Posted by: Josh Lehner | April 24, 2014

Graph of the Week: Oregon Counties Growing

In the aftermath of the Great Recession, it has taken awhile for the recovery to spread beyond Portland first, and then beyond the state’s other metro areas. However, with the latest county level employment report, 4 out of 5 counties in Oregon have experienced job growth in the past year. This share of all counties is nearly back to the levels seen in previous expansions. Of course the key question is the magnitude of the growth, not just an up or down measure. Next month the Employment Department will release an updated, benchmark series for the counties so we can examine the strength then. For now, most of Oregon is seeing growth and will continue to as the expansion continues and gathers at least a little bit of steam.


Posted by: Josh Lehner | April 22, 2014

Hours Worked in Oregon

While the state is getting there in terms of the total number of jobs, increased attention is being paid to the other measures of the labor market — e.g. see the spider chart. While jobs are very important, hours worked and wages obviously are as well. The typical pattern is that during a recession businesses lay off workers but also work the remaining employees fewer hours due to slack business conditions. In recovery, as sales and demand picks up, firms begin working employees more hours before hiring additional workers. For this reason it is important to also focus on the total number of hours worked in the economy, not just employment counts.

The first graph below illustrates the decline in total hours worked during the recession and then were each metropolitan area in the state is today (well, the past 12 months) relative to pre-recession peaks. Note: This is the same methodology and graphing style used previously to highlight trends by region and by industry across the state. 


A few items stand out. First, the state lost about 8.5 percent of its jobs during the recession, however total hours worked fell further, to nearly 12.5 percent. Today, jobs are about 2 percent below pre-recession peaks but hours worked remain about 4 percent below. This implies that both the average work week and number of jobs have improved but the jobs account for the majority of the improvement (see graph below).

Additionally, the Portland MSA recently returned to peak in terms of jobs, but it likewise has as well in terms of total hours worked. This means the average work week is all the way back, as are the number of jobs. One surprising finding in this work is Medford. As part of the epicenter of the housing bust, Medford lost nearly 12 percent of its jobs but the average work week pretty much held steady. So far in recovery Medford remains about 8 percent below in terms of jobs but just 2.5 percent below in total hours worked, due to the average work week continuing to increase.

Each metro in the state has its own pattern of recession and recovery, however impacting all areas is the changing nature of the industry mix. Manufacturing and financial activities both have above average work weeks compared to the total and these industries experienced much more severe recessions and slower recoveries thus far. One concern is that the economy is replacing these types of jobs with part-time, generally low-wage service jobs. As our office has shown — see job polarization and part-time employment — that while these trends are a part of the story, they are not the majority of it. Likewise the data presented here, also shows that aggregate hours worked is improving and in a place like Portland, it is all the way back, right alongside the number of jobs.

The second graph decomposes the improvement in total hours worked to show how much of the gains are due to a longer work week and how much are due to more jobs overall. This is slightly different from above in the sense that the first graph allows the peaks and troughs to vary across areas, given that local business cycles are not perfectly in sync. This graphs, however, shows the improvement in just the past 4 years and fixes the start and end dates.


Posted by: Josh Lehner | April 17, 2014

The Plight of the Unemployed

One reason the spider chart is so important is the fact that while the economy is recovering, not all is well with the labor market. As our office detailed much more thoroughly in the previous three part series on unemployment in Oregon (Part 1, Part 2, Part 3) the divergence in outcomes for the short-term and long-term unemployed is particularly pronounced. Furthermore, given the depth of the Great Recession and lackluster recovery, we saw the ranks of the long-term unemployed swell to historic highs (at least in recorded history).

How does the labor market function, if at all, for those unemployed relatively recently compared with the long-term unemployed? Well, according to updated BLS data, the good news is the probability of finding a job if one is unemployed has improved in the past couple of years, regardless of how long one has been unemployed. However, the longer the spell of unemployment, the lower the probability of finding a job and the higher the probability of dropping out of the labor force all together. Only those unemployed three months or less have a higher probability of finding a job than dropping out. All data are U.S. figures.


This second graph shows the probability of finding a job over time. A few things stand out. First, the probability of finding a job during the housing boom was lower than during the 1990s and it is even lower post-Great Recession. Second, for each duration of unemployment, the probability of finding a job today is higher than it was last year and certainly above the lowest rates back in 2010. Third, notice the 53+ weeks line back in 1999. During the tight labor market of the late 1990s — arguably the last time the U.S. economy was truly at full employment — we saw employers begin to hire the long-term unemployed in greater numbers. In order to bring down the long-term unemployed today, at least find them jobs and not have them drop out of the labor force all together, the labor market will likely need to tighten sufficiently for the new entrants and recently unemployed before employers will begin to hire the long-term unemployed.


Focusing just on the long-term unemployed, recent changes in their labor market prospects have been both good and bad. Since 2010, the probability of remaining unemployed has fallen nearly 7 percentage points. About 30 percent of this improvement is due to more of the long-term unemployed finding jobs, however the other 70 percent decline in the transition probabilities is due to an uptick in those dropping out of the labor force. So the long-term unemployed are able to find jobs at a slightly higher rate than in recent years, however they are also more likely to given up their search, at least for the time being.


These are certainly important trends to follow in terms of gauging the overall health of the labor market. Given the large numbers of long-term unemployed, keeping them in the labor market and finding jobs is a major economic issue today. The data above come from BLS research series, which are not updated every month, so our office will highlight them when they are, such as this latest update which became available this week.

Posted by: Josh Lehner | April 15, 2014

Tax Day 2014

Today is the personal income tax filing deadline, and I know you are nearly as excited as we are! As you know, Oregon is a personal income tax state, making up 87 percent of the General Fund this biennium. While 2013 saw a big improvement in the Oregon economy, our office’s and other states around the country’s expectations are for modest growth this year due to the pull-forward of investment-type income into 2012 as federal taxes were raised on high income households.

However when you combine a more volatile underlying economy, such as Oregon has, with an income tax, which is even more volatile, the opportunity always exists for an “April surprise,” either to the up or downside. Our office has experienced a number of those over the years. Nevertheless, so far, so good in terms of collections and expectations. However, as can be seen in the graphs below, the vast majority of income tax payments have yet to arrive at the Department of Revenue and/or be processed, at least in terms of the dollar amounts. It makes for an interesting time in our office as we are effectively in wait and see mode for the next few weeks, but at the same time our colleagues at the Department of Revenue are sending us the latest data on income tax returns.

For more on these graphs, please see Karen Weise at Businessweek who wrote about these trends and patterns last week after seeing our graphs from last year.

The first graph shows cumulative final personal income tax payments since the start of the year. Data is updated through yesterday (4/14/14). Only about one quarter of expected final payments have been processed so far, however by mid-May we should have a good idea where the final number(s) will land.


On the refund side of the ledger, about 60 percent of expected refunds have been processed at this point (data through last Friday, 4/11). Typically filers who receive refunds file earlier in the tax season. This is likely due to a) they’re getting a refund and want their money and b) these are usually wage earners with reasonably simple tax returns; the more complicated returns are usually filed at the deadline (with or without an extension).


Our office’s next quarterly forecast is scheduled to be released on Wednesday, May 28th.


Posted by: Josh Lehner | April 11, 2014

Oregon Spider Chart

Oregon’s economy is recovering. Job growth has picked up as have wages and the unemployment rate is falling (for both good and bad reasons). Even still, the labor market is not 100 percent healthy as we still have too much unemployment, in particular among the long-term unemployed, the hiring rate is not fully back and workers are not confident in the economy to quit their jobs, for fear they will not be able to find another and the like. Given that no single measure can convey the full situation, the Federal Reserve Bank of Atlanta developed an elegant spider chart last year that lets you visualize the improvements in the labor market across a wide variety of indicators. Below is our office’s Oregon version of the same chart, or at least including similar metrics.

What this chart shows is pre-recession peak levels of each measure along the green line (100%) and each measure’s recessionary trough at the red dot (0%). Over time as these measures improve along with the economy from the worst of the recession, they move outward from the center. Once the measures reach the green line, that indicates it has fully regained the level (or rate) last seen prior to the Great Recession.


10 of the 12 measures are Oregon specific ones, with the quits rate and jobs plentiful measures being national as no local data is available. Overall, the labor market is getting there. The leading indicators have improved the most, but now employer behavior is picking up. The number of help wanted ads is all the way back as employers are looking to hire. While hiring rates have improved, as too has total employment, average hours worked is lagging somewhat relative to these other measures. Where the labor market has shown the least improvements are in the confidence and utilization measures, which can broadly be considered the “feel good” part of the recovery. We are not to that point yet, but progress is being made, albeit slowly.

Posted by: Josh Lehner | April 9, 2014

Price to Rent Ratio, an Update

With growing demand for housing and a limited supply, prices are rising considerably in the past two years for both owning and renting. Below is an update of our office’s standard price to rent ratio that uses the FHFA home price index and the Owner’s Equivalent Rent from the consumer price index data (per SF Fed methodology). This data show that in the past two years (late 2011 – late 2013) home prices in Portland rose 12.4 percent and rental inflation increased 7.1 percent. This results in a rising price to rent ratio (indicating that buying is becoming more expensive relative to renting). Salem home prices rose 3.2 percent based on the same data, and has a more stable price to rent ratio in recent years.


The home prices and rental inflation factor used above are trying to gauge the overall price movements of the housing stock and it does a pretty good job of that. However, one interesting item of recent research has been to try and focus on the current market prices of what it takes to buy or rent today. Instead of average rents or prices over the entire housing market, examine just what it takes to find a place to live in today’s market. For new owners/renters/movers, are purchase prices rising faster than rents, or vice versus?

Luckily here in Oregon we have the great Multifamily NW apartment reports with rental price per square foot data, which is a fabulous resource and one I have been trying to find a way exploit more. It also includes data for more than just Portland so matching the rental price data with the recently sold home price data from Zillow, we can create price to rent ratios for Salem, Eugene and Bend as well. This measures price changes on a per square foot basis. One can do the same analysis but with just Zillow data, however their rental data only begins at the end of 2010 so there is not as much history.

PriceRentZillowMFNWExamining the housing market in terms of what it takes to find a place to live today, it shows a much more stable relationship between buying and renting. Although both are rising very fast, they are more or less moving in tandem.

On the outlook, our office is still looking at moderated price increases as eventually a supply response will come (more on this soon). Permitting for new apartment construction is up considerably post recession and I think the headline from the latest Barry Apartment newsletter out of Portland says it all: “Construction Ramps Up As Developers Play Catch Up.” At the least it appears there is no longer acceleration in price increases, even though they are still going strong.

Next week the Spring 2014 Multifamily NW apartment report will be released and we can see just how strongly rents increased in early 2014, as we know home prices continued upward. Future work will include exploiting the neighborhood level rental price data for Portland, as Zillow homes sold data is available as well.

Posted by: Josh Lehner | April 7, 2014

Graph of the Week: Income, Rents and Zip Codes

I have been spending a lot of time at the zip code level of all things housing lately, including prices, rents, sales, and supply in Oregon. Our office has lots of good work coming on that, however, one item that popped up recently was the impact of rising rents on folks across the income spectrum. FiveThirtyEight’s Ben Casselman examined inflation across income groups, based on consumption patterns and found overall the inflation may be rising fastest for the basket of goods the poor and the rich buy (for different reasons) than it is for those in the middle of the income spectrum. The primary reason inflation may be higher for the lower income households is due to rising rents (housing and utilities account for 45% of spending, compared with about 37% for the wealthiest households), based on the BLS inflation and consumer expenditure survey data.

To try and get at the same issue but looking though a different lens, I cross referenced the Zillow rental data by zip code with 2008-12 American Community Survey data on median household income (Spreadsheet: Zip Code Rents). While we know home prices and rents are rising quickly, we also know they’re not rising uniformly across areas. Different cities are seeing larger or smaller increases and neighborhoods within cities are not seeing the same increases, due to various aspects like perceived livability, trendiness, available supply and the like. Putting this all together, we see that rents are actually rising fastest in wealthier zip codes and slowest in lower income zip codes.


While this look at rising rents is a bit more granular and detailed, unfortunately it still does not fully address the underlying question of how rising home prices and rents are directly impacting individuals across the income spectrum as each zip code has both higher and lower income individuals in it.

Bringing it full circle back to Oregon, rents in Portland are rising faster than in Albany and within Portland rents for zip codes in and around downtown are rising faster than, say, east Portland. Adding in spatial analysis and maps may bring another level of understanding to the issue and the impact of rising prices. Our office’s housing affordability measure is hovering just in the threshold range of affordability, but drilling down to the zip code level reveals certain areas are more expensive relative to the median household’s ability to pay than others (more on this later). It’s a complicated issue with different neighborhoods within a city seeing different impacts, but one worth tracking and our office will address some of these topics with our future work on housing at the local level.

Posted by: Josh Lehner | April 4, 2014

Getting There. Financial Crises Edition.

Friend of the blog, Bill from the indispensable Calculated Risk, asked for an update on the financial crises comparison graph now that the U.S. has nearly regained all of the Great Recession’s lost jobs. The private sector is back but the public sector is still down considerably. Expectations are that the U.S. will be back to pre-recession levels in the next 2-3 months (although this leaves to the side the fact that the population has grown).

Just as a reminder the graph compares employment loss for Carmen Reinhart and Ken Rogoff’s so-called Big 5 financial crises with the U.S. Great Depression and Great Recession. Using Reinhart and Rogoff’s great work on historical financial crises, our office previously looked into comparing the Great Recession with these historical financial crises to see how the current cycle stacked up. Overall the Great Recession was your “garden-variety, severe financial crisis” as Rogoff once said, but in terms of employment, the U.S. has actually done fairly well when compared with these other, major crises. Not good enough overall to avoid mass unemployment and lackluster growth, but in the context of historical financial crises, the U.S. employment picture is better than most.


Posted by: Josh Lehner | April 2, 2014

Update: Job Polarization in 2013

New occupational data for 2013 was just released by the Bureau of Labor Statistics. Overall 2013 was not only stronger than previous years, since we already knew job growth accelerated, but the pattern of growth across occupations was also more encouraging.  Below is an update on job polarization in Oregon, following the same methodology and sources as our research report last year.

In 2013 Oregon outperformed the nation and the average state just a little bit overall but significantly when it comes to the high-wage and upper middle-wage occupational groups. This is the business cycle kicking in, or the state’s beta kicking in. Oregon tends to outperform during expansions (and fall further in recessions) and the data show this at the occupational level, particularly among the upper middle-wage jobs. These include construction and installation, maintenance and repair occupations which are increasing with the improving economy and housing rebound but also other occupations tied more to the public sector like protective services (firefighters, police officers) and teachers. These gains are also directly tied to the improving economy and stronger cycle as public sector budgets follow the general economy, albeit generally with a lag.


In terms of each occupational group in expansion, the graph below clearly shows the polarization pattern of strongest growth at the high- and low-wage levels. However if you compare with the graph from last year (with data through 2012) the middle-wage occupations have shown improvement. Additionally we have made one, relatively minor, methodological change in terms of the health support occupations. See below for a full explanation.


Finally, even as growth in Oregon has spread beyond the Portland Metro in the past year or two, it still largely remains within the other metropolitan areas of the state, namely Bend and Medford but also Salem as well. Rural Oregon, while seeing some job growth, remains further below pre-recession levels than the state’s metropolitan areas.


Overall, Oregon’s labor market in 2013 saw an improvement relative to the lackluster growth so far in expansion, however the state is still digging out from the Great Recession and has a lot more room for improvement.

* You may remember that health support had previously shown some large declines relative to everything else and the U.S. overall, which in consultation with the Bureau of Labor Statistics largely turned out to be a data and classification issue. Many of these jobs are now showing up in personal care as personal care aides, and not occupations like home health aides or the like, even though they effectively are the same job, or at least type of job. Given that the data clearly show that these are offsetting each other, our office is now including personal care aides in health support and not personal care. Somewhat confusing, I know, but this does capture the underlying trends better.

Posted by: Josh Lehner | April 1, 2014

Graph of the Week: Gold Mining in Oregon

Our office fields a wide variety of research requests, many of which we are unfortunately unable to help with due to time, staff, data availability, research topic and the like. However, recently we were asked about historic gold mining and production in Oregon. After some quick research and between our colleagues at the Oregon Department of Geology and Mineral Industries, which has what is commonly referred to as Bulletin 61 (pdf), and the U.S. Geological Survey’s Mineral Yearbooks, it turns out the data is available back to the 1880s. As our office is friends with nearly all things data and all things research, putting these together yields the following for Oregon gold production in troy ounces and the latest installment of the Graph of the Week. Keep in mind that these are reported gold production figures and are very likely an under count of actual production.


The two primary regions of the state that experienced a gold rush were Southern/Southwestern Oregon and Northeastern Oregon. During the late 1800s, Oregon accounted for 3-4 percent of U.S. gold production, based on what I could find (historical data can be spotty). From around 1900 through the 1940s, Oregon accounted for 1-2 percent of U.S. gold production and since the 1950s, hardly any reported gold mining at all. There has been mining but the specific amounts are withheld due to confidentiality/disclosure issues. In correspondence with the USGS, they write that the last time they saw commercial gold production from Oregon was in 1994 with most of it placer production and the last lode mine information was 1987.

So there you have it. While historic gold mining is way outside of our office’s wheelhouse, we try to help out the best we can and thought I should share what I was able to find. This is but a scratch on the surface of a much bigger, historical topic. The more I started to look into it, the more I found to discover so I’ll sign off here on an interesting part of Oregon’s history.

Older Posts »



Get every new post delivered to your Inbox.

Join 253 other followers