016, and the Australian economy is surging:
Australian business confidence jumped and an employment gauge in the survey surged to the highest in almost five years, signaling a healthy job market and reducing the likelihood of an interest-rate cut. The local currency gained.
The sentiment index doubled to six points last month, according to a National Australia Bank Ltd. survey of more than 400 firms conducted March 23-31. The business conditions gauge — a measure of hiring, sales and profits — climbed to 12, matching the highest level since before the 2008 global financial crisis. The employment index jumped four points to five, its best result since 2011.
“This is an especially good result in the context of a downbeat global economic outlook,” said Alan Oster, chief economist at NAB. “Low interest rates and a more competitive currency, even given recent strength, are expected to remain key drivers domestically. Consequently, our outlook for the economy remains unchanged — and with the non-mining recovery expected to progress further, monetary policy is likely to remain on hold for an extended period.”
Australia’s economy is proving resilient in the shadow of recent financial turbulence in China, negative interest rates in Japan and Europe and weaker commodity prices that have combined to increase global risk. The Reserve Bank of Australia cut rates to a record-low 2 percent in May last year in an easing cycle designed to cushion the economy from unwinding mining investment and encourage services industries to pick up the slack.
While gross domestic product grew a robust 3 percent last year and the unemployment rate has fallen to 5.8 percent, the Australian dollar has rebounded more than 10 percent since mid-January.
Last Australian recession—1991.
Funny how countries that maintain adequate long-term NGDP growth don’t seem to have problems with the zero bound. I wonder if the problems in Japan and the eurozone are self-inflicted? (NGDP growth has recently been weak in Australia, due to lower commodity prices, but the long-term expected trend is high enough to keep interest rates above zero. That trend rate remains well above European and Japanese levels.)
When the Chinese stock market crashed last summer, lots of commenters mocked my claim that China was doing fine, and insisted that it was entering a recession. This month the consensus forecast for 2016 GDP growth in China (private forecasters) rose from 6.4% to 6.5%. Even accounting for data problems, most experts think growth is at least 5%. In an economy with a flat labor force, 5% is not too bad. Retail sales growth is still running at double digits, and exports are picking up, although by less than this FT story suggests:
China reported stronger than expected trade data on Wednesday, the latest sign of a tentative revival in fortunes that paves the way for Friday’s release of first-quarter economic growth.
Exports surged 18.7 per cent in renminbi terms in March over the same month last year, after declines in both January and February. Imports also stabilised, dropping just 1.7 per cent compared with an 8 per cent fall in February.
In dollar terms, exports rose 11.5 per cent while imports fell 7.6 per cent for the period, reflecting the renminbi’s recent rise. The currency has gained 1.9 per cent against the dollar over the past two months.
China’s export sector has been buffeted by the slowdown in global trade, the dollar value of which has been shrinking since 2012 largely because of the slump in international commodities prices.
The International Monetary Fund this week warned that the world risked a “synchronised slowdown” but highlighted China as a rare bright spot among major economies. Chinese officials have been working to counter international investors’ increasingly negative outlook for the country’s economy.
Their cause has been boosted by a slew of better than expected data releases, including March inflation figures that showed producer price deflation had moderated.
This contributed to the IMF’s decision to revise upwards its forecast for Chinese economic growth this year, to 6.5 per cent from 6.3 per cent. At last month’s meeting of China’s parliament, Premier Li Keqiang projected economic growth of 6.5-7 per cent for 2016.
“China’s commodity imports should see further improvement soon,” said Zhou Hao at Commerzbank.
“Rare bright spot” doesn’t sound like a Chinese recession.
When I hear constant predictions that the Chinese bubble will burst any day now, I’m reminded of Redd Foxx. People need to stop predicting recessions, because recessions are unforecastable. The IMF has predicted zero of the past 220 negative growth periods between 1999 and 2014. This shows that the IMF is smart. They know that the most likely outcome is growth, and hence they predict growth. So do I! Recession predictions also have a corrosive effect on economics as a science. These predictions lead non-economists to believe that economists should predict recessions, and give undeserved reputational points to lucky permabears.
Peter Schiff said back in January that the US would have a big recession this year. If we don’t have a recession this year, people will forget Schiff’s false prediction, as well as false predictions of major crises in the US in 2015 and 2013, and recall his correct prediction of the 2008 recession. (Insert broken clock analogy here.) Whenever I hear that someone has accurately predicted a recession, my evaluation of that person declines. Economists should not be trying to predict recessions; the point is to prevent them. That’s why we need NGDPLT.
PS. If you have trouble with me praising the IMF being zero for 220, consider this analogy. If I’m standing next to a statistician at the roulette wheel, and he predicts the ball will not land on the green numbers (36 out of 38 odds) for each of 220 spins, then I will assume he’s a good statistician. If he occasionally predicts the ball landing on the green (2 out of 38 odds), I will think less of his statistical skills, even if it does land on the green.
HT: James Alexander
Tags:
13. April 2016 at 08:44
People need to stop giving Peter Schiff a platform, he deserves absolutely no credibility at this point – total fraud.
13. April 2016 at 08:52
But there estimates of turning points are statistically significant at the 95% level, therefore they merit pronouncement and publication
Haha, j/k, nice post.
13. April 2016 at 08:53
“Whenever I hear that someone has accurately predicted a recession, my evaluation of that person declines. Economists should not be trying to predict recessions; the point is to prevent them.”
-What if he tried and failed? And I don’t grade a person wrong for correct predictions. What if that person predicted the 1997 Asian crisis, down to the timing?
“When the Chinese stock market crashed last summer, lots of commenters mocked my claim that China was doing fine, and insisted that it was entering a recession.”
-I was not among them.
I don’t know what China’s actual growth rate is; nobody can trust Chinese statistics.
“If he occasionally predicts the ball landing on the green (2 out of 38 odds), I will think less of his statistical skills, even if it does land on the green.”
-This is nonsense. Recessions are not random. And what if predicting landing on green increases the statistician’s accuracy, Sumner?
I was predicting a recession beginning sometime between late 2016 and mid 2019 from 2014 on. I’m still reasonably confident it’ll happen (70%+), but I literally have no idea what the Fed is trying to do right now, so I’ve become less certain.
13. April 2016 at 09:03
Foxx looked very cool in his bow tie!
13. April 2016 at 09:03
E Harding,
Your less certain than 70% certainty that there will be a recession in a 3 year window? So that’s about punxsutawney phil type certainty, right?
13. April 2016 at 09:09
Harding, America has recessions about every 5 years, on average. We haven’t had one for a while. Predicting one in a 3 year window with 70% confidence is not exactly Nostradamus level forecasting.
13. April 2016 at 09:17
Brace yourselves, Austrins are coming.
13. April 2016 at 10:07
Isn’t it easy to predict recessions if you look for the right factors.
In the US – a highly diversified economy they only occur because of poor monetary policy. It seems to me we should never have another recession in the USA. But if one would occur it would occur because the FED ignored some kind of mini bubble and had tightening bias. 2008 seemed bad because we did have problems in housing/housing finance combined with a FED concerned with inflation with oil booming to $150.
Otherwise it would take a massive catastrophe to cause a supply side recession in the USA. 9/11 type stuff, but realistically even then monetary policy should be fine.
Other countries with more focused industry like Saudi Arabia or Brazil that are more exposed to a single factor like commodities are more difficult.
13. April 2016 at 10:19
The problem is as much the consumers of the recession predictions as the predictors themselves. Most people listening to the news aren’t engaged enough to interpret the implications of small changes in probabilities. So even if the permabears realize that their derived probability of recession is only slightly higher than the average, the only way to get heard and listened to above the din is to overemphasize their probability of a recession so it passes those heuristics barriers.
So what we actually need is a better education of probabilistic events so that wares of doomsday peddlers don’t look so attractive to their current consumers.
13. April 2016 at 11:14
Are recessions probabilistic events?
They seem 95% man-made thru bad monetary policy. With a 5% chance of a natural disaster large enough to cause a supply side problem.
13. April 2016 at 12:54
They seem 95% man-made thru bad monetary policy.
So, are the central bank governors chronically inept (and, if so, who are the non-inept who might replace them?) or is ‘bad monetary policy’ something one recognizes only in retrospect?
13. April 2016 at 13:18
Scott, do you mean “do not predict recessions [a certain time] in advance”, because surely the whole idea behind market monetarism is to utilise the information contained in market prices to determine how to change the money base. So if the Fed is not following MM and stock prices fall 30% and bond yields halve, then it might be reasonable for an economist to ‘predict a recession’, as you yourself have said would have been a reasonable view in mid-October 2008. Maybe you don’t think that would be a ‘prediction’ in any meaningful sense of the word?
13. April 2016 at 13:24
Keep in mind that Peter Schiff is a businessman. He’s selling fear to people. That being said, I don’t think that we should ignore these people. They do have interesting things to say from time to time. What people shouldn’t do is blindly follow their advice. You need to know when to ignore them and when to pay attention. I still remember going to a doom and gloom-type seminar back shortly after the recession ended and they were still selling fear to people. People who listened to their advice would have missed out on some of the best years for equity markets.
As for economists predicting recessions, consider a Markov-Switching model with two regimes. One has growth in GDP and the other has recession. This type of model can accurately capture some characteristics of recessions. It could tell you something like the probability that next month will be a recession given that you are likely not in a recession now. As growth slows down, you become less confident that you’re not in a recession and the probability that next month is a recession increases.
13. April 2016 at 14:32
‘While gross domestic product grew a robust 3 percent last year and the unemployment rate has fallen to 5.8 percent….’
No recession in 25 years but the unemployment rate is 5.8%!
High minimum wages don’t have disemployment effects. Kshama Sawant assured me.
13. April 2016 at 14:40
I should have pointed out that Australia’s $17/hr minimum wage (for adults) is about $13 in USD.
13. April 2016 at 15:05
Patrick, that’s not the worst of it. Minimum rates of pay in Australia are higher again on weekends and public holidays. Getting someone to wait tables on a regular public holiday will cost you $A50 per hour! That said, as elsewhere, many firms in hospitality operate on an informal basis paying less than those rates. But our zealous regulators are increasingly trying to catch and fine them.
Even so, I think the RBA has kept monetary policy too tight for the last 4-5 years. I see no reason why we can’t have sustainable 5% unemployment even without significant labour market reforms, provided fast food and convenience store businesses can manage to continue flouting the rules.
13. April 2016 at 15:09
Patrick: it used to be higher, during the Howard government, the relevant wage setting tribunal let it drift down in real terms.
13. April 2016 at 15:57
Scott,
I understand that you can give good theoretical and empirical arguments to justify your worldview, but I’m just noting for posterity that your post contains the following points:
A) “If we assume recessions are unforecastable, then the people forecasting recessions are making a mistake.”
B) “If Peter Schiff’s prediction comes true, he was lucky; he still doesn’t know what he’s doing. If it comes out wrong, it proves he doesn’t know what he’s doing. Let’s wait and see which way the data resolve this one.”
13. April 2016 at 15:58
Sean, Bad monetary regimes cause recessions, but it’s still hard to predict when they will occur.
Rajat, You said:
“So if the Fed is not following MM and stock prices fall 30% and bond yields halve, then it might be reasonable for an economist to ‘predict a recession’, as you yourself have said would have been a reasonable view in mid-October 2008. Maybe you don’t think that would be a ‘prediction’ in any meaningful sense of the word?”
Yes, it’s not much of a “prediction” when it occurs 10 months after the recession began. I did not predict a recession in December 2007, when it actually began.
Rajat, Isn’t the minimum wage lower for young people?
13. April 2016 at 16:07
http://www.paulnehlen.com/paul-ryan-betrayal/
🙂
Remember, Sumner wants Paul Ryan to be the Adlai Stevenson of the 21st century. And he had the gall to berate me about being concerned about immigrants being Berniebros!
13. April 2016 at 16:27
Scott, but it wasn’t much of a recession before October 2008, right? And I think you’ve said before something like that if October 2008 didn’t happen, the NBER may not have identified a recession beginning in 2007 at all. If the 1990-91 and 2001 smaller recessions are glossed-over by inclusion in the ‘Great Moderation’ period, as they often are, then don’t you think the act of calling the ‘important’ recessions before the NBER does may be of some value without doing too much reputational damage?
Yes, Australian minimum wages for people under the age of 21 are lower. Rates are about half the adult wage for 16 year olds and two-thirds the adult wage for 18 year olds. Here’s a slightly dated link: http://worksite.actu.org.au/youth-entry-level-wages/
13. April 2016 at 17:18
Bob, You said:
“If it comes out wrong, it proves he doesn’t know what he’s doing.”
Where did I say that?
You might wonder why I do post reminding people of false recession predictions. It’s a dirty job but someone’s gotta do it. Otherwise people will have the impression that some people are actually good at forecasting. Yes, these false predictions don’t prove anything, they merely put some realism into the picture. I often hear people talk about the brilliant predictions of Shiller, Roubini, Keynes, John Paulson, etc, without mentioning their false predictions. I like to debunk.
Harding, So Ryan supports free trade, which has been the GOP position in recent decades, and that’s suddenly a problem?
Rajat, You said:
“Scott, but it wasn’t much of a recession before October 2008, right? And I think you’ve said before something like that if October 2008 didn’t happen, the NBER may not have identified a recession beginning in 2007 at all.”
I’ve argued that September 2008 happened (Lehman) precisely because we were already in recession, and the recession got much worse after June 2008. It is true that I said the first 6 months were mild, and if we’d had growth in the second half of 2008 then no recession would have been recognized.
If you mean “is it possible to forecast that the recession is likely to get deeper”, then I’d say “yes.” But even that is harder than it might first appear.
13. April 2016 at 17:21
Rajat, Let me put it this way. If were possible to “rationally” forecast a recession at time=t, then the recession would probably have begun at time=t-1.
That’s because markets forecast things before individuals, and the market forecasting a recession also causes a recession to occur at roughly the time of the forecast. I’m not trying to be cute here, this is actually what I’ve observed all through my life.
13. April 2016 at 17:24
Great post Scot. You’ve been on a roll lately (especially with the recent post on “currency wars). I would really just highlight the stats 101 point that we shouldn’t be judging whether the sample mean is a good estimate of the population mean based on whether the next single number drawn from the population is the same as the sample mean. Arguably we should stop caring as much about the population mean as well.
13. April 2016 at 18:19
“So Ryan supports free trade, which has been the GOP position in recent decades, and that’s suddenly a problem?”
-Yup! Taking the GOP back to being the party of Lincoln and Coolidge. Didn’t you see that poll showing that Democrats are way more pro-trade-agreement than Republicans these days?
BTW, I don’t like this protectionist upstirring, either. Again, I’d rather have Ron Paul than Donald Trump. But I admit it would be so awesome if that smug-faced pseudo-Objectivist phony got ousted from Congress by primary challenge just like the last Republican Jew in the House of Representatives. Remember, it was David Brat who killed the path to citizenship.
13. April 2016 at 18:59
Thanks Scott, that makes sense to me. I think all I am saying is that people infer forecasts from market movements, but the market itself doesn’t announce a forecast in the way a human being does. So an economist ‘predicting’ a recession based on market movements is not really making that prediction as an economist, but rather than an informed commentator. Whereas many commentators in the media will either call a recession too far in advance or remain in denial well after the collapse in prices.
13. April 2016 at 22:00
Business cycles do end. Myself, I’m a lot more afraid of the next cycle’s end than this one. There’s a very real possibility that the next business cycle will end just as Social Security and Medicare go insolvent. It may not be likely per se (let us all hope the probability is below 50%), but it is a real risk.
Heck, in the worst case the U.S. could experience simultaneous ethnic/political and balance of payments crises. Or we could enact reforms and have no crisis at all; Congress isn’t completely stupid. At times, I really, really wish the next ten years would be over already.
13. April 2016 at 22:16
By the way Scott, what do you think of the argument that “free trade is bad if global demand is deficient” (secular stagnation thesis)? I came to the same conclusion years ago, but I don’t think it leads to protectionism. A deficiency in global demand is fundamentally a geopolitical problem, not an economic one. Economic policy is only effective in as much as it has geopolitical effects, and protectionism is hardly the only macroeconomic policy instrument that does that.
People always mocked Ben Bernanke for his “global savings glut” thesis. But I think it does make a lot of sense.
14. April 2016 at 03:03
What has Australia actually done to prevent recessions since 1991 (compared to other economies that have had recessions)?
14. April 2016 at 05:00
Federico, Thanks, and I agree.
Harding, You said:
“Didn’t you see that poll showing that Democrats are way more pro-trade-agreement than Republicans these days?”
And didn’t I tell you that those polls are meaningless?
You said:
“just like the last Republican Jew”
And you wonder why people don’t respect you.
Rajat, I once did a post entitled something like “It’s possible to prevent a recession after it has started”, which explains my views in more detail.
Joe, You asked:
“By the way Scott, what do you think of the argument that “free trade is bad if global demand is deficient””
I don’t agree.
BTW, when Bernanke said there was a global saving glut in the mid-2000s, he wasn’t saying it was a problem, AFAIK. He was simply describing why bond yields were 5%, and not higher.
Andy, The only thing I am aware of is that they’ve implicitly set their NGDP growth path higher than other developed countries, and thus avoided the zero bound. They did that with a slightly higher inflation target, and also a higher population growth rate (due to immigration policy).
14. April 2016 at 06:36
I’ll be honest Scott I was talking with someone about politics recently and I mentioned that Australia hasn’t had a recession in over 20 years and then later googled Australia recession and I can’t make heads or tails about that place. Is the economy there souring or soaring or treading water? I can’t tell. Could we get another post about it?
14. April 2016 at 06:42
The Australian situation is very encouraging, especially as an argument for NGDPT or NGDPLT. I bet BOJ is paying attention.
China’s deep trade recession seems to be stabilizing, and the reserves picture looks much better. Whether this is a respite before another crash, a return to growth, or something in between remains to be seen, but obviously huge institutional problems remain. Would really like to see those air pollution numbers start to drop in a serious way, and a lot more political liberalization, and RGDP reporting that makes a little more sense.
14. April 2016 at 07:29
But I admit it would be so awesome if that smug-faced pseudo-Objectivist phony got ousted from Congress by primary challenge just like the last Republican Jew in the House of Representatives.
Ryan is gentile; he likes Ayn Rand’s fiction but has eschewed high-test Objectivism; and he’s perfectly plain about what he stands for. (And what difference does it make that Eric Cantor’s Jewish? There’s no shortage of gentile Chamber-of-Commerce fellators in Congress, among them John McCain, Lindsey Graham, and the pretty little liar from New Hampshire).
14. April 2016 at 10:33
Eric Cantor was the last Republican Jew in the House of Representatives; i.e., no other Republican Jew has followed after him (so far) to the House. What’s disrespectable about that fact? The Israeli press covered it for like a week. Yes, Art, I am well aware that Ryan’s a Catholic. And I don’t think of McCain or Graham any better than of Cantor.
“And didn’t I tell you that those polls are meaningless?”
-What; you don’t think those polls predict voting habits?
14. April 2016 at 12:20
The biggest push back against Market Monetarism I get is some variation of ‘the central bank can’t control NGDP with any precision’. That’s sort of an empirical question, I mean maybe MV=PY, but NGDP in a modern economy is just inherently rather stochastic? It could be. But no grand theory of NGDP volatility is needed, I can just point to Australia. Why are they able to smooth out NGDP for decades? Probably because NGDP is actually fairly easy to smooth out, if that’s what you want to do.
Israel and Poland are good, Iceland is good, but Australia is easily the single best propaganda point for us. NGDP can be made predictable: if you will it, it is no dream!
15. April 2016 at 03:28
[…] Sumner warns that predicting the next recession is a mug’s game. The dismal record on forecasting certainly […]
15. April 2016 at 05:39
Harding, Not polls on views on trade policies. If it was Bush’s trade agreement, the views would reverse. If Obama says the sky is blue, then 34% of the GOP will say it’s green.
Justin, The point is to make NGDP less unstable, it will always have some volatility.
Everyone. Don’t put too much weight on Australia, or any other country or person (including me.) Any country (or person) can screw up at any time. Stick to the ideas, and the overall statistical evidence looking at many different cases.
If I become an MMTer tomorrow, just write me off as a mental case and start following other MMs, like Beckworth.
15. April 2016 at 21:31
[…] Sumner warns that predicting the next recession is a mug’s game. The dismal record on forecasting […]