Today, we are pleased to present a guest contribution written by Lydia Cox (Harvard University) and Kadee Russ (University of California, Davis), both formerly on the staff of the Council of Economic Advisers.
A study released in December by Aaron Flaaen and Justin Pierce, two highly respected experts in trade and labor markets, was widely covered in the press. “Trump Tariffs led to job losses, higher prices for businesses,” was the most succinct and comprehensive among the article titles.
Some of the coverage noted the statistic in Footnote 10 on page 19 of the Flaaen and Pierce study: By mid-2019, manufacturing employment ended up 1.4 percent lower than would have been the case without the tariffs levied in 2018-2019, likely due to a combination of increased costs of production and retaliatory tariffs. Yet it was hard to find in the articles an articulation of how many jobs this represents in level terms.
Exactly how many jobs is 1.4 percent of pre-trade-war manufacturing employment? Take 1.4 percent of 12.5 million, the number of manufacturing jobs at the end of January 2017, the month before the trade war began. The answer is 175,000 manufacturing jobs missing by mid-2019. Flaaen and Pierce’s study suggests that increased costs for imported inputs account for about two thirds of the total reduction in manufacturing employment. Retaliatory tariffs account for the remaining third.
In fact, more than 175,000 jobs disappeared: this figure nets out the roughly 40,000 jobs that may have been added or protected in industries benefitting from tariff protection. Furthermore, the estimate captures the decline in manufacturing employment only through mid-2019, but the tariffs have remained in place for a year since that time, likely leading to additional losses.
Digging a little deeper, we see evidence in Figure B5a that within this number, about 75,000 of these missing manufacturing jobs (about 0.6 percent of manufacturing employment, once we weight by the average cost share of steel) are associated with the Section 232 tariffs on steel and aluminum that went into effect in March 2018. A potential positive impact on employment from import protection is much smaller and not significantly different from zero. The steel and aluminum tariffs plausibly may have led to an increase of 1,000 jobs in these industries and kept a few thousand more from disappearing. See our Econofact memo for more on this.
Flaaen and Pierce’s study design filters out effects of macroeconomic conditions like variation in the value of the dollar and changes in economic growth overseas. By doing so it can plausibly be seen as a conservative estimate if the escalating tariff war led to the dollar strengthening against some currencies or dampened global growth, as downward revisions in growth forecasts by the IMF and OECD have suggested. Moreover, the estimate of 175,000 missing manufacturing jobs does not include adverse effects on employment that may have occurred due to the way that trade policy uncertainty dampened investment and industrial production during that period across many countries, according to macroeconomists Dario Caldara, Matteo Iacoviello, Patrick Molligo, Andrea Prestipino, and Andrea Raffo.
In addition to job losses, tariffs create a burden on households in the form of higher prices on goods and the inconvenience of having to substitute away from goods targeted by tariffs. Estimates assess the costs of the trade war from January 2018 to June 2019 at about $800 per household. Considering the macroeconomic effects of the trade policy uncertainty more than doubles this figure. The study by Caldara, Iacoviello, Molligo, Prestipino, and Raffo captures the overall impact of the trade war during that period, including trade policy uncertainty and attendant effects on national output through its adverse impact on investment and industrial production. They estimate that the trade war caused the U.S. economy be 1 percent smaller in 2020 than it would have been without the tariffs, equating to an average cost of $1700 per household. See this Econofact memo for a comparison and discussion of overall costs of the trade war, or more on the macroeconomic impacts of unilateralism here.
This post written by Lydia Cox and Kadee Russ.
Interesting that the increased costs of inputs into American goods actually hurts more than the retaliatory tariffs. I’m assuming something similar to this number has been reported before, but to be honest, I never remember having read it myself, and I think that ratio which differentiates the costs of tariffs inherently interesting. And I appreciate the authors Russ and Cox satiating this question. The Flaaen and Pierce paper itself and the “mainstream” writings of the two generous authors above (although the PBS article reads much more complex relative to most mainstream writings, “complex” in a good way) seem to be interested in separating the “core” costs of these bad MAGA decisions on tariffs from the “extenuating” costs of the tariffs (if that’s not too sloppy a joe six-pack paraphrasing there).
I also look forward to the future paper Russ is doing with Obstfeld (if I can manage to cognitively grasp it) and the future works of her cohort.
Trump promised the working class a lot of things in 2016. His policies have been a disaster for workers. Biden recently has started talking about economic issues noting how his Administration would be different.
It’s obviously not a “cure” or a vaccine, or a “guarantee” of any kind. But you take what you can get when it can make a difference:
Mix it with washing hands, social distance, mask, and maybe some 60%+ Ethyl alcohol based hand gel in the car when you’re driving around in public~~~you can adjust your probabilities.
I’m middle aged, I take 400-800 IUs of vitamin D everyday. One 400 IU pop in the morning and one 400 IU pop in the evening. Really I just randomly take them at least 6 hours apart. I could probably take more, but it’s all I want just to adjust my risk factors and not worry about calcium build up in my arteries.
I believe our good man Professor Chinn has a moderate and healthy interest in probabilities and odds~~~
The Federal Reserve data would seem to indicate that the 1.4% estimate was probably within the margin of error for their analysis. It wasn’t until COVID-19 hit and states shut everything down that manufacturing jobs plummeted. https://fred.stlouisfed.org/series/MANEMP
You might note a similar plateau in 2015-16. That was followed by manufacturing employment growth in 2018 with the tax relief taking effect.
Where you reading some other paper rather than their Federal Reserve study? For example, they do not mention Douglas Holtz-Eakin after that caveat they wrote which you took totally out of context. So don’t lie to us to suggest they did.
As far as your last line:
“That was followed by manufacturing employment growth in 2018 with the tax relief taking effect.”
The authors do not credit the tax cut at the very end of 2017 with growth in manufacturing at all. In fact, see their Figure 1: Measures of Manufacturing Activity: 2017m1 to 2019m8. Employment was rising BEFORE that tax cut.
Is Bruce Hall lying to us again? Or is he just over his head as usual?
Thanks to Google for helping me find the Fox Business discussion that little liar Bruce Hall failed to provide a link to. This is where he drew his weird quote:
In case people thought Douglas Holtz-Eakin was referring to our current weak economy, notice the date of this Fox Business cheerleading was December 29, 2019.
No wonder dishonest Bruce did not provide a link to this Fox Business cheerleading.
pgl: Oh. Now I see what Doug Holtz-Eakin was getting at. Well, let’s just say, those remarks have been rendered “inoperative” by current events.
Bruce Hall: A simple plea, to advance discussions in an intelligent manner. Please provide links (i.e., URL addresses) if you are quoting a person. I know Doug Holtz-Eakin; I worked for Doug Holtz-Eakin (twice!). I suspect there is some context missing in your quote.
BTW – take a look at Bruce’s favorite measure of manufacturing employment since the beginning of 2010:
We saw a rather strong growth in manufacturing employment under Obama-Biden. Why didn’t Bruce note this? Oh wait – he is Kelly Anne Conway’s little puppet so an honest discussion is not allowed.
Of course Bruce wants to believe that tax cuts led to this employment growth. Ah yes the Obama 2010 tax cut. I remember it well. Snicker.
Of course this growth in manufacturing employment came to an end just as Trump’s stupid trade war began, which is the point of this paper. But of course Bruce’s job is deny, deny, deny.
Moses does this better I do but any mention of deny, deny, deny deserves this classic from A Guide for the Married Woman:
Anybody can attribute effects to causes. Your attributions are, to be generous, let us say tendentious. Two good economist, using simple mathematical tools, put the job loss from Trump’s trade policies at 175,000. You throw dust in the air and try to put a positive spin on Trump’s policies. Rational people will give credence to Cox and Russ and write you off as a political hack.
Even before RBG is buried, Trump is to hot to trot to fill her position with some right wing loony tune woman with the speculation being he will nominate Amy Barrett. Part of her record as an Appeals Court justice is to find some way to abuse Title IX and the 14th Amendment to protect the rights of a rapist:
Robert Scott of EPI recently wrote this on US manufacturing employment:
“We can reshore manufacturing jobs, but Trump hasn’t done it – Trade rebalancing, infrastructure, and climate investments could create 17 million good jobs and rebuild the American economy”
Wow his title sounds like he is working with Biden
“While the Trump administration has claimed that the era of U.S. offshoring is “over,” the reality is that the United States has not begun to address the root causes of America’s growing trade deficits and the decline of American manufacturing. Decades of trade, currency, and tax policies that incentivized offshoring, combined with an utter failure to invest adequately in infrastructure and good jobs at home, have contributed to growing inequality and an eroding middle class. President Trump’s erratic, ego-driven, and inconsistent trade policies have not achieved any measurable progress, despite the newly combative rhetoric. On top of that, COVID-19—and the administration’s mismanagement of the crisis—has wiped out much of the last decade’s job gains in U.S. manufacturing.”
This is followed by a lot interesting data and analysis. On trade policies:
The USMCA—which was touted as a replacement for NAFTA—is unlikely to resolve longstanding U.S.–Mexico trade issues. America’s trade deficit with Mexico increased by more than 29% in 2019 alone (U.S. Census Bureau 2020c). And when it comes to important sectors like autos and auto parts, General Motors has been closing assembly plants in Ohio, Michigan, and Maryland while increasing its reliance on imports from Mexico (AP 2019; Samilton 2019; Mirabella 2019). In fact, GM has been ceding market share to foreign producers for decades, and has grown increasingly reliant on imports from Mexico and other countries. Meanwhile, market share has been captured by foreign producers. Recently, BMW, Mercedes/Infiniti, and Kia opened plants in Mexico—a missed opportunity to reshore production to the United States (Szczesny 2019; Mexico Now 2018a, 2018b). And the supplier networks for these plants will be built in Mexico, not the U.S.—further eroding America’s auto industry. Offshoring to Mexico is also taking place in aerospace and other sectors, with aerospace exports from Mexico increasing 10% in 2019 (Krause 2020). While the USMCA significantly improves domestic labor protections in Mexico compared with the earlier version of NAFTA, its overall provisions are inadequate to stem these offshoring trends. The Phase One China trade deal is a bust, too. China promised to increase purchases of U.S. goods and services by $200 billion over 2017 imports. But Beijing is unlikely to meet these targets (Craymer and DeBarros 2020). And the deal doesn’t even address China’s egregious, systematic labor rights violations.Beijing has also strategically adjusted to the Trump tariffs. China is simply exporting more goods elsewhere, and the U.S. trade deficit with China’s trading partners rose rapidly in 2019. In fact, China’s overall trade surplus with the world climbed significantly in 2019 (Setser 2020a). China also reduced the value of its currency by 10.0% against the U.S. dollar since March 2018, helping to offset the tariffs (Fed 2020a). The tariffs remain a “signature” element of the Trump trade agenda. And they’ve helped sectors like steel and aluminum (Scott 2018a, 2018b). But the president misses a key point: If you increase tariffs without taking steps to prevent the dollar’s appreciation, the overall benefits can be simply neutralized.
This is followed with a critique of Trump’s tax policies and several proposals that sound a lot like he is indeed working with Biden.
China reducing its currency was why parts of China’s system told Trump to go in 1918.
It is interesting that such a substantial portion of the job losses due to the Trump’s trade war have been due to the higher cost of inputs. This was worried about at the time some of the tariffs adopted, such as those on aluminum and steel, precisely for these reasons, with several companies in industrries, such as autos and appliances, warning of this problem. However, those job losses tend to be nearly invisible, and thus may not have much political heft. The new jobs for 40,000 in protected sectors are highly visible and get a lot of attention.
Thus the Sept. 12 Economist has an article in which a reporter went to Youngstown, Ohio to talk to people working in the construction industry who used to support Dems but switched to Trump in 2016. Most of them are still for Trump, if anything more strongly so. But they astoundingly cite one falsehood after another: “He fixed health care.” “He built the wall,” and, of course, “He has brought back manufacturing jobs by protecting us from China,” etc. Depressing.
It still looks like he will win Ohio, although he does not deserve to, especially with the Lordstown GM plant closing there.
Uh, it doesn’t look like anything in ohio. It’s been a light red state since 1990. Presidential votes come down to the large independent vote in ohio.
Sorry, Rage. You are wrong about Ohio. The last time it backed a losing candidate was 1960 when it went for Nixon over Kennedy, a very close race. It has been a super swing state, the one that probably put GW Bush over in 2004. The GOP tilt is recent, with it going for Trump in 2016, even though he lost the national popular vote by a substantial margin. This year it seems to be even further into the GOP camp, with many Dems not even wanting to contest it, although it does seem o ey have a chance. But it is highly likely this will be the first time since 1960 it does not go with the ultimate winner.
You do not know what you are talking about, Rage, and hardly for the first time.
The funny thing is that Trump was not particularly great for Ohio manufacturing:
Ohio saw an ever-so-slight uptick in manufacturing employment in 2017 followed by mostly flat employment in 2018 and then an ever-so-slight decline in 2019. Overall Ohio saw much larger gains in manufacturing employment during the Obama years, most likely due to the economy recovering from the Great Recession.
The attitude of those Youngstown, OH voters represents a larger problem across the electorate. Economics, particularly trade economics, requires a lot of abstract thinking and an ability to think in terms of counterfactuals. Those are not skills that one learns with just a high school education. Precious few develop those skills even with a 4-year degree. Economists need to find a way to explain non-intuitive economics in a concrete and intuitive way. That’s probably a more urgent task than economists writing more 30 page papers on a narrow topic filled with differential equations that only six people will read.
“The Trade War Has Cost 175,000 Manufacturing Jobs and Counting”
Total Employment in 2019 was roughly 150 million. So 175,000 is 0.00116666667
sammy: But from Jan 2018 to June 2019 – the trade war – manufacturing employment rose 280K. 175K is 62.5% of that amount. So I’d say that was pretty horrible.
I must say, you are the most dedicated apologist for bad policies I have have ever interacted with.
There is also the basic bottom line matter that Trump is running around the nation right now loudly proclaiming that his trade war has increased employmentr. It is one of the main things he argues regarding his economic policy and promises more of it. His crowds wildly cheer these claims. But they are false. That is the bottom line: Trump openly and repeatedly lying while you justify it on the grounds that maybe the lie is not all that large.
Sammy is worse than Lawrence Kudlow? Make him the head of the CEA!
Menzie, representing the typical Dem position says: “…you are the most dedicated apologist for bad policies I have have ever interacted with.” Meanwhile, they conveniently ignore the policy results comparison of this and the past administration. Conveniently, they want us to believe the trade war only began when we fought back.
PGL, how those ole soybean prices going?
CoRev: NFP at this corresponding time in Obama administration, 0.2% higher than at start; Trump, -3.2% lower.
Says the economic genius who twice voted for Nixon and took part in those tri-cornered hat rallies ten years ago.
Soybean prices? Seriously – you forgot how to check this out:
Still well below CoRev’s forecast.
“they conveniently ignore the policy results comparison of this and the past administration.”
You must be dumber than I gave you credit for. I have already noted manufacturing employment grew a lot from 2010 to 2017. I also provided something very fair and balanced from EPI. I guess it went over your head. But do try to read it again.
“PGL, how those ole soybean prices going?”
It is odd that CoRev is too much of a coward to ask this question of Menzie. I have never pretended to be THE expert on soybean markets even though I know enough to realize that CoRev never has had a clue.
Yes – spot prices very recently rose above $10 a barrel a full 15 months AFTER CoRev’s bold forecast that did not materialize. But as I have tried to tell the village idiot CoRev many times, spot prices are not the same thing as future prices.
Huh – maybe our host can provide a discussion of soybean futures. I’m willing to bet that future prices are less than current spot prices. Even if CoRev still to this day does not know the difference.
CoRev Conveniently, they want us to believe the trade war only began when we fought back.
Fought back? Can you explain why fighting back (i.e., escalating a trade war) makes any economic sense at all? Again, did you learn your economics on the playground? It sounds to me that you’re arguing by some facile power politics metaphor instead of actual economics. In any event, Trump gave away our most effective “weapon” when he killed TPP. Of course, you wouldn’t know that because you get your economics from future Nobel laureates Sean Hannity, Laura Ingraham, Mark Levin and Tucker Carlson.
Soybean prices: a three year “blip”?
You may get better results by removing outliers from your sample. Giving weight to a sample point which is obviously in error may reduce the validity of your work. sammy is bad data.
Apparently, the word “manufacturing” had too many syllables for Sammy to handle. “Total” was more in his wheelhouse.
You think Sammy used to play Porky Pig? He had trouble with words with more than one syllable too!
Trump told us he would increase manufacturing employment by a lot. Even a modest decrease is far from what we were promised. But do make excuses for sheer incompetence. It is what you do.
Is there where Bruce Hall got this nonsense?
“The Federal Reserve data would seem to indicate that the 1.4% estimate was probably within the margin of error for their analysis.”
I doubt Brucie boy does not even know what “margin of error” means in context of this paper. I’m sure you don’t. But here is a homework assignment for the two of you. Actually READ the paper and mansplain to us where the authors talked about “the margin of error” for their analysis.
We’ll wait as I’m sure this will be fun!
“Total Employment in 2019 was roughly 150 million.”
I guess you do not realize that a lot of these 150 million were not jobs in the manufacturing sector. Government employees, service sector, etc. Why not take 175 thousand relative to the world’s population?
“manufacturing employment rose 280K.”
So, overall, his policies must have been good, right? Especially compared to the Obama years (your policies?).
Manufacturers Added 6 Times More Jobs Under Trump Than Under Obama’s Last 2 Years. https://www.forbes.com/sites/chuckdevore/2019/02/01/manufacturers-added-6-times-more-jobs-under-trump-than-under-obamas-last-2-years/#ebdf7785635a
So a paper saying Trump theoretically cost 175,000 jobs, or 0.00116666667, is pretty weak proof of bad policy, more akin to a rounding error.
That’s like criticizing a basketball team that won 250-0 but should have won by 251-0, if they just had hit that free throw at the end of the game.
“So, overall, his policies must have been good, right? Especially compared to the Obama years (your policies?).”
Seriously Sammy – your lies are so apparent that they are beyond stupid. Earlier I took Bruce Hall’s favorite measure back to January 2010 and note from then to January 2017, manufacturing employment rose by 909 thousand. From January 2017 to January 20202, it rose by 475 thousand before it just fell off the cliff.
Yes the liars at Forbes can take the slowest two years of the Obama period – which even a lying piece of garbage like you knows is dishonest. In a few months, we can look at Trump’s last two years. How is that going to look?!
“That’s like criticizing a basketball team that won 250-0 but should have won by 251-0” I bet you were the center for the losing team here. Yes – your defensive skills are nonexistent and your offense is worse.
Chuck DeVore writing his usual intellectual garbage for Forbes way back in Feb. 1, 2019? This is your idea of reliable updated information. BTW – Chuckie boy confuses nominal wages increases with real wage increases.
Congrats Sammy – you have found someone dumber than even you!
sammy Look VERY closely at the graph in your Forbes link. Aren’t you just a wee bit concerned with the way Honest Chuck displayed the BLS data and the way he calculated job growth? Think it over. Hint: suppose there are 100 manufacturing jobs and next month one job is added. Then the next month two jobs are added. Using Honest Chuck’s math job growth in the second month is 100% greater than in the first month. Are you comfortable with that? Is that how you calculate growth in your 401k? Here’s a pro tip. Anything from Forbes is not worth lining a bird cage. It’s target audience is people who don’t actually know a lot about economics but like to think they do, not unlike watching CNBC. And the real clincher that he’s playing you for a sucker is the insinuation that growth in government jobs adds less to GDP than growth in manufacturing. It’s an appeal to your unconscious belief in the old Marxist labor theory of value.
Chuckie Cheese wrote ‘The sluggish growth in manufacturing in the latter half of the Obama years led to President Obama remarking in June 2016 that manufacturing jobs “are just not going to come back.”’
Chuckie provided a youtube link but did he listen to Obama’s full response. Sammy didn’t because had he done so he would have heard Obama talking about the increases in manufacturing jobs under Obama’s Administration. And of course had Sammy checked that FRED link, he would have seen that Obama was telling the truth.
It seems Chuckie pulled a quote out of context and created a misleading graph and Sammy the Sucker fell for it.
Chuckie Cheese wants us to think Obama was massively increasing government employment. Socialism! OK, government employment rose from the end of 2014 to the end of 2016 but Chuckie did not tell us about the decline in government employment from the end of 2008 to the end of 2014:
Let’s be clear. Chuckie Cheese is a liar. And Sammy is so stupid he falls for these lies.
I bet Sammy even knows who Chuck DeVore even is. He has been a right wing politician ever since he graduated from Claremont McKenna College. CMC has a decent undergraduate economic program with a definite conservative bent. But something tells me that some of DeVore’s writings would embarrass CMC. First of all his 2013 book claimed that if the rest of the nation followed the Texas model, we could see prosperity. Really? Everyone nation should drill for oil and grow cattle?
Then there is his latest:
“Chuck DeVore: Trump energy policy key to reviving economy — Biden plan would destroy jobs, raise energy costs”
Now that’s a title. The economy is nothing but the energy sector? DeVore really has been sipping a lot of Texas tea.
“One of President Trump’s most important accomplishments in office has been the energy revolution that has led to sharply increased oil and natural gas production and job creation in the U.S. But our days of energy insecurity and dependence on other nations in unstable areas of the world to meet our energy needs will return if former Vice President Joe Biden wins the November election.
The Trump administration’s deregulatory policies, combined with a rapidly innovating domestic energy industry, led to the U.S. becoming a net exporter of petroleum products in 2019 — including gasoline, diesel fuel and crude oil — for the first time since 1949.”
Excuse me but the shale oil revolution started well before Trump took office. And while we have a trade surplus oil, the rest of the economy had a deeper trade deficit. Yes DeVore’s brain is consumed with oil. This fool is critical of Biden for daring we develop other sources of energy. To Devore, it is oil and nothing else.
A curious aspect of this f9ossil fuels focus by Trump is that the increase in private capital formation during his first year in office, 2017, was overwhelmingly driven by a rise of investment in that sector, mostly oil and natural gas, something like half the increase, with a lot of that going to a few locations, especially North Dakota and Texas. But that has fallen off since with the decline in oil prices, which remain around $40 per barrel, not high enough to spur another round of that. And indeed, for all the hoopla about how the Trump tax law was going to spur private capital formation, including pulling corporate money from abroad, it really has not happened since 2017, with other parts of the economy responsible for most of the growth that happened up until the pandemic hit.
Thought the name sounded familiar. DeVore was a California assemblyman until he was “termed out” in 2010. He was so popular among California Republicans that he received 19% of the vote in the 2010 Republican primary, finishing far behind Carly Fiorina and Tom Campbell. His political career here? Over and done.
He then left for greener (or browner or slicker) pastures in Texas where, it seems, the state decided some time back that economic diversification was preferable to over dependence on oil. Alas, such diversification seems to have resulted in political and cultural diversification too.
Little did he know he arrived in time to witness the Californication (popular term in Idaho) of Texas.
From folks who might know about soybean markets
Soybean prices are temporarily up but these guys think this is about to change. Notice the comment section has Trumpian jerks screaming. Sammy
“”The trade subsidies represent one of the biggest ever farm sector bailouts not related to a natural disaster. For perspective, the trade aid that farmers have received is nearly triple what the Treasury department estimated it ultimately cost taxpayers to bail out the auto industry during the financial crisis of 2008.
“for many farmers, the aid represents both a lifeline and an awkward reality of government dependence.” Reuters Business News 3/11/20
“Government payments to farmers have surged to historic levels under President Donald Trump as the Agriculture Department floods the industry with cash to stem the financial losses from Trump’s tariff fights and the coronavirus pandemic.” Politico 7/16/20
“Soybean subsidies in the United States totaled 46.1 billion from 1995-2019” EWG Farm Subsidy Database
Anyone remember when Ronald Reagan insisted gubmint WAS the problem? . Bruce? CoRev? Sammy? Bueller?
5-star comment award from the panel of judges.
I have no idea why farmers support Trump. Not only have cash receipts from farm commodities fallen during the Trump years, but net farm income has come to depend on government payments (see BEA Table SAINC45). But even with those added government payments net farm income still fell. It averaged $91.3B over the Obama years and only $84.2B during the first three years under Trump (BEA Table 7.15). And those are nominal dollars not adjusted for inflation. What’s wrong with farmers? Are they old men with failing memories? Are they innumerate? Do they spend too much time watching Fox & Friends instead of working in the fields? MAGA hats fit too tight and blocking blood flow to the brain? What’s their deal?
Latest Iowa poll has the race 47/47. Trump still has a big lead among the Iowan men but the ladies strongly favor Biden. I guess the next Supreme Court nominee will decide the 19th Amendment is inoperative.
As a native Iowan, I can honestly tell you, watching Iowans vote for donald trump and Steve King is personally upsetting to me. Iowa has a better public education system than most states do—and Iowans should know better. The deep south has an embedded excuse—their public schools are predominantly worthless.