When energy prices rise, it usually eats away at consumers’ discretionary spending which has economic impacts, even if total, overall spending remains intact. One item UCSD Professor Hamilton, over at Econbrowser, tracks regularly is the share of energy expenditures as a share of consumer spending. He uses the rule of thumb of 6% as a warning signal – that is, when consumers spend more than 6% of their budgets on energy, it has economic consequences. For those interested, Professor Hamilton also frequently posts on many different aspects of energy markets, including prices, production, transportation and the various factors that affect these items.

The purpose here isn’t to discuss the impact on consumers’ budgets but rather to examine the current state of energy prices and their relative outlook. The standard graph our office normally uses looks like the following. Until the middle of the last decade oil and natural gas prices more or less followed each other but have since diverged. Note that the forecast figures used here come from the Annual Energy Outlook 2013 produced by the U.S. Energy Information Administration. [Other proprietary forecasts our office has access to do not differ widely from this outlook, particularly for the longer term.]

While it is easy to show the lines diverging when on different axis and measured in different units – price per barrel of oil compared to price per million BTU (MMBtu) – one issue raised with our advisory groups is how do they compare on an energy equivalency. Well, based on my readings, it turns out that this does depend somewhat upon the quality of the products being used so the exact calculation may vary, however a commonly cited factor is 5.8. One barrel of oil is equivalent to 5.8 million BTU of energy and conveniently, natural gas prices are commonly quoted in millions of BTUs. Thus, you take the price for a barrel of oil and divide by 5.8 to arrive at a relative comparison based on energy equivalency between oil and gas. What this means is that today’s barrel of oil, which goes for about $96 (WTI prices) is equal to about $16.55 per million BTU, while the Henry Hub price of natural gas today is about $4. (The data used in these graphs are quarterly and a couple months old, hence the small differences.) That means that natural gas is about 80% less expensive than oil on an energy equivalency basis and the outlook for energy prices projects this difference to be sustained for the foreseeable future.

The issue becomes, what do we do with this price differential and how long can it last? Over time these two products are substitutes so one could reasonably expect there to be some sort of convergence in prices again, however that is a long run solution and takes time. The reason for the drop in natural gas prices is largely due to a supply shock; the increased domestic production has provided larger and larger quantities of gas on the market, putting significant downward pressure on prices. Oil prices on the other hand are being somewhat constrained by the lack of new supply coming online coupled with increased international demand. As Professor Hamilton notes:

Global field production of crude oil … stagnated … between 2005 and 2008. It is up a couple of million barrels since then, but more than 100% of this increase has been consumed by China alone, forcing the U.S. and other countries to reduce our oil consumption.

These forces work to keep oil prices elevated and we have this large gap in prices on an energy equivalency between natural gas, which is effectively being set in a domestic market, and oil, which is being set in a global market.

So what are the benfits of this? From an article on CNBC.com:

Cheap natural gas not only cuts costs for companies that use it as a raw material or feedstock for other products such as chemicals, it also means lower power prices as utilities use more gas to generate electricity.

The article goes on to note that electricity for industrial users on the East Coast is substantially less expensive than some European countries, including about 30% less than in Germany and nearly 60% less than in the U.K. This in turn makes manufacturing less expensive in the U.S. than some international peers and may even lead to the popular topic of discussion: reshoring or insourcing. Or at the least it may slow the tide of offshoring or outsourcing of manufacturing jobs. Cheaper energy prices at home also lowers the cost differential with developing countries, particularly with rising labor costs in places like China. This is all well and good and tells a nice story, however how are we doing on this? Recent research work put together by Goldman Sachs’ economic group sheds some disappointing light on the subject, at least so far. For energy-intensive industries, such as metals and plastics where this cost savings would most likely make a larger impact, their production has actually lagged overall manufacturing production in the U.S. (This figure is taken from Business Insider’s article based on the Goldman Sachs research.)

Additional work done by Cleveland State University (HT: WSJ) shows that even with a shale boom in Ohio, jobs in these regions are not following suit. Furthermore, from a Washington Post article last year, a paper from 2009 finds that “only one tenth of U.S. manufacturing involved energy costs exceeding five percent of the total value of shipments.” To some, all of these findings appear to point to the story that these lower gas prices are not a big deal or will not produce large benefits, however I think a) this isn’t necessarily the takeaway from this research and b) that does not mean manufacturing is doomed. To be fair, as the Business Insider article relays, Goldman Sachs is still bullish on manufacturing overall and writes “[b]ut we believe the reason for this will be broad economic improvement that benefits all sectors, especially the more cyclical ones, not a structural U.S. manufacturing renaissance.”

I think that is a fair assessment. These really low gas prices are certainly beneficial to companies (and homes heated by gas too), improves the relative international cost comparisons, but so far is not a providing a structural renaissance, as some had maybe hoped it would. It certainly has in terms of employment and income for workers in the gas fields (North Dakota unemployment rate: 3.2%) but this has not necessarily translated into a much broader improvement, or at least not yet anyway. Over a much longer time horizon, one would expect there to be some convergence in prices as oil and gas are substitutes. However the costs to switching infrastructure like this are generally large and not done overnight (how often do you replace your home’s furnace?) In the short term, it does appear this price differential is here to stay and natural gas customers do benefit from the increased domestic supply, which does help both business’ bottom lines and household finances.

P.S. I also think the Ohio report is flawed definitionally as many of the workers likely live in the bigger cities, so they show up on employment counts there, not in the more rural areas. Cleveland is classified as a weak shale area while Columbus and Cincinnati are counted as non-shale. The authors do mention this as a possibility, and I would not be surprised to see that it is the answer.

This morning the S&P / Case-Shiller Home Price Indices for January 2013 were released and showed continued home price appreciation across the major cities in the US, including Portland. On a year-over-year basis, the 10 City Composite index is up 7.3% from January 2012, while Portland is up 8.3%. Prices in Portland are effectively back to the same levels as seen in Spring 2005, based on the index. They are up that same 8.3% from the bottom, coincidentally reached last January, but are still down around 23% from their peak in 2007. This continued home price appreciation is just another confirmation that the housing market recovery is here, and is boosting household balance sheets as asset values rise.

As both a major cause and consequence of the Great Recession, our office and our advisors have spent a considerable amount of time thinking about and discussing the housing market and its outlook. One common theme that emerged was that it was highly unlikely we would have another asset bubble in housing again. There was a resetting of consumer expectations in the value of homes, we took this large downward plunge as the bubble burst and moving forward as the economy improved, so too would the housing market. As jobs returned and population grew, this would lead to increased demand, which would lead to more new construction, more homes for sale and modest home price appreciation just a bit over the rate of inflation. While our outlook for the medium and longer run housing market remains this way, in these early stages of the housing market recovery we haven’t seen this nice, clean cut, ebb and flow of the supply and demand. In fact, there is increasing evidence that the rising demand is taking home builders by surprise, which may mean the market will be more supply constrained in the near future. We’ve known for awhile now that we’re actually under-building new homes, even after accounting for the run-up in new construction during the bubble (see this post from nearly 2 years ago.)

Why is this and why are we seeing such strong home price appreciation in recent months? I don’t have all the answers but a lot of it really comes down to basic Econ 101, supply and demand. First, one major factor that does affect the demand for housing is the ability to finance the mortgage. Mortgage rates continue to hover in the mid 3 percent range, and possibly more important today, lending standards are loosening somewhat. You still are required to fill out paperwork documenting your income and the like – I haven’t heard of any recent NINJA loans – but the banks and brokers are underwriting mortgages for a somewhat wider audience than in recent years. This is a welcomed development and does support housing demand in the market.

The graph below shows annual data for Metro Portland based on the year-end RMLS reports. The actual data here are not perfect measures of supply and demand but they are the data I was able to pull off the reports and certainly act as good proxies for the real underlying series. I will not rehash the changes over the past decade, however it is hard to remember, at least for me, that 2005 was the top of the buying market in terms of volume. 2007 was the top of the market for prices and this continued price appreciation from 2005-2007, even in the face of declining demand which was supported through the loans with the loosest underwriting standards and disproportionately ended up in foreclosure down the road, spoke to the bubble mentality. In fact, as Case and Shiller wrote back in 2004:

During a housing price bubble, homebuyers think that a home that they would normally consider too expensive for them is now an acceptable purchase because they will be compensated by significant further price increases.

I want to focus on that last change, from 2011 to 2012, as this is the most interesting and most important aspect in the housing market today. We are seeing an increase in housing demand (red line increasing). Household formation has picked up, investors are looking for deals and to buy in bulk for rental properties, and buyers want to get in at the bottom of the market. Additionally, we are also seeing declining inventory of homes as new construction, while growing briskly, is down considerably from anything approaching a normal level. The market has worked through the majority of existing supply and homeowners who would like to sell, also do not want to sell at the bottom of the market. Active residential listings in RMLS in early 2013 are down nearly 50% from two years ago. When you have increasing demand and decreasing supply, the natural market movement is for price increases.

This is the basic story of the housing market today. The good news (for sellers, underwater homeowners, the economy more generally) is the bottom of the housing market is clearly in the past. We’re seeing stronger than anticipated home price appreciation today given this somewhat clunky ebb and flow in supply and demand. However over the medium and longer term we expect the market to move back into an equilibrium of steady increases in both supply and demand, resulting in price appreciation more in-line with the rate of inflation. Until we reach that place, we can likely expect more of the same, so long as the economic expansion continues, leading to stronger housing demand. Additionally, as home values rise and more and more homeowners come above water on their mortgages, this will likely help increase supply. These developments are both very interesting and important to the economy overall, so keeping track of the market is high on the priority list as we begin to develop the next forecast.

The following post takes a look at Oregon Lottery sales from both a high level perspective in terms of consumer spending trends and from a somewhat longer perspective over the business cycle. What follows below is the baseline outlook our office has for both Lottery sales and available revenues to the state for budgeting purposes. This is the same outlook that was incorporated into our office’s latest quarterly economic and revenue forecast and also the same outlook I recently delivered to the rating agencies in explaining the forecast. To download the full report in PDF format, please click here: LotteryOutlook_0313

Introduction

There is no perfect comparison for Oregon’s unique video lottery system. As a result, our office looks at a number of different spending and gaming related benchmarks for comparative purposes. Useful indicators include: employment and earnings among Oregon’s hospitality firms, sales tax collections in neighboring Washington, gaming expenditures nationwide, and the performance of video machines at tribal casinos and in Nevada.

Comparisons with other gaming systems are not perfect. National data incorporates many types of gaming activities across regions. Las Vegas relies upon large casinos and out-of-state tourism. Oregon relies more upon locals and spending in the bars and restaurants across the state. However, even with these underlying differences, broader consumer spending trends have become clear in recent years.

In the aftermath of the Great Recession, there has been a resetting of consumer spending on gambling and gaming activities. This is true at the national level based on the Bureau of Labor Statistics’ consumer expenditure data. This is true in Nevada, based on slot machine revenues from Las Vegas and elsewhere in the state. And this is also true in Oregon, based on sales at the 12,000 video lottery terminals throughout the state.

Great Recession Losses

The first chart gives a high level look at the recession’s impact on gaming revenues in Nevada and Oregon. The chart shows the percentage change in spending from the pre-recession, or business cycle peak levels. Both Nevada slot machine revenues and Oregon video lottery revenue fell by 20-25% during the financial crisis. Since hitting bottom in late 2009, Oregon’s video lottery revenues have begun to slowly climb back, and are up 8.6% from the depths of the recession. Clark County, Nevada (Las Vegas) slot machine revenues bottomed in 2010 and likewise have begun to increase again, however are up just 4.2% through the end of 2012. Slot machine revenues elsewhere in Nevada have yet to fully turn the corner and remain about 25 percent below their 2007 highs.

This overall pattern of large recessionary losses and a slow climb out was not wholly unexpected, particularly given that a slow recovery has been built into most underlying economic outlooks for the past few years. What was somewhat unexpected was the overall severity of the gaming declines, and the even more protracted recovery in revenues relative to underlying economic variables, such as income growth and of the broader “entertainment dollar.”

The Entertainment Dollar

The following graph illustrates U.S. consumer spending on entertainment as a share of income, based on data from the Consumer Expenditure Survey. Spending on entertainment in recent years exhibits a pro-cyclical pattern with this discretionary spending increasing more than income during the good times of the housing bubble and falling further during the recession. In 2011, broader spending on entertainment stabilized across consumer units at the national level. This broader stabilization of the entertainment dollar can be seen in other items such as sporting event attendance, increased tourism and eating out. Within the Consumer Expenditure Survey data this stabilization across the nation is also seen in both Alcoholic Beverages and Food Away From Home. Even the number of visitors to Las Vegas has reached pre-recession levels (HT: Calculated Risk), although visitors are not spending as much as they used to. The number of tourists is back, even if gaming revenues are not.

Without a sales tax in Oregon, we cannot track spending patterns in real time as well as other states. However, employment at leisure and hospitality firms has regained over 70 percent of recessionary job losses, and employment in video lottery establishments is all the way back. For comparison purposes, according to sales tax data from our neighbors to the North (Washington), sales tax collections from leisure and hospitality firms are essentially back to pre-recession levels on an inflation adjusted basis, and certainly from a nominal basis.

These measures show that consumers are spending more money on broader entertainment items, both in total and also as a share of their overall consumption. Where we have yet to see stabilization is the share of consumer spending on gaming activities, relative to income or overall consumption. Spending on gaming itself has increased – see the first graph – however these gains are below the overall gains seen in income growth or consumption.

The Gaming Dollar

Oregon is, again, somewhat unique as line games were introduced in 2005 and therefore the run-up in spending as a share of income was more pronounced due to this new product being deployed across the state. For those unfamiliar with the nomenclature, line games are, for all intents and purposes, effectively slot machines as reels spin and a player wins by reaching the proper combination of images on the reels. The introduction of these games into Oregon’s existing video lottery network substantially increased the level of sales across the state as this marked a new product in the marketplace. There was also a certain novelty aspect to these machines as many states do not have line games. Overall, these new video lottery terminals resulted in a large increase in the share of income devoted to these activities.

Oregon Video Lottery Outlook

At this point, our outlook is for sales to continue to improve as the overall economy continues to improve. Consumers will spend more money on both the broader entertainment category and on gaming activities as employment continues to increase and income growth follows. Also, as population growth continues (adding more individuals and more income to the state), our expectations for lottery sales improve. However, growth in gaming will not keep up with the rest of the economy. The gain in video lottery sales is expected to be just a bit under the total personal income growth rate in the state, and thus the share of income spent on video lottery will continue to erode somewhat.

So far in recovery the economy has been adding jobs back slowly and wage and salary gains have been weaker than previous expansions. Income gains, particularly in percentage change terms, have also been weighted toward investment and asset-based types of incomes due largely to the stock market rebound in recent years. Investment income has a smaller impact on lottery sales than does housing wealth or wages.  Also, the Oregon Lottery is only one provider of games in an increasingly competitive marketplace.

However, given the strong sales seen during the housing boom, there is upside risk to the sales outlook. To the extent that employment and wages grow more quickly than the baseline economic forecast, a pickup in the share of income spent on video lottery is not out of the question. As seen in the graph above, during the height of the past expansion the share of income spent on video lottery was much higher. This illustrates that there is certainly capacity for video lottery sales to grow both in terms of the level of sales and in terms of the share of consumers’ budgets.

We were not and are not expecting this share to rebound strongly, however we are expecting it to stabilize as the economic expansion strengthens in the coming years.

Oregon Overall Lottery Outlook

The overall lottery outlook for the state largely follows this same general pattern as transfers from video lottery account for about 85 percent of available resources. Two further important items to highlight are the impact of the federal payroll tax cut expiration and the video lottery terminal replacement program that Lottery is undertaking over FY2014-17. First, the payroll tax cut expiration is expected to shave just over 1% off of growth in 2013 which amounts to about $7.7 million in sales or $5.0 million transfers. (Although, as shown previously, recent sales have picked up in late February and early March) Second, over the next two biennia Lottery will be replacing the 12,000 existing terminals throughout the state, some of which will be nine years old when replaced. Due to advancements in technology, the current machines are becoming obsolete in the marketplace. This replacement plan is expected to cost approximately $215 million over four years, of which Lottery will self-fund $85 million. The remaining $130 million will be deducted from Lottery earnings prior to being transferred for general revenue purposes. The biennial impact of the replacement plan is $71.2 million in 2013-15, or approximately 7.2% of available revenue to be transferred, and $59.2 million in 2015-17, or 5.5%. The projected available resources, after accounting for the replacement plan, can be seen in the graph below.

Again, to download the full report in PDF click here: LotteryOutlook_0313

On Tuesday morning I gave a presentation to the Institute for Continued Learning at Willamette University. I promised that I would post the slides for their use, however I also figured now was a good time to share the slides more generally for those of you who are interested. The slides are available in PDF format below and run about 45 slides in length. They cover recent economic performance, Oregon’s short run and long run outlooks and student debt, as discussed previously. If you are a regular reader of the blog, you will likely recognize the majority of the facts and figures, but they are all now in one place. Plus there are a few new ones thrown in more good measure. If you have any questions or would like to discuss this further, please contact me.

To download the slides follow this link: Willamette 031913

This morning the Oregon Employment Department released the revised January employment data and preliminary February estimates. Overall these are very good numbers and are certainly encouraging. Although the unemployment rate held steady at 8.4 percent (US in February was 7.7%), the gains in the establishment survey are stronger than we have seen in recent years and even stronger than average gains during the past two expansions in Oregon (more on that in a minute).

I would like to back up here for a minute and discuss the outlook and recent economic performance. First, expectations were for a softer economy in late 2012 and into the first half of 2013. This was largely due to two factors: the waning manufacturing cycle and federal policy. As seen in the graph below, manufacturing took a step back during the summer, which would generally lead to slower growth moving forward. The graph shows new orders for nondefense capital goods, excluding aircraft, which basically looks at the pipeline for future manufacturing work after excluding two large (and sometimes volatile) groups that can distort the underlying trends: defense and aircraft. We have discussed the waning manufacturing cycle previously (with housing taking its place as a driver), and while the industry is operating at a high level, the actual growth rates have been lessening, however as seen in the last couple data points in the graph, these cuts have reversed themselves. Other manufacturing indicators such as the ISM Purchasing Managers Index which were hovering around the expansion/contraction breakeven point for all of the second half of 2012, have also picked back up in early 2013.

The other main item expected to weigh on growth was federal policy. Both the fiscal cliff negotiations which resulted in the expiration of the 2% payroll tax cut and the higher marginal tax rates on higher incomes and the potential impact of the sequester, which did come to pass.

So, how has the economy performed in recent months? Well, 2012q4 GDP came in very weak (current estimates show 0.1% gain) but early 2013 is surprising to the upside. Consumer spending and job growth in January and February are better than expected. Job growth over the two months has averaged about what the country has seen the past two years and retail sales were very strong in February.

The same is also true in Oregon. The job gains – 5,400 in January and 6,800 in February – are quite large from both a recent economic performance and from a more historical perspective. The two months combined, 12,200 jobs, are the strongest two months since November, 2005 and March, 2004 before that. If these gains do reflect the underlying economy – are not offset by losses next month, or revised away – it is certainly encouraging for both current conditions and the outlook, which does call for a pickup in growth in 2013 and 2014.

The fact that these gains in both employment and spending have occurred in the face of what many expected to be impactful tax changes, is somewhat surprising, but also good news. I am still skeptical that households have fully worked through the impact of the 2% payroll tax hike, and do expect it to impact spending this year, however so far the evidence that we have is surprising somewhat to the upside. We will certainly be keeping our eyes on these changes, particularly as we head into the May forecast which effectively sets the revenue for the 2013-15 BN.

Finally, even with the good news in recent months, the economy is not free and clear. The European situation continues to evolve over time and generally not to the good. Many European countries are back in recession (or never left) and even though the U.S. has managed growth during this, the dark scenario regarding Europe is, effectively, another financial crisis, from which the U.S. would not escape. Generally, the U.S. has not been pushed into recession due to slow growth elsewhere however when financial markets lock up and credit freezes, the U.S. economy would certainly be impacted.

On the domestic front the sequester has now gone into effect and is expected to weigh on growth moving forward. The first order impact on Oregon is expected to be somewhat muted given the relatively small federal presence in the state. As seen in the table below, based on work The Pew Center on the States did late last year, Oregon ranks low in most categories when comparing the size of the federal government to the state’s economy. Oregon ranks in the middle of the pack with regards to federal grants, however federal spending as a share of GDP, Oregon ranks 3rd to last at 2.1% compared to the national average at 5.3%. The one area where Oregon does rank higher than the average state is in terms of our workforce. Given the large federal presence in terms of land ownership in the state, higher levels of BLM and USFS workers – not to mention the BPA workers – are to be expected.

One more for a Friday afternoon. In today’s Oregonian there was the article about Oregon Labor Commissioner Brad Avakian testifying in Congress about the minimum wage. I cannot seem to find the print version of the article online as there was this portion where a member of Congress, if memory serves, mentioned that Oregon’s number of employees per restaurant was significantly lower than the U.S. average and has been declining. Either directly saying or implying that this was due to the rising minimum wage. I’m not here to debate the merits of our minimum wage system, but given research I have down done I was struck by that comment. It’s just a particular statistic that seems a bit odd to use. It really could be that Oregon is relatively under-served by the industry and possibly due to the wage. However it could just be that Oregon has a lot more smaller restaurants and less larger ones (possibly due to less nationwide chains? I’m not sure.) With a quick internet search I was able to find a similar remark on Oregon employment size in a Los Angeles Time article:

But testimony from the National Restaurant Assn. makes the case that Oregon’s minimum wage increase actually hurt the state’s economy. That’s because, it says, the minimum wage forced employers to lay off workers — the average number of workers in Oregon’s restaurants has steadily declined from 16.4 workers per establishment in 1996 to 13.8 workers in 2011.  Overall, U.S. restaurants employ an average of 16.9 workers per establishment, the same level as 1996.

Anyway, back in the fall I was doing cross state comparisons for the Leisure and Hospitality industry for the Citizens’ Initiative Review Commission, in particular for the casino measure. They had asked how Oregon compared to other states and what I found was Oregon actually has more Leisure and Hospitality than other western states (except for Nevada, think Las Vegas) so when this appeared in the paper today, I was somewhat puzzled. I went and downloaded the QCEW data for NAICS 722 – Food Services and Drinking Places which has employment, number of business establishment and wages. This includes both bars and restaurants however Oregon data for just restaurants in the QCEW is not available online except for 2011 and 2012 so to get a longer time series I’m using the broader category that includes bars.

First, as seen below, yes, Oregon does have less employees per establishment than the national average. This verifies the claim above and it is technically accurate. However does that mean Oregon’s Leisure and Hospitality industry is in decline (and due to the minimum wage)?

I would argue that Oregon is not under-served by the industry when looking not at the size of each individual establishment but the overall size of the industry relative to the population. The graphs below show the number of establishments and total employment on a per 1,000 population basis. This gives the relative size of the industry to the base population of potential customers. It is here that you can see Oregon actually has more than the national average.

This is the result I have in my head for the industry in Oregon. Possibly it has to do with our foodie culture, our brewpubs (although some of those are in beverage manufacturing, at least the brewery portion, if I’m not mistaken), and smaller, more abundant restaurants. And part of it may have to do with the minimum wage, in terms of scaling a business larger, although I’m not entirely sure why that may be the case but I wouldn’t rule it out categorically. Now what all of this means for actual business conditions and profitability, etc, I do not know. I just know that Oregon tends to have more and, yes, smaller eating and drinking establishments than the U.S. relative to our population. I am also not sure if this is good, bad, or indifferent, but simply arguing our smaller establishments are bad for the economy seems like the wrong way to go about talking about the subject.

For more information on Oregon’ minimum wage here is an Employment Department blog post from last month and they have more articles on their website.

Here is something we haven’t posted in a number of months, but is our staple recession comparison for Oregon. This graph shows job loss in the state for every post-WWII recession in terms of the percentage of jobs lost. When the line returns back to the 0% level, that indicates the state has regained all the lost jobs from the recession and the horizontal axis indicates how long it takes to return to the previous peak level of employment. Data goes through January 2013 and incorporates the latest benchmark revisions from the Employment Department (and BLS).

Just as the generally positive and encouraging economic news continues to roll in – particularly in the face of federal tax and fiscal policy – this graph really brings home the point of how much further we still have to go. Sure, we’re continually adding jobs and seeing wages grow and the unemployment rate tick down, albeit slowly, but today in Oregon, a little over 5 years since employment peaked at the end of 2007, the level of employment in the state remains a bit over 5 percentage points lower. It is now at the same relative level as seen during the early 1980s recession. That recession, at least for Oregon, was our real Great Recession as the state lost over 12 percent of its employment, the unemployment rate crested at 12.1 percent and the timber industry was hit hard, back when there were around 80,000 direct wood products manufacturing jobs, and the whole industry underwent a major restructuring.

It took a bit more than 7 years for the state to regain all of those lost jobs in the 1980s and if you were to overlay our office’s latest forecast on top of this graph, it would follow a similar path moving forward. We now expect that Oregon will not regain all of its lost jobs until the summer of 2015. Our overall outlook for a slow growth hasn’t really changed too much in recent years, although the exact return to peak data has fluctuated within about a year or so. (Relatively small differences in assumed growth rates do compound over time) For example, here is a post from early 2010 – actually this blog’s first real post – indicating our belief that the return to peak date would be more like 6 years, however by the end of 2010 our forecasts had called for even slower growth coming out and returning to peak at around 7 years.

So now why are we more optimistic moving forward? In a word, housing. Housing is clearly on the upswing and as Calculated Risk likes to point out all the time, it is the best leading indicator of the economy that we have. Beyond housing, as detailed more thoroughly the other day, household debt is coming more in-line with income, indicating household balance sheets are being repaired, which will support spending and economic growth moving forward. One significant headwind – or at least one item everyone thought would be a significant headwind – was the expiration of the 2% payroll tax cut.

Most workers are now taking home 2% less in each paycheck, and expectations were that it would squeeze household budgets further resulting in lower spending and/or lower savings rate. Well, so far at least, this doesn’t seem to be happening. The February retail sales report was released the other day and indicated continued increases in sales, above expectations. As that leaked Wal-Mart memo discussion mentioned, due to the last minute nature of federal tax policy changes at the start of the new year, the IRS delayed processing returns early in the year so they could get their systems updated and ready to go with the recently passed laws. The IRS (and Oregon for that matter) are now processing as usual and personal income tax refunds are being processed and sent back to taxpayers as expected. Both of these items do have an influence on personal income and therefore consumer spending, at least in terms of the timing of spending.

Oregon obviously does not have a sales tax, so therefore we cannot track local spending in real time like other states can, however we do have video lottery which in a sense acts in a similar manner. Video lottery is a form of discretionary spending and recent sales data can be seen below. I will have a full report and post about the Lottery outlook next week however given both the payroll tax expiration and the stronger February retail sales, I thought this graph nicely showed both items in Oregon.

Following the tax cut expiration, there was a clear decline in discretionary spending in the state, however as the tax returns are now being processed as usual, the refunds – which come early in the tax season, the big payments come late in the tax season – are supporting income and consumer spending and we do see an uptick in video lottery sales on a year-over-year basis. Our latest Lottery forecast does have a negative quarter built in for this Jan-Mar period with slow growth through the remainder of 2013, so this pattern was not wholly unexpected. I think most economists have been surprised by the strength of consumer spending in recent months in the face of the 2% expiration. I do not believe that households have fully worked through these issues, however recent sales data nationally, in neighboring western states and in this lottery data is encouraging from a big picture level. Also the impact of rising housing wealth does have a positive impact on consumers, their balance sheets and spending patterns, so with home values on the clear upswing this certainly helps today.

Is it possible that the majority of households have finally got their balance sheets back in order and are willing and able to spend even in the face of what was thought to be a sizable headwind? I don’t know but it is beginning to appear that we’re at least close. However, none of these data include the impacts of the federal sequestration which are only now beginning to be implemented. That impact is another item/issue to watch in the coming months to gauge the strength of the economy.

Following up on yesterday’s post on debt more generally, the following dives into student loan debt specifically. This is the only type of household debt that continued to grow throughout the Great Recession and now makes up the 2nd highest level of household debt behind only mortgages. If you have read our latest quarterly forecast publication, then you have already seen the following as this is excerpted and edited based on that. I would like to given a big tip of the hat to the Wells Fargo Economics Group for their recent report on student loan debt which served as a primer for this work. Their group provides very valuable commentary and information on economic issues and subscription is free, if you are interested. As for this post, I have also included a set of slides in PDF: StudentLoanDebt.

One subject that has generated concern in recent years is the high level of student debt that individuals are incurring today. Not only are students taking on larger debt burdens as the cost of school continues to increase at a much faster rate than overall inflation, but the number of students nationwide and in Oregon is very large, in part due to the lack of current job opportunities. The amount of student debt outstanding is approximately $1 trillion and accounts for 8.5 percent of all consumer debt in the U.S. (mortgages, autos, credit cards, etc). According to the College Board, 57 percent of bachelor degree recipients at public schools in the 2010-11 school year borrowed money to pay for school and those who borrowed graduated with an average of $23,800 in debt. These figures are up from a decade ago when 52% of graduates borrowed, with an average of $20,100 inflation adjusted dollars in the 2000-01 school year. In terms of enrollment, students on OUS campuses increased 22 percent from 2007 to 2011 and enrollment at Oregon’s community colleges increased even further, nearly 37 percent, over the same period. These gains reflect a general upward trend in enrollment, which spikes higher during tough economic times as job opportunities are fewer and further between.

With so much debt being incurred, the question some ask is why do individuals choose this path? The simple answer is, to paraphrase a recent conversation by the Governor’s Council of Economic Advisors, the only thing worse than taking on student loan debt is choosing not to go to college. The economic benefits of higher education can be seen below as the unemployment rate decreases for individuals with more schooling and earnings are also higher. This, in short, is why so many individuals incur student loan debt: to improve their economic situation.

There are a number of potential problems along this path, however. One, a student may not graduate, leaving him or her with few new skills to market and a much higher debt burden. Second, even with a degree, employment opportunities can be hard to come by. In fact, the unemployment rate for recent college graduates (bachelor’s or higher) has essentially mirrored that of the overall population in recent decades. While it is certainly true that college educated individuals have a lower overall unemployment rate, it takes time for new graduates to find work. Third, not all degrees or training a student obtains in post-secondary schooling are created equal in terms of job opportunities and potential wages earned in the future. Incorporating potential future earnings into a cost-benefit type analysis of incurring student debt may be beneficial for some students.

To the degree that there may be a potential asset bubble of bad student loan debt, as some have argued, it is likely to be found in the three areas mentioned above. All of these lead to a situation where the degree and/or training program(s) one attends does not lead to an economically better situation, or at least not good enough to reasonably finance the debt incurred. The theoretical reason educated individuals earn more money is that they have better skills and are more productive workers. In economic speak, students increase their human capital through schooling to raise their marginal productivity in the workplace (or equivalently signal their productivity to employers by succeeding at school), which in turn yields higher income and provides a return on the initial investment. To the extent that this is true, which the unemployment rate and earnings data suggest, the benefits outweigh the costs for higher education.

In addition to the individual returns to education, educational attainment also brings with it significant societal benefits for communities. As a result, a strong economic case can be made that public investment in education can enhance efficiency by better aligning private incentives with those of society. That said, if market incentives are altered incorrectly, the potential for inefficiencies such as asset bubbles arise. Poorly targeted or oversized educational subsidies could lead to bad educational choices.

The bottom panel above shows the default rate for Oregon students who have federal student loans that came due in 2010. These rates are our office’s calculations with categories based on the underlying data from the U.S. Department of Education, Federal Financial Aid office. While the exact breakdown by type of loan is unavailable, Stafford and Plus loans comprise the bulk of the sample. Overall, the pattern is generally what one would probably expect, with exceptions at both the top and bottom ends of the spectrum. Students from 4 year universities (both public and private) on average have a lower default rate, just as they have a lower unemployment rate and higher earning potential. Next are the majority of career schools, which generally offer training programs and/or associate degrees. Examples of Oregon schools in this category are Everest College, Sumner College, Warner Pacific College, etc. Schools with above-average default rates include culinary, art, and natural medicine institutions. Cosmetology, hair and beauty schools have an even slightly higher default rate. The pattern of default for these four major categories of schools is most likely the combination of degrees obtained and future earning potential.

Finally, both religious schools and community colleges may be somewhat of a surprise. Religious schools (bible colleges, seminaries, etc) have the lowest default rate of any category although the reason is not entirely clear. Some types of financial aid do have restrictions on the majors a student may choose, in particular some religion-related majors are excluded, so the lower default rate may be a reflection that a higher percentage of these school’s students do not qualify for federal aid. Or the fact that students from these schools simply have a lower default rate which implies that their job placement – at least their financial situation – is better than most other types of higher education institutions.

On the opposite end of the spectrum are Oregon’s community colleges. These schools have the highest default rate out of any of the categories. This is problematic since traditional four year colleges are not a good fit for many students. As a result, no significant increase in educational attainment can be achieved without a large role being played by community colleges and professional programs. State and local governments have invested significant resources into these community colleges and workforce training programs are run through them in many locations with the hope that these actions result in a stronger workforce and better jobs.

Returns on education are a function of more than just the quality of the program and ability of students. The availability of job opportunities is also a factor. This factor is clearly evident in 2010 default rates which are elevated relative to rates seen in better economic times. Differences in regional job opportunities may also help to explain the elevated default rates among Oregon’s community colleges, many of which are located in rural regions that are struggling economically.  These colleges generally enroll students from their local communities and are more closely tied to regional economies than are large, four year institutions. Regions around the state outside of the Portland metropolitan statistical area have not shared equally in the economic expansion so far. While these regions have stopped losing jobs, they have yet to begin adding jobs on a consistent basis, making loan repayment more challenging. The scatter plot below shows Oregon community college’s default rate on the horizontal axis and the primary county’s unemployment rate (or counties) on the vertical axis.

There is a positive correlation between these data series, indicating that the higher the local unemployment rate, the higher the default rate on student loans. Every county in the sample with an above-average unemployment rate is home to a college that exhibits an above-average student loan default rate.  To the extent that default rates reflect underlying, local economic conditions and not something fundamentally different, is encouraging. As local economies around the state improve, so too should student loan repayments. However, it also speaks to the need to attract firms that hire skilled workers alongside any efforts to improve the skill set of the local workforce.

Overall, going to school generally improves the lifetime earnings and wellbeing of students. Given this, the increase in student debt should be manageable moving forward, provided the economy continues to improve and provides sufficient job opportunities for all individuals to find work at reasonable wages. Should growth not pick up and provide more jobs, or the economy experiences another recession in the near term, the U.S. and Oregon are in for a rough time and bad student debt will become a larger financial issue.

Finally, the last item takes a quick look at regional educational attainment in Oregon. Oregon as a whole has a small advantage over the US in terms of bachelor’s degree recipients and even though Portland is considerably higher than the state average (and the U.S. average), relative to other major cities, Portland is just a bit higher as well.

Again, sides here in PDF if you are interested: StudentLoanDebt.

One item discussed on the blog last summer were the declining delinquency rates for debt across Oregon. That was important because as households increased their debt during the previous expansion, when the bottom fell out it became much more difficult to finance those larger debts. Declining delinquency rates imply improved household balance sheets and consumers are able to stay current on their debts, pay them down or off, or in the other case the debts may be written off by the creditor as the individual or household defaults. Improving household balance sheets are important for both the health of the economy and growth moving forward. The following post takes a quick look at the same debt data, available from the Federal Reserve Bank of New York, but examines the level of debt relative to income to get a sense of the size and how balance sheets are improving.

While the NY Fed has further information at the national level, in terms of the type of debt, at the local level the data available consists of automobile loans, credit cards and mortgages. This first graph shows the ratio of these debts to total personal income in both Oregon and the U.S. Over the period for which we have this data, debts were increasing faster than income growth through about 2007, hence the rising ratio. As the recession hit you can see stabilization in the ratio as income falls, but so too does the level of debt (combination of not taking on more debt, banks lending less, paying off more debt and outright defaults). In the first couple years of the expansion, this ratio has continued to decline (and fully expect 2012 data to show this trend continuing) as individuals and households are repairing their balances sheets. This includes the same items mentioned above – taking on less new debt, paying down existing debt or defaulting – but also with gains in personal income helping to support his balance sheet repair. It’s hard to tell over this time period what the appropriate or even the average level of debt to income “should” be but in terms of improving household finances, a declining ratio was certainly in order for most Americans, and we do know that the U.S. savings rate has increased during the crisis and has held up relative to the levels seen during the housing boom.

Oregon’s overall debt is somewhat higher, relative to income, than the U.S. for two reasons. One, yes, Oregon’s personal income is a bit lower than the U.S. average as discussed in the media and reports. For any given debt level, this fact alone would push the ratio higher for Oregon relative to the rest of the nation. Second, and this is the primary reason for the higher ratio, our mortgage debt is simply larger in Oregon. This could be due to higher home prices than average, larger loan balances (less paid off) or some combination of the two. Based on the latest American Community Survey (using the 2009-11 3 year estimate) it appears to be both. Oregon’s median home value is higher than the US (about $245,000 vs about $180,000) and the percentage of Oregon housing units with a mortgage (69.5%) is higher than the U.S. (67.2%).

The graph above looks at the state total, but how do individual counties compare? Below are the same debt to income ratios for 2007, when the ratios peaked, and 2011, which is the most current year we have data for. Some regional patterns do emerge based on this. The higher ratios are generally seen in Central and Southern Oregon and the Portland suburbs (of, if you prefer, the Portland MSA ex Multnomah). This, from my vantage point, speaks to the housing bubble. Bend and Medford experienced large housing booms (and busts), which coupled with strong in-migration, led to higher mortgage debt levels both because prices were higher but also because more recent purchases generally have higher debt balances as owners have not lived there long enough to pay down the debt. You can see this same dynamic playing out in other locations like Hood River, Polk and Yamhill counties that experienced stronger population growth. On the other end of the spectrum are many Eastern and Southern Coast counties. Home prices are generally lower in these areas and population growth during the past decade has also been lower.

Another aspect you might have been able to see in the graphs above is the fact that every single county has seen their debt to income ratio improve over these years. To better show this improvement, the following graphs the same data for 2007 and 2011 but sorts the counties from the largest improvement to the smallest improvement in the ratio. (The difference between the dark blue and light blue.) The larger the improvement in the debt to income ratio, the higher the rank.  It is encouraging that the largest improvements in these ratios have been in some of the counties where the ratio had been highest (again, Central and Southern Oregon and Portland outside of Multnomah.) As the underlying economies continue to improve (or in some cases, begin to improve), expectations are for these debt levels, at least relative to income, to continue to improve.

The smaller improvements are generally seen in the Eastern and Coastal counties which did not have as high of debt levels to begin with, so seeing smaller improvements is not as much of a concern. The two counties that stick out somewhat are Clatsop and Columbia with their relatively high debt levels but low improvement. This likely speaks to both housing costs and economic performance. Both counties have either higher home prices (Clatsop) or experienced larger population growth (Columbia) which work to increase mortgage debt levels. Both counties are also not experiencing as strong of economic growth in expansion as the statewide figures show. While Clatsop outperformed the state during the crisis the improvement in the past couple years has been slower in the county. Columbia suffered a worse recession than the state as a whole and its unemployment rate remains about 1 percentage points above the Oregon figure.

Next, we turn to comparing Portland with a few select western MSAs. These are sorted the same way as the graph above – by improvement in the debt to income ratio from 2007 to 2011. It is interesting to see that the bubbilicous regions are back down to around the same levels are seen in other major western MSAs. This is encouraging from a bigger picture, household balance sheet perspective as these region had the furthest to go for improvement and, at least based on this ratio, the majority of the work appears to be done.

Finally, the last graph shows the overall ratio for NW states. Washington’s debt to income ratio is a little bit higher than Oregon’s while Idaho’s is a little bit lower than Oregon’s. As mentioned above, this is likely due to overall housing costs and less about auto or credit card debt.

Overall, it does appear that both U.S. and Oregon households are making improvements in their balance sheets. Debts on automobiles, credit cards and mortgages are declining relative to income which bodes well for both households and for economic growth. Other types of debt may be increasing – in particular student loans, which is the subject of a subsequent post – but these main categories are becoming more in-line with the income that is used to finance them. As the economic expansion continues, expectations are that these debt levels to likewise improve.

Apologies for the radio silence. I’ve been busy both professionally and personally but have a number of good things lined up in the coming weeks. Up next are two posts on debt in Oregon. One will examine debt on auto loans, credit cards and mortgages, using the same NY Fed data seen on the blog before, however this looks at debt to income ratios. The second post will be on student loan debt and educational attainment. Both in Oregon and in the U.S. as a whole. In about a month I will revive the Job Polarization report from last year that still sits in draft format as 2012 occupational wage data is to be released by BLS at the end of March. I hope to have that report issued in either the first or second week of April. In the coming weeks we will start to see the bulk of income tax returns being submitted and processed and we will provide some updates along the way. Our forecast calls for a fairly big April due to stronger growth in 2012, particularly in investment type incomes and also the pull forward impact of higher federal tax rates. Finally what has kept me the most busy recently has been the Lottery outlook. I have a research report on Lottery and gaming trends that will likely be released next week or the following.

In the mean time here is a photo of me doing economic/finance related things in New York, where I was recently. On the left is the Charging Bull, an iconic symbol, while on the right is the plaque for Zuccotti Park the physical origins of Occupy Wall Street. It turns out that “park” in NYC doesn’t necessarily mean grass and trees, as it really is more of a plaza and I walked by the first time without realizing it. The last picture is in front of the Federal Reserve Bank of New York building, which is also down in the financial district, as you might imagine.

Lots of other good things being produced by our office that I will try to keep you posted on. Also as we head toward the May forecast, which effectively sets the revenue outlook for the 2013-15 budget and also the basis for the kicker calculation, lots of discussion on the economic outlook.