* Production in the US motor vehicle sector has fallen by nearly 10% since the beginning of the quarter, reflecting the impact of supply chain disruptions in Japan following the natural disasters there. The setback in vehicle output is likely to shave approximately 4 points off the growth rate of industrial production in Q2. We also think it will take a bit more than 1/2 point off real GDP growth for the quarter, and are lowering our Q2 growth forecast to 3.0% (annualized).
* The drop in vehicle production probably explains some (though likely not all) of the weakness in manufacturing surveys released in recent weeks. It has pushed up jobless claims and could have a modest negative effect on May manufacturing payrolls. A more limited supply of vehicles in the summer months is likely to result in slightly higher vehicle prices and slightly lower sales than would otherwise be the case.
* Luckily, these effects should be short-lived. Production is likely to bottom out within the next few weeks and move higher over the summer. This will provide a small positive boost to growth in the second half of the year.
Several weeks ago we estimated the potential impact of the Japanese earthquake and tsunami on US economic activity. (See "The Effects of Japanese Supply Chain Disruptions on US GDP", US Daily, April 20.) Today's comment updates our estimates of the impact on growth and expands the scope to consider other possible effects on the US economy.
As expected, the hardest-hit sector of the economy appears to be motor vehicle production. Shortages in supply of key components—notably auto microcontrollers—have led to production shutdowns at US facilities, particularly those owned by Japanese manufacturers. The latest Federal Reserve data show that production of vehicles and parts fell 8.9% from March to April. Weekly data from other sources suggest car production will be roughly flat in May but truck production could be down perhaps another 2%. All in, the drop from the pre-earthquake level of US vehicle production in March to the trough level post-earthquake in May or possibly June should be around 10%.
This decline in production is likely to impact the US economy in several related ways:
1. Industrial production. The impact of the vehicle shutdown on overall industrial production depends on two key factors: the magnitude of the production cut, and the share of industry that is affected. Measured on a quarterly average basis, the drop in vehicle production will be less severe, perhaps around 7% (not annualized). (This is because the March level of production is somewhat above the Q1 average, so comparing to Q1 results in a smaller decline than comparing to March alone.) For those plants that have cut production, we expect most suppliers to follow suit; the alternative would be for them to continue producing and build inventories. This means that as much of 10% of industrial production could be affected (this is higher than the vehicle production share of GDP because suppliers of items like glass and steel may not be classified in the vehicle sector). A 7% drop in production across the vehicle supply chain therefore implies about a 0.7% drop in IP overall; converting this to a quarterly annualized basis would mean a 2.8 percentage point hit to the quarterly growth rate of IP in Q2. Industrial output has been growing at an average rate of approximately 5% over the past two years, so this would suggest IP growth of about 2% in the quarter. A range of potential impacts (using somewhat more precise figures) are given in the grid below, which repeats a table from our previous analysis; the likely impact should fall somewhere in the shaded range.
2. Manufacturing surveys. Lower vehicle output has likely contributed to recent weakness in manufacturing surveys. How much? If 10% of all industrial activity is in some way related to vehicles, and all of these firms go from expanding to contracting in one month—an extreme set of assumptions, to be sure—that would result in a 10-point decline in a survey like the ISM manufacturing index (where a response of decreasing activity is scored as a zero), and a 20-point decline in surveys such as those conducted by the New York and Philadelphia Federal Reserve Banks (where the number of responses indicating lower activity is subtracted from the number of responses indicating higher activity). As Japanese transplants make up more than one-third of US vehicle production, a more plausible potential impact might be about one-third of the above figures—still significant, but not enough to explain the May drops in the regional manufacturing surveys by itself. Of course, the impact will vary from one survey to another based on the share of vehicle-related producers in the sample.
3. Imports and inventories. The interruption to Japanese production should push down US imports of vehicles and parts. Monthly imports from Japan in these categories are roughly $4bn; a temporary reduction of half or more of this amount is quite plausible. Lower imports and lower domestic production should reduce inventories of vehicles and parts by a comparable amount.
4. Inflation. Tighter vehicle supplies will give manufacturers and dealers more pricing power. Vehicle prices have risen almost 2% in the first four months of 2011 (some of which occurred before the earthquake and its effects), and could rise further over the summer. The move so far has been worth roughly 12 basis points (bp) on the headline CPI and 16bp on the core CPI.
5. Consumer spending. The combination of higher net vehicle prices and a smaller selection of vehicles on dealer lots may slow spending somewhat, although our auto analysts expect this impact to be fairly small (see “Adjusting estimates on Japan, maintaining Attractive coverage view,” Americas Automobiles Equity Research, April 14). Reports on vehicle sales for the first half of May do suggest a notable slowdown from April, although how much of this is due to supply chain issues versus other factors (such as high gasoline prices) is unclear.
6. Labor market. We estimate a cumulative addition of about 70,000 new jobless claims over the last four weeks from vehicle producing states (see "Rise in Jobless Claims Mostly Temporary", US Daily, May 16, for more discussion of the auto sector's role in the recent surge in jobless claims). US payrolls in the vehicle and parts manufacturing sector are about 700,000, so if the 10% March-to-trough reduction in output were fully reflected in employment, this would also imply a drop of about 70,000 jobs. While significant short-term layoffs are typical in the vehicle production sector, particularly around annual model changes in the mid-summer, this number seems quite large. We suspect that employment will not actually decline one for one with production in the short term; feeding the claims increase into our standard payroll models suggests an impact of perhaps 20,000-30,000 on May manufacturing payrolls. We caution this estimate is quite tentative; we will be able to refine it with information from the ISM manufacturing employment index and the May ADP Employment Report.
7. GDP. The overall impact on GDP growth in Q2 is apt to be around 50-60bp, slightly higher than the 30bp-50bp we estimated in our previous comment on April 20. The main reason for the difference is a larger estimated GDP share for the vehicle supply chain--about 2%, versus the 1.3% we estimated previously. (This has the effect of increasing the figures in the right-hand column of the grid below; the other numbers are identical to our previous calculations. Note that simply dividing Commerce Department data on "motor vehicle output" by GDP will overestimate the share because the latter includes margin on used vehicles and does not net out imported parts.) Taking this larger estimate and other recent data on board, we are revising down our second-quarter US real GDP growth forecast to 3.0% (annualized), from 3.5% previously.
Luckily, these effects should be short-lived. US vehicle production is expected to bottom out within the next few weeks and move higher over the summer. This will provide a small positive boost to growth to industrial production and GDP in the second half of the year. New jobless claims have already fallen back somewhat, and any effects on payroll employment should unwind over the next several months. Higher vehicle inflation and any supply-chain related weakness in vehicle sales should be temporary effects as well, probably fading by late in the year.