Posted by: Josh Lehner | November 17, 2020

Oregon Employment, October 2020

Tomorrow our office is releasing the latest economic and revenue forecast for the state. We have a lot of new material regarding the outlook to keep you busy during the holiday season. But first, our friends over at the Employment Department released the October jobs data this morning.

Overall, the report continued to bring goods news and the recovery remained intact. To the extent we did see some relative slowing the September data that was fire-related, it looks to have reversed, as expected, in the October data.

A couple additional notes. One, this month’s report incorporated the first round of revisions to pandemic-related data. It looks like the severity of the recession was a little bit worse than the initial estimates indicated, but overall the the cycle remains pretty much the same. (Our office built in these expected revisions to our forecast that will be released tomorrow.) Two, OED highlights the declines in education employment in Oregon. It was clear that education employment was down — see our previous work for more on why and how — but we really need to wait until a month or two into the school year to get the most accurate picture. Education and related employment will not fully recovery until the pandemic is over and schools return to in-person learning.

Speaking of the pandemic, we are at a crossroads of sorts. On the one hand COVID cases are surging across Oregon and around the country. On the other hand, the vaccine trial news continues to be better than expected. As such the end of the pandemic is likely in sight. We just need to make sure we are able to keep the amount of permanent economic damage to a minimum between now and then to ensure a stronger, faster recovery once it may begin in full force. These next handful of months will be our most challenging yet.

The key measures to watch for that permanent damage are the number of business closures and associated layoffs. While the little bit of information we have on real-time closures (bankruptcy filings, liquor license renewals, video lottery retailers reporting revenue) look OK, or at least relative to initial expectations, we know the longer the pandemic drags on the more damage that will accumulate. In fact, now that we are into our ninth month of the pandemic, we are starting to see things like long-term unemployment increase. There are real economic, human, and social costs to long-term unemployment. And we know that the longer someone is unemployed the lower the probability they will find a job becomes, and the higher the probability they will drop out of the labor force entirely.

In order to not end on a down note, I wanted to share one final chart that tries to show that not everything is terrible. I’m borrowing this concept from our counterparts in Washington who use a similar chart that I thought was useful and interesting.

While total spending in the economy remains down a little bit, we have clearly shifted our consumption basket away from services like going out to eat and getting haircuts and more to physical goods. We have been buying RVs and camping equipment to get outside over the summer. We have been sprucing up our yards, remodeling our homes, stocking up on toilet paper, and cooking and drinking at home to a greater degree. And as much as ever before, we are doing so by buying online as opposed to going into stores. As a result, employment in sectors that sell these products, or move, store, and deliver them actually have higher levels of employment today than a year ago. Note that in past cycles, state government tends to increase initially in recessions as more workers are hired to process claims for needs-based programs before traditional budget cuts reduce employment in the year(s) after.

Stay tuned, we have lots more information on the economy and revenues tomorrow with our forecast release.

Posted by: Josh Lehner | November 13, 2020

Working from Home on the Decline, and Outlook for Downtowns

Working from home has been a key economic response to the pandemic. As detailed before, about 1 in 3 workers are theoretically work remotely. We know that number was higher earlier this year as working from home wasn’t so much about optimal business operations as it was to stop the spread of the virus and still get some work done. The key question remains just how large of a permanent shift we will see in the years ahead. It will increase. And recessions have a way of speeding up structural changes, but where exactly it lands is still very much to be determined.

While we know working remotely wouldn’t stay at these high levels forever, even if telecommuting a day or two remains an employee perk, the question was how quickly would workers be recalled to the office? The answer it turns out is, at least to me, sooner than expected. Since May, the Bureau of Labor Statistics has been asking some supplemental questions about the virus and remote work. If you look at Americans with a job, the share who are teleworking due to the pandemic has been declining in recent months. It has fallen by more than a third, going from roughly 35% to 20% of all employees.

If you dig into the data, it shows this decline is across the entire economy. The share of workers in every industry, or in every occupational group has fallen. The overall decline is not due to compositional factors, like the fact a lot of the job growth recently is in leisure and hospitality as bars and restaurants reopened. That sort of industrial or occupational mix accounts for less than 10% of the decline, while the within industry or within occupational declines resulting in more than 90% of the overall drop in telecommuting. We’re still on track for remote work to set all sorts of records in 2020, but workers being recalled to the office is maybe happening even sooner than expected.

This issue has big implications for downtowns, or commercial centers more broadly across the country. These hubs of economic activity are reliant upon demand from local residents, yes, but also from commuters who spend the workday in the area, and also demand from being destinations for those looking to shop, eat, and partake in nightlife and entertainment. This destination demand includes both those traveling from within the regional economy and also tourists.

With the pandemic, working from home lowered the daytime demand for restaurants and shops as workers commute less. This removed one key source of sales for businesses in downtowns and commercial districts. It also means it boosted relative demand in suburban areas as the workers stayed home. Then on top of commute trends, the drop in tourism — particularly air travel — and local residents staying home more to socially distance themselves, it all hit these hubs of commerce especially hard.

Note: Other Services includes things like barbershops, nail salons, parking lot/garages, and laundry services which are more prevalent in downtowns.

So what does all of this mean? It’s still too early to know for sure. On the one hand, the fact workers are being recalled to the office in recent months bodes well for businesses in these commercial districts, leaving to the side any discussion of pubic health. It is also likely an indication that there is no massive change in business operations on a permanent basis*, or at least not yet.

On the other hand, a partial rebound in the lunch crowd, while helpful, is unlikely to keep all of these restaurants and shops afloat. At best it is a necessary, but insufficient condition for jobs centers and commercial districts to rebound. There remains key risks to the outlook for these locations and commercial real estate more broadly given the ongoing shift away from brick-and-mortar retailers and a potential drop in demand for traditional office space and high rise apartments. We know a full rebound is unlikely until the pandemic is over. At that point, people will go out to eat more, take in a concert, go on vacations, and the like.

While our office expects no real long-run damage to cities and downtowns, that doesn’t mean there won’t be issues this year and next. As City Observatory notes, there are a lot of reasons to be optimistic about cities. In the meantime our office is watching employment trends in office-based work, and tracking measures regarding working from home, dining out, and travel. To the extent there are larger, permanent changes, it provides opportunities to repurpose existing spaces to meet whatever that future demand is.

*Note that the BLS data only refers to workers who telecommute due to the virus. It does not include those working remotely not due to the virus. As such, to the extent permanent full-time remote work is already increasing, that is not captured here.

Posted by: Josh Lehner | November 6, 2020

Macroeconomic Policy Matters

As our nation gets clarity on the outcome of the election, a few economic-related items stand out. While the economy is currently digging out from a severe recession, I think many voters understand or believe that this cycle is different. It was an exogenous, or outside shock that sent the economy into a tailspin. So when some election models use the current state of the economy, or economic growth directly before an election as a predictive variable, it could overstate the expected weakness for the incumbent if today voters are gauging the economy in a broader sense given the nature of the cycle. And I think in that light it is hard to argue that Americans, and Oregonians of all stripes have not seen good economic growth in recent years. We clearly have.

This brings to mind a discussion that has been taking place in the economic twittersphere this year. Macroeconomic policy matters. It’s hard to pin down exact policy levers to generate a strong economy — most policies are about creating incentives to achieve desired outcomes, or directly supporting residents and businesses when issues arise. But as our office has said before, a strong economy works wonders, even if it cannot cure all ills. And I think in this context, the cumulative impact of the decade long economic expansion that just ended has maybe not been appreciated enough, especially the material impact it has on those who have generally been left out due an underperforming economy for much of the past 20 years.

What I mean by that is the U.S. economy had not been at full employment since the late 1990s. We had a very brief period in the mid-2000s of a decently strong economy, and then in the late-2010s an extended period of a strong economy. But in between that, the economy was downright bad to terrible. As a result, we have kind of redefined normal. From, say, 2000 through 2015, the normal state of the economy was basically income growth that kept up with inflation, but nothing more for the typical worker or household. But that all changed in recent years.

The economy, near the end of the decade long expansion, broke through that malaise after the labor market became sufficiently tight. We were finally seeing real, solid gains across the spectrum. Firms had to dig deeper into the resume pile, and cast a wider net to attract and retain workers, because most people who wanted a job, had a job. Labor was not as easy to find. The result is wages started rising. Given that wages are the only real source of income for most households, the stronger labor market where more people were working, for more hours, and at higher pay, really started to move the needle on income growth. We got so used to a weak economy, that when we finally saw a strong one, it really resonated with those most impacted. Now, whether or not one should assign credit for this process to whomever is currently holding office is a different discussion, and I digress…

So with that said, I’ll leave you with a few charts to this end. The economic expansion that begin in the summer of 2009 and ended in the spring of 2020 really did start to bear fruit in recent years.

You can see that in income growth for all parts of the state, where they are now higher on an inflation-adjusted basis than back in 2000. (Once more detailed 2019 ACS data is available, I will update our more granular regional income work.)

Looking at growth rates over 5 or 10 year periods shows just how different a strong economy delivers for the typical household than a weak one does.

And this third chart is one you have seen before, but we cannot lose sight of the fact that while large racial disparities remain, the strong economy was really delivering results for Oregonians of all races and ethnicities. Poverty rates for Black, Indigenous, and People of Color are at multi-decade, meaning multi-generational lows. That is something to celebrate, even as we realize there is more work to be done to help address ongoing disparities.

To bring this full circle, it is clear that a strong economy delivers results and puts more money in the pockets of everyday Americans. We still need public policies to help address issues, inequities, increase efficiencies, and the like. But macroeconomic policy matters considerably. Full employment matters. Americans notice when it does happen. And to paraphrase the economic twittersphere, we should try doing that!

Stay tuned. Our next economic and revenue forecast is released in a couple of weeks, Wednesday, November 18th. We will have more thoughts on the potential economic impacts of the election as we know more.

Posted by: Josh Lehner | October 27, 2020

Zoom Towns are Real (Graph of the Week)

Stories about pandemic migration and working from home abound, even as we are months away from any hard data. This can be a struggle, balancing these anecdotes and speculation when we lack any real evidence. And the issue isn’t that these things are not happening. They clearly are. The issue is that Oregon sees hundreds of thousands of people move every single year. What really matters is when any changes are widespread enough to be more than noise.

To that end, we just don’t know yet. We will get Oregon population estimates from Portland State in December, but need to wait until April for the full set of underlying data tables. And the 2020 ACS data won’t come out until September 2021.

The key question is to the extent we do see pandemic migration, will that be enough to offset the fact few people moved during the shelter in place phase of the cycle, let alone offset the traditional slowdown in migration during recessions? Statewide, our office does not believe so, but our forecast may be too pessimistic given this is not your normal business cycle. Plus regional patterns will always differ from statewide figures.

Our view is further clouded by the fact the DMV is appointment only these days due to social distancing. As such we have lost our traditional real-time migration indicator: surrendered driver licenses. That leaves us looking at things like home sales as a measure of mobility, even as it is hard to disentangle local moves vs out-of-state migration.

Here we see ongoing strength in housing due to the fact higher-income households have been less impacted by the recession to date, interest rates are at record lows, and there is the strong demographic tailwind of Millennials aging into their 30s and 40s.

Even so, all of these factors bring home sales nationally and in some popular metro areas to a point where 2020 will show solid to very strong gains. But there are some places experiencing eye-catching growth. Home sales are up double digits among Conor Sen’s “Zoom Towns” — popular, smaller, scenic areas from which workers can telecommute. Keep in mind that given the ongoing strong demand and pending sales numbers, the final year-end figures will be up even further across all locations.

UPDATE: The chart has been updated to reflect better Missoula home sales data. The previous chart showed 22% growth which was incorrect. The true number is 9.8% per local MLS statistics. A big thank you to Brint Wahlberg, a Missoula realtor, for spotting the error and correcting the data. My apologies for the error.

One of the key questions earlier this year was whether we would see people double down on existing patterns of migration and places to live, or whether we would see new patterns emerge. Keep in mind that it is still early in the cycle and the winds may change as it drags on, but early returns indicate we are doubling down.

Now, I want to touch on a couple related issues. The first is inventory, which is down about 40 percent a lot of places in Oregon, and 20 percent nationally. Normally inventory does not actually restrain sales. The main reason inventory is low is because sales are strong! But this year in some places it may actually be an issue. Take Hood River for example. You can currently count on your fingers the number of homes for sale, even at luxury price points. This extremely lean inventory is problematic for the market. But more generally, lower inventories make it a sellers’ market, and prices appreciate faster.

Second, I think the major increases in luxury homes are distorting some of the overall market figures. Some media articles out of Montana indicate the $1 million+ properties are a hot segment. In Bend, sales have been very strong above $750,000 with pending sales of $1m+ through the roof. Now, it’s hard to say that this means all that much for middle-income families, but it does drag up the average, and even the median home price due to a big compositional shift.

Third, it remains an open question as to what all these home sales actually mean for migration and population growth. Most buyers are usually local. They tend to be either first-time or move-up buyers. However a part of the strength is tied to migration as well. Increases of 10-15% in the Zoom Towns is not due to local demand alone. Although an added question is whether these sales represent true migration and population growth, or simply increased demand for second homes. That distinction may not matter for the housing market, but it does for the local economy.

Bottom Line: The housing market is among the strongest segments of the economy. Ownership is expected to remain strong due to demographics, low interest rates, and the nature of the cycle. Even so, popular, scenic areas have experienced an even greater influx of demand. Zoom Towns appear to be real and pandemic-related migration looks to be doubling down on existing patterns of growth.

Posted by: Josh Lehner | October 23, 2020

In the News: Oregon High-Tech

Yesterday Intel, the pillar of the state’s tech industry, announced their latest quarterly earnings and continued discussions on manufacturing challenges and the possibility of partnering with other firms for future production. See Mike Rogoway’s piece in The Oregonian for more.

This is something we recently discussed with our advisors as well and what it may possibly mean for not only the high-tech outlook in the state, but even the broader Portland economy. The upshot of their thinking is none of the potential changes are likely to happen overnight. What’s really a potential risk would be if, for whatever reason, the Portland region was no longer the main hub of research and development. It’s not just the direct manufacturing jobs that would be at risk, it would be all the engineering, management, office jobs and the like. To date there has been no discussion along those lines, which is encouraging. Plus the overall tech market looks to be strong in the coming years, helping support sales and activity.

With that in mind, it gives me an excuse to post a few charts on a Friday.

First, it’s always important to keep in mind that Oregon’s tech specialty has always been hardware, or tech manufacturing. The state has seen strong growth in software in the past decade, but that growth has largely matched national trends.

Second, a breakdown of tech jobs in Oregon by occupation shows that the manufacturing jobs that actually do the manufacturing, so-called production occupations, account for about 1 out of every 4 jobs. To the extent the vertically integrated business model that has served Intel (and Oregon) so well over time changes, the first impacts would potentially be felt among these jobs. That is, if the manufacturing of future chips is outsourced, one may think local production jobs would be lost. But as stated earlier, it’s really the broader cluster of occupations that comes with R&D and headquarter-type operations that accounts for the bulk of local employment.

This third chart drills down on these occupations for just Computer & Electronic Product Manufacturing, and the pure Software firms. It’s interesting to note that the manufacturers have just as many jobs in architecture and engineering, and computer and math occupations as do the local software firms. This is likely one reason Oregon does see some start-ups spin off, it’s because there is a large pool of workers with the skills and ideas to be able to do so.

Finally, just a reminder of what the outlook looks like. Our office expects tech manufacturing employment to hold relatively steady. It remains a key part of Oregon’s economy. The sector undertakes big investments every handful of years that help drive productivity gains not only in the industry but also in the statewide data. But, in keeping with the manufacturing story overall, those productivity gains do not tend to result in more employees. Or to the extent some firms are hiring, we know some of the other firms are downsizing, keeping sector employment relatively steady. The overall growth in high-tech jobs in Oregon will continue to be on the software side.

Note: this is an older, pre-COVID chart. Our office’s outlook for tech manufacturing remains relatively unchanged today, but we have seen some recent software job losses that aren’t reflected in this chart.

Posted by: Josh Lehner | October 19, 2020

COVID’s Impact on State and Local Governments

This started with a seemingly basic question: “Why is public sector employment down so much this year?” The normal pattern we see is that the public sector is more of a stabilizing force in the economy. Job losses and budget cuts come with a delay as it usually takes time for lower levels of economic activity to translate into fewer tax collections. Those impacts usually hit the budget the fiscal year after the recession starts. However, so far in 2020 local governments have shed nearly as many jobs as the private sector. Both the size of the losses and swiftness with which they came is highly unusual.

After digging into the data it is quite clear that the local government job losses are not a result of your standard budget cuts. That traditional recessionary dynamic is likely to come, but will hit next year, not this. The losses today are directly related to the pandemic and social distancing.

Most surprising are the declines in local education. Yes, bus drivers and lunch workers are impacted by distance learning. However the bulk of the declines appear to be teachers themselves. If all we did was switch from in-person to online schools, why are teaching occupations down double-digits? Some is likely due to lower enrollments — students switching to purely online options, homeschooling, or redshirting potential kindergartners. Some districts may be holding vacancies open in anticipation of future budget cuts.

Even so,the primary reason seems to be the lack of using substitutes. With online school, teachers feeling under the weather can better manage a few live meetings with their classes or send additional work remotely, whereas standing in front of the room all day is more challenging. Nationally, subs account for about 13% of all K-12 teaching jobs. Informal conversations with a few local districts confirms that sub use is down considerably, and really only used for durations lasting more than a day or two.

In terms of higher education, the impacts of the pandemic, social distancing, and online schooling are clear. There’s no sugarcoating it, enrollments across Oregon appear to be down double digits. Official counts will be available later this fall. Informal conversations indicate that full-time employee layoffs are minimal to date. Fewer adjunct instructors and professors are being used. The bulk of the layoffs are tied to student workers, recreational centers, dorms, student unions, and the like.

Besides education, the public sector does a lot of things. Employment here is down largely due to zoos, convention centers, recreation facilities, public pools, libraries and the like being limited during the pandemic. The losses in public administration are relatively small to date.

All of that said, there is still the traditional recessionary dynamic at play. Those impacts will largely come next year, not this. Depending upon a given jurisdiction’s population and economy, it may fare better or worse when it comes to local trends in sales, property and lodging taxes. These differences are already emerging in neighboring jurisdictions, across the state, let alone around the country.

The key question in the outlook is whether these traditional budget issues will be in addition to the pandemic and social distancing ones, or simply replacing them.

Full set of slides are below.

Posted by: Josh Lehner | October 15, 2020

Oregon Employment, September 2020

Yesterday we got the September 2020 jobs report from our friends at the Employment Department. The topline numbers continue to be good, or at least probably good enough. The recovery is likely self-sustaining even as it slows as expected. That said, it doesn’t mean we’re out of the woods yet. There remains a Great Recession sized hole left to fill. And remember, the recovery from the Great Recession was self-sustaining too, it was just stuck in a low gear and we grinded our way out over the course of a decade. Our office expects this recovery to be faster, and more complete but that is largely predicated on limiting the amount of permanent damage in terms of business closures and layoffs until the pandemic is managed and brought under control, and a medical treatment of some sort becomes widely available by the middle of next year.

As we’ll get into more in the coming weeks, a few things are emerging in the data. First is there is a clear gap when it comes to employment among women and men. Second, now that we’re in the seventh month — how is it only been seven months? — of the pandemic, we are starting to see a rise in long-term unemployment, commonly defined as six months or longer. Third, the virus remains in control. Yes we are learning to grow around the virus and resume some activities. However large swaths of the economy cannot recover until the pandemic is over. This includes some sectors you may not fully expect, like local governments.

Overall, job growth was decent in September. The state added 5,100 jobs. Normally that’s about as strong as you would ever see, but today it represents a gain about 1/3 as big as Oregon saw in July and August. Overall the state has now regained 45% of its lost jobs. This truly does remain good news overall, even if the path forward isn’t well lit.

I’m switching up this first chart to include our office’s forecast. We fully expect growth to slow further in the months ahead. I suspect that some of the slowing in September is fire-related in terms of less economic activity due to many of our friends and neighbors being evacuated and the smoke keeping the rest of us indoors. Not so much as jobs were cut outright, but more in the sense Oregon saw somewhat fewer gains than the average state. If true, that relative slowing vis-a-vis the nation should reverse in the October data. Even so, its important to keep in mind that the baseline outlook calls for more slowing in the months ahead. That doesn’t mean the recovery is in jeopardy, even if it will feel that way. The combination of any seasonal component to the virus, plus the seasonal issues related to the colder, wetter months when we are forced indoors more, plus some hangover from the lapse in federal aid, and that underlying traditional recessionary dynamic will all weigh on growth in the months ahead. After that, growth should pick up and, again, provided the permanent damage is fairly minimal, the overall recovery will be faster.

One silver lining continues to be that much of the job loss and corresponding increase in unemployment still looks to be temporary. The headline unemployment rate is back down to 8.0% and the core unemployment rate — a measure of permanent layoffs and damage — ticked down as well, even as it continues to be somewhat elevated. This is the part of the recovery that will take the most time, even as many of those on temporary layoff continue to be recalled.

As always, a huge thank you to Tracy Morrissette over at the Oregon Employment Department for his great work digging into the detailed data so we can look at core unemployment in Oregon. Thanks Tracy!

Finally in a good news/bad news update, we continue to see a split in labor market outcomes. Low-wage industries suffered the most during the shelter in place phase of the cycle, given the sudden stop nature of the shutdown. It is here that much of the job growth in recent months is taking place. This is good news as those workers are brought back following temporary layoffs. The bad news is that middle-wage sectors are only seeing very modest gains and high-wage sectors are not seeing any growth in recent months. They may not have suffered as much to date, but they’re also not bouncing back. This is a concern and likely speaks to more permanent layoffs in these industries. Keep in mind that job losses of 5% are more severe than Oregon suffered in the 1973, 1990, and 2001 recessions. Further research shows that many of these layoffs in middle- and high-wage sectors appear to be lower-wage workers within those industries. But overall, the slowdown in the September data is very clear in the chart below.

All told, Oregon’s economy continues to recover and heal. Employment is up and unemployment is down. The recovery is likely self-sustaining but it is expected to slow in the months ahead. The virus itself and the amount of permanent damage that accumluates will determine the overall strength and duration of the cycle. Stay tuned for more next week on some of the issues related to the gender gap and local governments.

Posted by: Josh Lehner | October 12, 2020

Economic Dynamism in Oregon

The key economic concern today is one of permanent damage. How many firms close and how many permanent layoffs occur? While much of the damage has been temporary to date, it’s that underlying traditional recessionary dynamic that will go a long way to determining how long the recovery takes. And in past severe recession in Oregon, like the early 1980s and the Great Recession, we saw both an increase in closures and a drop in new businesses forming. This prolonged the recovery process as it takes time to replace firms and rebuild the network of employees and supply chains in the broader economy.

Unfortunately we lack good, timely data on firm closures. It takes months to realize that a business is not reporting payroll, paying taxes, renewing business licenses and the like. What little information we do have shows closures are rising, but not yet considerably so. One reason may be that we are still early in this cycle, and firms do not close overnight when sales drop. They burn through reserves, maybe take out a loan, and then if those fail to tide the company over, then they close for good. Additionally, Oregon companies received $7 billion loans/grants from the Paycheck Protection Program and while that money is long gone, it was also $7 billion of temporary cushion.

What good, timely data we do have is about new businesses forming. This is one of the key metrics our office is following over on the COVID-19 tracking page. Here we see encouraging news to date. There has been no real drop-off in start-up activity in 2020. It may well happen as the cycle drags on, again it’s still early. But so far the data points toward a rise of business closures but no real decline in new business formation. While not good, this would still be better than those past severe recession, and be less of an drag in recovery.

The rest of this post comes from our office’s June 2020 forecast document, where we discuss some of the implications of firm closures and the broader issues at play. The slides at the end are a full update of these broader issues related to entrepreneurship, start-ups, closures, venture capital and the like. This is something we last really dug into on the blog more than 5 years ago. The update is largely based upon newly released, and revamped Census data on business dynamics.

Finally, the economy was already struggling with lower rates of productivity in recent decades. While an imperfect relationship, productivity gains are a key driver of employee wages and a rising standard of living. Many economists point toward the low rate of entrepreneurship as one key reason for low productivity growth. After all, new products and services are usually brought to market by young companies who seize the opportunity to improve efficiencies and generate profits.

COVID-19 may further entrench these trends. A byproduct here would be that the large businesses that have already won out in today’s economy may further strengthen their positions. Big companies generally have better access to capital markets to raise money to survive, while small firms rely upon the owner’s own worth, and programs like the PPP. Furthermore, benefits that workers generally want, particularly during a pandemic, such as health care, paid time off, the ability to work from home or remotely, and the like, are much more common among large firms than small. When combined overall, these issues may potentially paint a dark picture for small businesses, start-up activity, productivity, and ultimately for workers and job growth.

Posted by: Josh Lehner | October 1, 2020

Wildfires: Impacts in the Forecast

When the fires were raging and the state was blanketed in smoke, our office wrote about some of the implications and impacts associated with wildfires. We fleshed out the discussion in our forecast document (see PDF pg 11) and during our forecast release the other week. See that for the bigger picture perspective. Today I wanted to dig into some of the underlying details a bit further. Keep in mind we still do not know the full extent of the damage, or how things will play out over the next couple of years. But here is what we know now, and what we built into the forecast.

First, what makes this year’s fires different is the fact that they were/are much closer to cities and towns than in recent years. We’re not only losing timber and recreation areas, we’re losing structures and towns and evacuating tens of thousands of Oregonians. As of early September, if you look at the areas within the fire perimeters and those in the Evacuation 3 zones (full evacuation) they encompassed around 2 percent of the state’s population, and around 1 percent of total real market property value and employment. That’s a huge portion of the state to be fully evacuated.

More than one million acres have burned so far. We know fires burn in mosaics so there will be patchwork devastation across the state. Some stands of trees will be in good shape, while others will be gone. Some homes will remain standing, while others will be burnt to the ground. In terms of acreage and ownership, the following table comes from Mason, Bruce, and Girard, a natural resource consulting firm, from which Mark Rasmussen serves on the Governor’s Council of Economic Advisors. Our friends at the Department of Forestry have similar type numbers looking at ownership by fire, not by county.

Broadly speaking there are four main channels in which the fires will impact Oregon. There is the direct impact on sectors reliant upon the forests. There is the destruction of properties that will need to be rebuilt. There is the impact of the evacuations themselves. And then there is the longer-term implications, largely related to migration.

Let’s first look at the sectors that are impacted directly. This leaves to the side for today those sectors impacted by the smoke, where Oregonians couldn’t go outside for a week. But in terms of sectors reliant upon the forest for work, it’s 2-3% of the state’s economy. The timber industry — logging and the mills — account for the bulk of these figures. The potential impacts here are that harvest levels may be impacted or reduced for years to come, depending upon the level of devastation and ownership of the land, etc.

The other 0.5-1% of the impact comes from forest-based outdoor recreation. Overall BEA estimates indicate outdoor recreation is nearly 3 percent of Oregon GDP but half of that is the travel and tourism component (gasoline, lodging, food) and then the actual recreation portion includes things like water parks, festivals, tennis and the like. The hard thing to parse here are activities like cycling and boating. Clearly some of that is forest-based. I enjoy paddleboarding on lakes in the summer and the lakes are located in the forests. But what about fishing on the Columbia? Is that forest-based? And what is the breakdown between cycling around town, or commuting to work vs mountain biking in the forests? Clearly the available data overstates the forest component of outdoor recreation which is anything kinda related to being outside. So some very rough cuts of the data would indicate it is around half a percent of state GDP.

In terms of the destruction of property, what we currently know is the following, This is a very rough cut and estimate, but indicative of what we were looking at as we were adjusting our forecast at the eleventh hour.

Just over 4,000 residences have been burned. If you take median home value by census tract for the different fires and apply that to the number of homes, you get around $1.1b billion in total property values. Mike Cowles, the Lane County Assessor, was gracious enough to help me look at the properties impacted by the fires in his county. What it showed was among the residential properties within the evacuation areas, his data indicated the assessments were 50/50 in terms of the value of the land versus the value of the improvements (structures). So taking that 50/50 split and applying to the 4,000 homes indicates around $575 million in lost homes. Additional research from the Insurance Information Institute (III) shows that homeowners insurance typically includes coverage for personal belongings equal to 50-70% of the value insured of the structure. So, roughly, that’s another $340 million in lost personal belongings, bringing the total to nearly $1 billion. In looking at III research on wildfires, it indicates that during big wildfire years insurance seems to cover about 75% of the losses. I could not find a breakdown of residential vs nonresidential here but that does factor into those numbers. Regardless, the losses and devastation is big, and the final numbers will be even larger than this back of the envelope work from a few weeks ago. Plus none of this yet includes the commercial or industrial losses, or the value of the standing timber itself.

The next table shows what our office built directly into our forecast.

Our office included an additional 2,000 housing starts over the next 2 years. Now, we know we’ve lost more than 4,000 homes but we are only including 2,000. Why? One reason is these were ad hoc adjustments made at the last minute to the forecast and the estimates of the number of homes lost kept climbing by the day. That’s a factor. But also, two key questions relate to whether we will see every lost home rebuilt, and whether there is enough capacity within the construction industry to build an additional 4,000 homes in the months and years ahead. Will Oregon end up building all of the units? Probably. But we may need to reallocate construction capacity away from projects in Portland, Salem, Eugene, Medford and the like, and use those workers and materials in the rebuilding towns. To the extent that happens, then total construction industry production won’t increase as much given we will see some temporary reductions in the main cities until the rebuilding is complete.

The other direct impact our office built into the forecast was related to the Corporate Activity Tax. In the short-term we will see a boost to food and lodging sales associated with the evacuations, and the visiting firefighters helping to battle the fires. Additionally the rebuilding phase will result in an increase in the sales of construction materials, firm revenues and the like. All told our office increased the CAT forecast by $18 million total in 2019-21 and 2021-23 combined. However much of these impacts are in the short-term and will fade out in the year(s) ahead.

Finally, a key concern for the outlook remains any long-term implications from the fires. Our office previously wrote about the migration concerns and how that could impact the long-run outlook for the state.

The scariest potential impacts for Oregon is that fewer households and investments may be attracted to the region moving forward. Oregon’s primary comparative advantage remains its ability to draw skilled workers away from other states. To the extent that local quality of life has been reduced, or if Oregon is perceived as a riskier or costlier place to live and do business, this advantage will be less pronounced. Increased risk lowers growth prospects due to uncertainty and higher costs. If these fears are realized, our office’s long-run outlook would need to be lowered.

Over the summer, ECONorthwest released an economic analysis of Cascadia, which is a different type of natural disaster, but their report included a really good summary of the economic literature. Included there was an interesting paper by Boustan et al (2017) that examined county level impacts of natural disasters. One of the findings in the paper was that fires were associated with larger impacts on migration than most other types of natural disasters. That’s certainly the main long-run concern our office has. To what extent do residents impacted by the fire pack up and leave Oregon? To what extent do the fires deter other people from moving to Oregon in the first place due to the potential threat of fires? The answer here is unknown. Our office has not built in any sort of long-term impact to date, especially considering other climate-related issues likely indicate Oregon is relatively more attractive. Even so. here is how we wrote up one potential silver lining in the document:

Now, one potential offsetting factor for Oregon today is that most, not all, but most of the impacted and fire-ravaged communities are located within the Eugene, Medford, and Salem metro areas. Yes, these communities are largely smaller, and more rural in nature, but they are not far-flung nor geographically isolated. Given that the primary cities, and job centers have not been lost for the metros, it is likely that many residents of the devastated communities still have somewhat brighter job prospects and thus may be more likely to remain and rebuild. It remains too soon to tell the extent of the damage today. Longer-term effects are still to be determined, particularly should such fires become increasingly frequent across the West.

Our office will continue to monitor the available information and how it impacts the state’s economic and revenue outlook. The Governor has put together a Wildfire Economic Recovery Council which our office will help with as needed as well.

Posted by: Josh Lehner | September 25, 2020

Not Fun Friday: Low-Wage Workers and Income Disparities

One of the big takeaways from our latest forecast is the fact that the nature of this recession, combined with income inequality means that while there is considerable economic pain, we have seen hardly any in terms of actual state tax revenue collections. I just wanted to do a quick follow-up today given some new information.

First, our office has been publishing the following chart every month lately when the employment report is released. It is designed to to show that workers in lower-wage sectors have borne the brunt of the recession. We know that many consumer-facing industries are impacted the most by social distancing and the like. Even so, jobs in middle- and high-wage sectors are still seeing sizable losses, just not as severe as those in the lower-wage sectors.

The chart above is based on employment at the industry level and dividing the economy into thirds based on average wages by industry. While imperfect, this is what we have readily available at the state level. And it certainty shows indications of what is happening in the economy.

Well, yesterday, economist Ernie Tedeschi — a great follow on Twitter and one of the best researchers when it comes to digging into the microdata — followed up by looking at employment based on individual hourly wages and not industry averages. What Ernie found shows even larger disparities. Low-wage workers, regardless of industry, haven’t just borne the brunt of the recession, they have borne almost the entire recession alone. Employment among individuals earning more than $16 per hour is essentially all the way back to pre-COVID levels, while employment among individuals earning less than $16 per hour remains down 25-30%. This means the deep hole the overall labor market is in today is entirely due to the lowest-earning third of workers falling off a cliff while most everyone else is relatively OK.

Now, one reason our office has been focusing on the sector level view is because we have good data there. The data Ernie is crunching comes from the household survey. Here in Oregon our sample from the household survey is already pretty small. To get hourly wages from the survey you have to focus on what is called the out rotation group. When I filter the data for just those Oregonians with a job, in the out rotation group, with properly coded hourly wage information, the sample size is around 100 a month. It is very small. That means it is prone to a lot of noise.

Even so, I think the following chart clearly shows that what is happening nationally is also happening here in Oregon. High-wage jobs have seen essentially no impact from the recession. Now, whether the last month of data where low-wage jobs are up a bit and middle-wage jobs are down is truly representative, I cannot say. But in the larger picture, the trends are clear even with the noise.

Bottom Line: We know that the sectors most impacted by the pandemic are generally lower-wage industries. Such workers have borne the brunt of the recession to date. However, if both sets of data are correct and not revised substantially, what the findings show is that the impact of COVID-19 is even starker than first believed. The way the two data sets square is that low-wage workers in high-wage industries are the ones being laid off there, while high-wage workers in low-wage industries have been spared to a great degree. Regardless of industry, workers earning lower wages have borne almost the entire recession. Some of this is due to the nature of their jobs and occupations, the pandemic, their ability to work remotely, their overall work experience, their within firm work experience and the like. But regardless of the exact reason(s), the patterns are clear.

« Newer Posts - Older Posts »