Thursday, March 2, 2017

Housing permits in Metro Denver up 18 percent in 2016

Housing permits in the Denver metro area increased 18 percent from 2015 to 2016, rising from 1,462 to 1,732 permits. As we can see in the first graph, this brings total permit activity back to levels similar to those found in 2004 and 2005, and well above what we saw during the recessionary period of 2008 and 2009:

As demand for both apartments and for-sale housing grew after 2009, we saw gradual increases in permit activity, and the permit total is now at a 14-year high. Permit totals have not been this high since 2002 when they reached 1,770. 

The composition of the permit total is very different today from what it was prior to 2007, however. As we see in the next graph, prior to 2007, the total number of single-family permits greatly outnumbered the number of multifamily permits. Since 2011, single-family and multifamily permits have been rather similar, with multifamily permits even exceeding single-family permits in 2012 and 2013. 

We can contrast this to the situation before 2007 when the very large numbers of new single-family permits from 1998 to 2005 were a factor in what became sizable declines in home prices from 2008 to 2010 that only made a quick turnaround on the back of massive federal monetary and fiscal stimulus in the wake of the 2008 financial crisis. Since then, the single-family market has remained well below its former peaks even though population growth continues as a sizable rate in Colorado. 

Much of this single-family activity has now been replaced by multifamily production. In fact, over the past five years, nearly half of new permits have been for multifamily housing, which is highly unusual:

Moreover, we see this show up in the homeownership numbers in Colorado. due in part to the boom in apartment building that has taken place since 2012, homeownership has gone into decline over the past decade. 

While homebuilders were quickly adding new supply, we saw the homeownership rate sit as historic highs above 70 percent. . 

Now, the housing focus has shifted more in the direction of multifamily, but overall permitting has  rebounded to 80 percent of where it was at the 2001 peak. Nevertheless, continuing growth in both home prices and rents suggests that there is still room for more housing production. 

Foreclosures in Colorado's metro counties fell 6.9 percent from 2015 to 2016.

Foreclosure activity continued to fall in 2016 in Colorado's 12 metropolitan counties. In Colorado's 11 most populous counties plus Broomfield County,  new foreclosure filings were down 6.9 percent while foreclosure sales at auction — the point at which the foreclosing property is auctioned off to the bank or other buyer — were down 26.5 percent. The first chart shows all counties measured, and the year over year change from 2015 to 2016:

Except for Mesa and Denver counties, foreclosure filings fell in all counties. And, when it came to foreclosure sales at auction — the final and most serious phase of foreclosure — only Mesa county reported an increase from 2015 to 2016.

Annual foreclosure totals in these counties have fallen off significantly in recent years:

If we look at monthly totals, we see a similar trend — with foreclosure activity falling to some of the lowest levels we've seen since I began tracking monthly numbers in 2008. The first graph shows just how far totals have fallen since they peaked in 2008 and 2009. In December, we can see that new foreclosure filings ticked up slightly, while sales totals remained near all-time lows:

For all counties combined, filings were down in December 2016 by 17.4 percent, year over year, while sales at auction were down 29.2 percent. 

Here are filings: 

 And here are sales totals:

Only Mesa and Broomfield Counties reported increases in filings (the Broomfield numbers were tiny overall), while only Adams, Boulder, and Pueblo counties reported increases in sales. In general however, it's safe to say that foreclosures in the state continue to be at at least ten-year lows.  

As a final note, we can see which counties have the most or fewest foreclosures in relation to the number of households overall. The following chart shows the foreclosure rate given as the number of foreclosure sales adjusted for the number of households in each county (based on 2015 household totals): 

By this measure — in which a higher number indicates fewer foreclosures per person — we see that with zero foreclosure sales, Broomfield has the lowest rate, but Douglas County also shows a very foreclosure rate compared to other counties, with more than 57,000 households per foreclosure sale. Pueblo and Mesa counties — as has been the case for several years — show some of the highest foreclosure rates with Mesa showing just over 2,300 households per foreclosure sale. 

Generally, higher income areas with more demand for real estate tend to show lower foreclosure rates while areas with lower incomes tend to have higher foreclosure rates. 

This data is just the latest confirmation that demand for real estate in Colorado continues to be strong. In an environment such as we have today, even a household that goes into foreclosure can usually find a buyer before the property enters the foreclosure process. Thus, high demand for homes leads to few foreclosures, even as some households become delinquent on their loans.  

Note on the data: all foreclosure numbers are collected from the public trustee in each county measured.

Rent growth in Metro Denver falls to 4-year low in 4th quarter of 2016

According to the latest data from the Apartment Association of Metro Denver, the average rent in Metro Denver during the fourth quarter of 2016 was $1,347, which was down from the third quarter average rent of $1,368.

The fourth quarter's average rent was up from the fourth quarter of 2015 with the average rent rising 4.2 percent from $1,292.

With a year-over-year increase of 4.2 percent, however, average rent came in at the smallest year-over-year increase since the first quarter of 2013, when the average rent grew 4.1 percent:

Rent growth has been falling for the past six quarters after reaching a record high of 13.2 percent during the second quarter of 2015.

The average rent has also fallen, quarter over quarter, for the past two quarters. After hitting $1,371 in the second quarter of this year, the average rent fell to $1,368 during the third quarter and then to $1,347 during the fourth quarter:

If we adjust for inflation, we see a similar trend, although we can note that in real terms, the average rent fell between 2001 and about 2011:

Flattening rents are likely a result of substantial increases in new housing construction that took place in metro Denver since 2012. In October 2016, for example, new housing permits reached the highest level recorded since 2001. 

Not surprisingly, we also find that vacancy rates have increased as well. During the fourth quarter, the metro Denver' vacancy rate rose to 6.2 percent, which was up from the 3rd quarter rate of 5.1 percent. 2016's fourth quarter rate was down, however, from the vacancy one year earlier. The vacancy rate had surged to 6.8 percent during the fourth quarter of 2015, but fell again during the first quarter of 2016. Nevertheless, 2016's fourth-quarter rate of 6.1 percent is the second-highest vacancy rate recorded since 2010. 
All of this points toward some stabilization in the market, although there is still little to indicate that we're entering a bearish market in multifamily housing, barring a sizable economic shock to employment or incomes.

Home-loan payoffs in Colorado increase in 2016 after fourth-quarter surge

In Colorado, a release of a deed of trust occurs when a real estate loan is paid off whether through refinance, sale of property, or because the owner has made the final payment on the loan. Increases in release activity occur as refinance and home-sale activity increases, and rising release totals generally indicate increases in the demand for home loans and real estate.

Every quarter, I monitor release activity in 21 counties based on population size and to ensure that as many regions of the state as possible are represented. More than 90 percent of all occupied households in Colorado are within the twenty-one counties chosen:

The number of home loans paid off in Colorado was up 14.4 percent from the fourth quarter of 2015 to the fourth quarter of 2016. Comparing the full year of 2016 to 2015, the total was up 2.8 percent. 

Looking at all counties' measures, we find that during the fourth quarter of 2016, most areas reported increases in releases of deeds of trust: 

Year over year, the largest increase was in La Plata County and the largest decline was in Alamosa County.

Taking the year as a whole, the changes were less dramatic: 

Statewide, releases were up 2.8 percent for the year overall. Again, La Plata County showed the largest increase, but Eagle County, in this case, shows the largest decline. 

The fourth quarter of 2016 showed a sizable surge in release activity as can be seen in the first graph: 

For the fourth quarter of 2016, there were 93,147 releases, which makes the quarter the second-highest in releases since I began tracking the quarterly totals in 2008. During the fourth quarter of 2015, there were 81,398 releases. Over the past two years, there has been a sizable trend upward in release activity. 

Although quarterly data only goes back to 2008, I do have annual data going back to 2000. Looking at annual data, we find that 2016 was not an especially remarkable year, in spite of the fourth quarter's surge: 

Releases did hit a 3-year high in 2016, but this was comparable to what we saw in 2012 and 2013. Totals remain well below where they were in the run-up to the housing bubble in 2002, 2003, and 2004. 

So where is release activity most concentrated? In raw numbers, we naturally expect the most activity in the most populous counties. However, when we adjust for the total number of households in each county, we find: 

In this case, a small number means more release activity given the population size. So, the county with the most release activity per household is Summit County with only 6 households per release. The least-active county is Alamosa County with 48 households per release. 

If we had to identify factors that lead to more release activity, we could likely point to median income and employment trends as factor, as well as overall demand for real estate in that area. There isn't a flawless correlation here, but counties like Douglas, Weld, and Jefferson continue to see fairly solid employment and income numbers while some counties at the bottom of the list — Pueblo and Alamosa Counties, for instance — tend to have lower incomes and/or less job growth. 

And finally, it's always helpful to compare release activity against trends in mortgage rates. Since refinance activity is a large factor in release activity, we find that, historically, falling interest rates have tended to spur more release activity.

This appears to still be the case, and we can see that in the quarters following declines in the mortgage rate, we see increasing release activity:

There is a lag here, so the 4th Q's surge represents the result of the overall decline in mortgage rates that occurred through most of 2016. Note also how in the quarters immediately following the increases in the mortgage rate in 2013, releases fell to a multi-year low. 

In late 2016, mortgage rates began to climb and remain above four percent during the first quarter of 2017. This is likely to lead to a leveling off in release activity, although this may not be apparent until after the first quarter this year. Moreover, a continuation of strong employment trends may mitigate interest-rate-based downward pressure on home sales and refi activity — and prevent sizable declines in release activity as well. 

Note on sources: all release activity is collected from the public trustee in each state listed. 

Wednesday, March 1, 2017

Snowpack is above average, but eastern Colorado experiencing drought conditions

You'd have to be no paying attention at all to not have noticed that California has experienced an amazing turnaround in its drought conditions.

According to the national drought monitor, as recently as November 2015, the California drought map looked like this:

Now, after months of well-above-average precipitation, the California drought map looks like this:

Colorado's turnaround was not so abrupt, and went from severe drought in 2013, then to virtually no drought at all in 2016. Here's 2013:

And here's May 2016:

Over the past several months, however, the Front Range has begun to see dry conditions again. Here's the current week:

This will lead to concerns over wildfire conditions, although fortunately, precipitation in the mountains has been high enough that water supply-concerns do not appear to be substantial at this time. If we look at recent data on snowpack, we find that while Colorado started out very slowly this season, it is now well above where we've seen it in recent years.  The black line is 2017 and the red line is the average:

And, even the second-most pessimistic forecast shows total snowpack coming at at slightly above the median measure for the year:

The water supply is increasingly important in terms of residential use, although agriculture uses over 90 percent of the state's water while producing a mere one percent of the state's production and less than two percent of the state's jobs. 

Sunday, January 8, 2017

Vacancy rate in metro Denver falls to 1-year low of 5.1 percent

In yesterday's post, we looked at apartment rent growth in the metro Denver area, and we noted that rent growth, even though it's been lessening in recent quarters, has been quite strong over the past two years.

Not surprisingly then, we find that apartment vacancies have been relatively low. During the third quarter, metro Denver's vacancy rate was 5.1 percent. That's down from the second quarter rate of 5.4 percent in the second quarter, but it's up slightly from last year's third quarter when the vacancy rate was 5.0 percent:

The metro-wide vacancy rate is at a one-year low, but remains higher than during most of what we saw during 2012 and 2013 when population growth was strong, and new housing units were not being produced in the larger numbers that are being produced now. 

Traditionally, a vacancy rate of five percent has been called an "equilibrium" rate site it tends to imply neither a tight market, nor a market with excess supply. As an example of a tight market, we might look to the vacancy rates during the 1990s when vacancy rates frequently fell to levels near four percent (or even below). 

With a vacancy rate of five percent, we're likely seeing a reflection of the overall economy in which job growth is solid, and so is new construction in housing. The new construction means that, even with new job growth, demand is not driving vacancy rates down to historic lows as they did in the 90s. 

Indeed, one of the larger drivers of the vacancy rate has long been employment.  In the second graph we can see how the vacancy rate frequently moved along with the unemployment rate over the past 25 years. when the unemployment rate is low, vacancies tend to be low. When unemployment is high, the vacancy rate tends to be high.  In the wake of the 2008 financial crisis and recession however, we saw that the market tightened even as the unemployment rate increased. This was likely due partially to the accompanying foreclosure crisis which drove many former homeowners into rental units. It was also likely due to the crash in housing-unit production that followed the 2008 financial crisis. Today, however, it is likely that we will see the two variables track more closely as housing-production levels return to more normal levels in the absence of large numbers of foreclosures. 

Saturday, January 7, 2017

Metro Denver rent growth falls to 3-year low in third quarter

According to new data from the Metro Denver Apartment Association, the average rent in the metro Denver area grew 5.9 percent from the third quarter of 2015 to the third quarter of 2016. That's the smallest year-over-year gain in the average rent since the second quarter of 2013.
The year-over-year growth rate has also been falling for the past five quarters. Year-over-year rent growth had reached an all-time high of 13.2 percent during the second quarter of 2015. Since then, the growth rate has gotten smaller each quarter, and as of the third quarter is now at a level not seen since the second quarter of 2013 when the year-over-year growth rate was 4.3 percent:

In the first graph, we see that rent growth had reached all-time highs last year, but things have since tapered off (the graph shows percentages):

At 5.9 percent, the growth rate for this year's third quarter is at a level that would have been considered very run-of-the-mill throughout the 1990s. Considering that new housing permits reached a 15-year high in metro Denver during October 2016, it may be that the period of surging rents in metro Denver is over for now. 

In the second graph, we see rent growth in non-inflation adjusted dollars over the past 35 years (the graph shows dollars): 
Although rent increases have been getting smaller, they do remain in positive territory, so we do see that the overall trend has certainly been upward. 

During the third quarter of 2016, the average rent in metro Denver was 1,358 dollars, which was up 5.9 percent from the third quarter of last year (when the average rent was $1,291), and down 0.2 percent from the second quarter (with an average rent of $1,371). It is very unusual, however, for the average rent to fall from the second quarter to the third quarter, further suggesting that the apartment market really is softening. 

Inflation-Adjusted Rents 

It is always a good idea to look at rental rates with an eye on the inflation-rate as well. In recent years, rent growth has generally outpaced the Consumer Price Index (CPI), and we see that is still the case as of the third quarter. (The CPI was up only 1.1 percent, year-over-year for the third quarter). However, we can see that rent growth, when we take inflation into account is less than it is in nominal terms. Adjusted for inflation, the average rent in metro Denver looks like this: 

As the CPI has been growing very slowly in recent years, we find little difference in the recent trend between the nominal numbers and the inflation-adjusted numbers. What's different in this case is the fact that rents can go down for years at a time  in real terms — as happened from 2001 to 2012. That has not the been the case since 2012, however, and even when adjusted for inflation, we find that rent growth has been substantial in recent years. However, as is the case with the nominal rents, the rate of year-over-year growth in rents has been falling for the past five quarters. 

The takeaway here is that rent growth continues to be positive, but the trend suggests we're already moving beyond last year's record-level rent growth and may very well be moving into a period of more moderate rent growth, especially in light of increased activity in single-family and multi-family construction. 

Home Loan Payoffs Down 2.6 Percent in Third Quarter of 2016

The number of mortgage loans paid off in Colorado was down 2.6 percent during 2016’s third quarter compared to the same period of last year. Mortgage payoff rose 21.6 percent from the second quarter to the third quarter of this year. 

Public trustees in Colorado released a total of 87,192 deeds of trust during the third quarter of 2016, indicating a sizable increase in the amount of home purchase and refinance activity since the first and second quarters of this year. Typically, a release of a deed of trust occurs when a real estate loan is paid off whether through refinance, sale of property or because the owner has made the final payment on the loan. Release activity rises as refinance and home-sale activity increases. 

In the first chart, we see release activity in the third quarter compared to the third quarter of last year: 

In this case, we see that Summit County and La Plata County reported the most release activity over the period, thus suggesting more sale and refi activity in those areas. The largest decreases, on the other hand, were found in Boulder County and Eagle county over the period suggesting less sales and refi activity. Overall, release activity fell in 11 counties while increasing  in 10 counties. 

The second chart shows release activity relative to the overall number of households in each county. A lower number indicates more release activity per household. 

In this case, there were only 7 households per release of deed of trust in Summit County, making it the most active county. The least active county was Pueblo County:

The third chart shows total releases for the first three quarters of this year compared to the same period last year. In this case, the largest increase was found in Alamosa County, and the largest decline was found in Eagle County. In the counties surveyed, activity fell 1.3 percent from 2015 to 2016. 

In this case, it appears some of the most expensive counties — in terms of real estate — have been experiencing some of the largest declines in activity while the largest increases are seen in the lower-cost counties such as Pueblo and Mesa. There are exceptions, however, as in the case of La Plata county, which is a relatively expensive county that is nonetheless seeing continued increases in release activity. 

Being largely driven by refi activity, release activity is sensitive to changes in mortgage rates.  In the next graph, we can see that following periods of decline in mortgage rates, release activity tends to rise. Release activity tends to fall in periods following increases in the mortgage rate: 

For example, note how release activity fell significantly during late 2013 and early 2014 following an uptick in mortgage rates during 2012 and 2013. Similarly, an increase in mortgage rates in 2015 likely contributed to a slight drop-off in release activity in early 2016. In 2016, however, mortgage rates have fallen, which helps explain the uptick in release activity during the third quarter of this year.  

Other factors are important too, however, and these include overall demand for housing, employment, and population growth. In Colorado, even in the face of increases in the mortgage rate, we often continue to see sustained real-estate transaction activity when jobs creation remains solid and population totals are steady or increasing. Overall, 2016 has seen little change from 2015 in terms of statewide release activity.

Barring a significant chance in economic conditions, this situation is likely to continue. 

(Totals for releases of deeds of trust are collected quarterly. This report tracks releases of deeds of trust as reported by public trustees in Colorado. The report includes twenty-one counties which are chosen based on population size and to ensure that as many regions of the state as possible are represented. More than 90 percent of all occupied households in Colorado are within the twenty-one counties chosen.)

Friday, December 30, 2016

November Foreclosure Sales in Colorado Metro Areas Continue to Fall to New Lows

During November 2016, there were 147 foreclosure sales in Colorado's metro areas, which means foreclosure sales totals are near the lowest levels ever recorded since the report was first initiated in 2007.

Foreclosure activity in general throughout 2016 has shown numerous declines throughout the year as home prices increase, job losses remain muted, and demand for real estate remains strong.

Foreclosure sales are the final stage of the foreclosure process when foreclosing properties are sold to a third party or become bank-owned.

In November, foreclosure sales were down 29 percent compared to November 2015 in the metro areas, falling from 207 during November 2015 to 147 during November 2016.

For the first eleven months of this year combined, there was a total of 2,138 foreclosure sales in Colorado's metro areas, which was down 26.3 percent compared to the same period of 2015, when there were 2,899 foreclosure sales.

Foreclosure filings, which are the first step in the foreclosure process, were up slightly in November 2016 compared to November of last year. There were 469 foreclosure filings in Colorado's metro areas, which is an increase of 0.9 percent over November 2015's total of 465. For the first eleven months of this year combined, foreclosure filings were down 5.8 percent, falling from 6,050 to 5,697 from 2016 to 2016.

2016's falling foreclosure totals continue a multi-year trend in foreclosures which have been declining since 2009 and are now at some of the lowest levels ever recorded.

The Colorado counties measured for this metro-area report are Adams, Arapahoe, Boulder, Broomfield, Denver, Douglas, El Paso, Jefferson, Larimer, Mesa, Pueblo, and Weld counties.

The report can be found below:

Wednesday, December 21, 2016

Colorado Metro Foreclosures Hit New Low in September 2016

During September 2016 in Colorado's metro Counties, there were 438 new foreclosure filings. That's the lowest total ever recorded since I began tracking monthly totals in 2007. September 2016's total was down 13.2 percent from September 2015 when new foreclosure filings totaled 505.

"Filings" refers to "notices of election and demand" which are the first step in the foreclosure process. Filings had peaked at 4,030  in December 2007, but September 2016's total was down 89 percent from that peak. 

Foreclosure "sales," which are the point in the foreclosure process where foreclosing properties are sold off at auction, are also down considerable from the 2008 peak. during September 2016, there were 158 foreclosure sales in the metro counties. That's down 38 percent from September 2015, and it's down 94 percent from the peak of 2,706 in January 2008. 

Due to high price inflation for homes — and the related phenomena of high demand, moderate new construction, and rising population — foreclosure totals continue to decline, and show few signs of increasing substantially barring a significant change in current economic conditions. 

The counties included in this analysis — which constitute around 85 percent of statewide activity — include Adams, Arapahoe, Boulder, Broomfield, Denver, Douglas, El Paso, Jefferson, Mesa, Pueblo, and Weld counties. This data is collected from each county's public trustee office.

Tuesday, October 4, 2016

Tax Return Data Shows Declining Average Tax Collections in Colorado

Most of the employment data used in economic trends analysis depends on survey data using samples. This data can be useful, but it's always helpful to have other sources to provide context to the data.

One way to analyze trends on an annual basis is to use tax return data. Tax return data, of course, is especially reliable because misrepresenting income on tax returns can bring stiff fines and even jail time. 

Specifically, there are at least two helpful data points here. One is the number of filers for each year. This tells us how many people (or joint filers) had enough income to necessitate filings a return. When the number of people with jobs declines, the number of filers will decline. What the job market is adding many new workers, the number of filers will tend to increase. 

Looking at the total number of filers putting in income tax returns, we see the following trend: 

Not surprisingly, as job growth surged at the end of the last economic boom, we saw total filers peak in 2008. After that, it took six years for total filers to return to where they were. Now, the total number of filers includes more than just full time workers. It can include people with other sources of income as well. Nevertheless, this does gives us some insight into how much job creation was taking place during this period. 

What we do find is that the last economic downtown was indeed rather severe, especially when we note that during the post-2002 downturn, filer totals returned to the former peak after only three years. 

Since 2014, filer totals has exceeded the peak.

As a second measure, we can also look at average tax collections. We find this by dividing gross tax collections (including individual income tax, employment taxes, and estate and trust income tax) by the number of income tax filers. This is just an estimate of total income taking place, of course, but using the same methodology over time, it does provide us with a trend. 

 I have also adjusted this for inflation (using the CPI). What we find here is that real real average income tax collections have been declining over time. 

Some of this may be due to people voluntarily retiring and accepting less income — and thus, a lower standard of living. But in many cases, it may also reflect declining real wages and a decline in the number of work hours. For example, we do know that nationwide, the number of workers who are involuntary part timers has been significantly higher during this cycle than during past cycles. 

In an earlier post, we also noted that real incomes in Colorado have been flat or declining, depending on the measure. We see a similar trend here to that found in median household income in Colorado (see original post):

Note: All data comes from annual "Internal Revenue Service Data Book." Calculation based on total  income tax collections found in Table 5. This is then divided by filer totals (individual income taxes) found in Table 3. 

Tuesday, September 27, 2016

Colorado's homicide rate increases, Pueblo reports highest rate

The FBI released new homicide data yesterday showing the US nationwide homicide rate increased to 4.9 per 100,000. That's up from 2014's rate of 4.4 per 100,000.

Nevertheless, the US homicide rate remains near a 50 year low, and 2015's rate is comparable to homicide rates found during the late 1950s when the US was in the midst of a historical lull in homicide rates.

In Colorado, we find a similar situation. 2010's homicide rate was the lowest recorded since 1960, with 2015's rate up from that. However, the overall trend has generally been flat over the past ten years:

In 2015, the Colorado homicide rate increased to 3.2 per 100,000, up from 2014's rate of 2.8 percent. However, 2015's homicide rate is what in no way unusual compared to the last decade of data. These totals are among some of the lowest we've seen in decades.

We see little to suggest that the legalization of recreational cannabis use has led to any sizable increase in homicides, or that gun control measures in recent years have had any impact that is measurable.

Colorado's homicide rate increased 14 percent from 2014 to 2015, compared to a ten percent increase for the nation overall during the same period.

Compared to other states, Colorado has the 16th-lowest homicide rate in the nation. Nearby states in the northern Rocky Mountain region have some of the lowest homicide rates in the nation with Idaho coming in at 1.9 per 100,000, and Utah at 1.8 per 100,000:

For additional reference here is a map of all states:

As The New York Times and New York magazine recently reported, the nationwide increase in homicides was largely driven by increases  in violence in seven large cities nationwide. Denver was not among them.

Regional Differences

A with the nation overall, Colorado's homicide situation varies geographically. Within the state's metropolitan areas, the homicide rate ranges from 0.9 per 100,000 in Boulder to 8.0 in Pueblo. Pueblo's homicide rate is higher than both the statewide rate and the nationwide rate:

Homicides increased in Pueblo, Colorado Springs, Denver, and Fort Collins. The rate in Greeley was unchanged. Homicides decreased over the year in Grand Junction and Boulder. Colorado overall remains remarkably safe in terms of homicides. With homicide rates around 2 per 100,000, northern Colorado and Western Colorado are among the safest places on earth.

Pueblo's rising homicide rates have been blamed on increasing gang activity.

Statewide in 2015, there were 176 recorded cases of "murder and nonnegligent manslaughter." That's up from 151 cases in 2014.

Of those 176 in Colorado, 108 were reported in the Denver metro area (with 53 in the city of Denver itself). 26 were reported in Colorado Springs and 13 in Pueblo. Pueblo's homicide rate is much higher, of course, because its population is considerably smaller than that of Colorado Springs or metro Denver.

In areas outside of Metro Denver, we're generally dealing with small numbers. Pueblo's total homicides increased 30 percent from 2014 to 2015, but that's an increase from 10 to 13 over the period. This hardly indicates an established trend toward an epidemic of violence. Nevertheless, one would prefer to see movement in the opposite direction.