Last week’s post on NIC’s data group introduced some of the teams and processes involved in collecting the most accurate and reliable data available on America’s seniors housing and care properties. This second half of the story delves into processing and reporting on the data.

A team of three full-time analysts is responsible for processing the data from the Quality Assurance team and completing quarterly reporting on it. The team also produces specialty reporting throughout each quarter, such as the Seniors Housing Actual Rates Report and the Skilled Nursing Data Report. They also support all of NIC’s presentations and webinars with accurate, current data, analysis, and visualizations.

Working in collaboration, the Quality Assurance and the Data and Analytics teams jointly execute each quarter’s data finalization process. Data and Analytics Manager Leighann Garcia explains, “The two teams bring two different perspectives to the data. In the Quality Assurance department, they tend to have a property-level expertise, while the Data team looks at it from a higher level, looking at broader trends over time and at the aggregate, rather than at individual properties. Combining those perspectives works well when we’re finalizing the data before reporting on it. From there, we produce about 45 different reports each quarter.

Research Manager Brian Connolly added that, “the delineation between the Quality Assurance team’s property-level view and the Data Team’s aggregate view is deliberate. It frees up the Data Team to view the data without any bias that might come from looking at a particular property. On the flip side, the QA team can look at the property level without any bias that comes from the aggregate. We put that in place on purpose. That’s why the partnership between the teams works so well. Both teams find important nuggets, and, as a group, we are better able to direct our data collection efforts as a result, particularly as we identify long-term trends that might need clarification. When the Analytics Team looks at the finalized data for publishing and reporting, they know it’s reviewed and as clean as possible. When they see trends, they can be confident they’re not just seeing something that was missed in the QA or data collection process.”

Connolly’s team conducts regular audits, to further ensure nothing is overlooked. “We review the entire database, auditing each market individually versus state data. We’ll go to a given state’s department of health, for example, and find all their lists and compare them to our database. That’s one way for us to proactively research each of the markets outside our normal process of data collection and make sure our database is as complete as possible.

With clean data in hand, Garcia’s team generates reports throughout every quarter. “Reports fit into one of several categories,” she explained. “The first of these is reports that go to clients who have requested a customized report. We also generate reports that get uploaded to the NIC MAP^® Client Portal. Some reports are for internal use. For example, we provide our marketing department with data for their content and generate a few specialized reports for our senior leadership. We also work very closely with our Outreach Team, providing data for their webinars and presentations. We’re closest to the data, and often call out trends from the very beginning of the process.”

NIC is careful to spread out the collection of data points for each metropolitan market throughout the quarter. The way they structure their agreements with data providers – and execute their own collection processes – ensures that no metropolitan market’s data is collected within a single month, for example. All of NIC’s data collection occurs as evenly as possible throughout time and across markets.

Partnerships with leading software vendors in the space have streamlined processes by which NIC receives data. Customized reports, designed to automatically feed data to Connolly’s team, have significantly boosted efficiency, while helping to reduce human error. Because NIC either talks to or receives data from nearly every property in its database, every single quarter, accurate and efficient data collection processes are essential.

Reports are generally built in three different ways. Some are custom developed in SQL, others use the data tools in Excel, while still others are coded in a programming language called R. Using these tools, the team generates reports, some of which are complex and very time-consuming. Once generated, every report goes through a quality assurance process to be validated.

“One key role we play is subject matter expertise on our data,” added Garcia. “While we don’t advise on strategy or make predications, we can help clients if we know what they’re trying to achieve with our data. A client that’s running 14 reports manually in every one of their subscribed markets might come to client services and ask if we can save them time, for example. We’d be able to help them with a custom report.”

Looking ahead, the data group plans to expand the database to include, among other things, seniors housing financial benchmarking data. NIC’s Seniors Housing Financial Benchmarking Initiative will create a time series database that allows for consistent financial benchmarking of select property-level operational revenue and expense categories. “This will be another way our clients can benchmark their communities,” said Connolly.

Through everything they do, NIC’s data teams share a passion for accuracy. “NIC is really focused on getting the numbers right,” said Dan Raney, Director, Product Solutions & Technology. “It’s personally offensive to us if the numbers are wrong, even for a single property. I really believe that the care we invest on data accuracy at NIC is second to none.”

“And also, our transparency in the event that we have an oversight,” added Garcia. “Those things do happen. There have been errors, but we’re very forthcoming and transparent when that happens. I believe our customers appreciate that.”

New initiatives, new sources of data, and new reporting products will all be executed by NIC’s dedicated data group, on top of the considerable effort that goes into daily, weekly, and quarterly data collection, QA, analysis, and reporting output which NIC clients have come to rely on, year after year.

A key topic of conversation among seniors housing investors and operators is the impact of inventory growth on occupancy rates and the ability to grow rents and revenue. Related to inventory growth are rising costs of construction, which may be one factor influencing a slowing of construction activity. This blog highlights some recent data to support those discussions.

Seniors housing construction may be plateauing. NIC MAP^® data from the fourth quarter of 2018 continues to suggest that construction activity may be slowing. There were roughly 37,000 units under construction at Seniors Housing properties (Majority Independent Living (IL) and Majority Assisted Living (AL)) for the NIC MAP 31 Primary Markets. That figure is down from the record high of 43,826 in the fourth quarter of 2017 or by nearly 6,500 units. Moreover, construction as a share of inventory slipped back to 6% of open inventory in 4Q2018, down from the recorded high of 7.3% in 4Q2017.  

Construction starts (units that have newly broken ground) for both Majority IL and Majority AL are also trending lower. In the fourth quarter, AL starts totaled nearly 1,552 units, the fewest starts since the first quarter of 2012. On a four-quarter aggregate basis, starts totaled 9,393 units, the fewest since 2014. As a share of inventory, this amounted to 3.3%. The last time it was below 4% was 2012. IL starts, on a rolling four-quarter basis, totaled 7,287 units in the fourth quarter. As a share of inventory, this equaled 2.2%. The last time it dropped below 2.5% was 2014. 

It is important to note, however, that construction data sometimes is restated—both up and down. Construction starts are some of the most commonly restated data points as they are particularly difficult to capture, so these numbers are likely going to see some changes. We occasionally find out that a project has broken ground after it has done so or that a property indicated ground break pre-maturely. Part way through a project, a property manager may adjust their plans for unit mix based on pre-leasing patterns, unforeseen challenges, or other factors. Changes to unit mix can also mean changes to a property’s majority property type designation. 

Construction costs are increasing. The Weitz Company is a major architectural/engineering/construction firm in the United States. Based on data they collect from seniors housing projects they work on and cost estimates from RSMeans, Weitz produces estimates of construction costs for seniors housing by metropolitan area. Larry Graeve, Senior Vice President at The Weitz Company, and Amy Burk, Estimator, prepare Special Issue Briefs for the American Seniors Housing Association (ASHA) focused on construction costs of Seniors Housing.

Their recently released special brief on Winter 2019 Seniors Housing Construction Costs indicates costs rose by 6 to 8 percent year over year. Weitz cites tariffs and labor challenges as contributing factors to the rising costs. As the Weitz reports highlight, geographic variations in cost are notable. The data in the graphic below present figures on a city index of 100, with an example of Wichita, KS (city index 85.6) in the furthest right two columns, demonstrating local variability by product type. For example, there is a cost range of $134 to $159 per square foot for independent living in Wichita, while Philadelphia, which has an index of 115.2, translates to a cost rate of $180 to $214 per square foot.

Local variation exists in seniors housing construction. Not only do costs differ by region, but construction activity also varies due to local development cycles and broader macroeconomic factors. The heatmap below shows local market variation.  Although many markets are “cooling” (as evidenced by the blue tones), some metro area markets still have heightened activity. One case in point is Atlanta, which of the CBSAs in the Primary Markets, had the highest level of construction as a percent of its inventory for 4Q18 at 14.8% (it was 6% for the Primary Markets). Close behind Atlanta is Sacramento, which remained at its record high level of construction as a percent of inventory (12%) for the second consecutive quarter.

Conclusion: many factors could be at play. Construction activity for seniors housing appears to be slowing—by both measurements of starts and active construction underway. While this would be good news for many operators and investors who have been facing the consequences of new supply and competition, we remain cautious until several more quarters of data validate this trend. The slowdown, should it stick, may reflect rising costs, as showcased by the Weitz data as well as other cost and wage data not discussed in this blog. Other factors impacting construction activity may include a tighter lending environment by many of the larger banks and lenders, a more competitive employment environment for retaining construction workers, more cautious investment attitudes at this stage of the development cycle, and possibly concerns about the broader national economy. Additionally, markets experiencing declining occupancies may be discouraging developers from pursuing new projects in potentially challenging areas.

Hundreds of businesses, including capital providers, developers, owners, operators, service providers, and others, rely on the NIC MAP^® Data Service as they research, analyze, and assess their markets and prospects across the country. Clients know they can trust the accuracy of the data NIC makes available, as they leverage the ever-expanding utility of NIC’s online platform; but they may be unaware of what happens behind the scenes in order to produce such quality data. There’s a lot going on, so we’re breaking this into two parts, and will post the second half next week.

In fact, NIC has invested heavily, since its earliest years, in its data collection and quality assurance practice, to achieve its mission for improved transparency into the seniors housing and care property markets. NIC’s efforts have helped propel the sector to its current status as a leading property type in the commercial real estate industry. Quality data has increased access to capital, lowered costs, and, ultimately, helped improve access and choice for America’s seniors.

Today, NIC supports a talented and experienced in-house data group dedicated to producing the finest quality data available, and enjoys close, long-term alliances with some of the industry’s leading vendors of data collection services. In 2018, 22 people, employed both in-house and by vendors, produced four quarterly data releases, two Seniors Housing Actual Rates Report releases, all of the data for the Fifth Edition of the NIC Investment Guide, and many custom data requests, both for internal and external clients. All the while, the group worked daily to collect thousands of data points, assure quality across its entire platform, add new data from new sources, research and analyze trends, streamline processes and develop new reports and new features.

To shed light on what it takes to produce the quality and volume of data that NIC offers, NIC convened the leadership of its data group to discuss their approach – and share how NIC’s data investments are providing value to clients – and the sector overall – every day.

Dan Raney, Director, Product Solutions & Technology, started off by explaining that NIC collects data from a number of sources: “Many of the larger contributors are able to give us detailed reports from their property management systems, but we also work closely with the ProMatura Group, who gets information directly from properties over the phone. They have a team of 15 people dedicated to data collection for NIC, and don’t work on anything else. “

The in-house group that manages NIC’s data is divided into several distinct teams. There is an internal research and data collection team that focuses just on construction data. Three team members are responsible for researching projects in the pipeline, from the early planning stages through construction starts. They comb through a complex collection of resources to identify and track proposed and planned construction in 140 U.S. metropolitan markets. Once a project breaks ground, the ProMatura Group tracks the property through its opening day.

NIC’s systems have built-in checks designed to flag and block any data that sits outside given parameters. The three-person Quality Assurance team looks at thousands of records daily, in order to assess whether the data has errors in it – or whether indeed they point to actual shifts in the market. “Our teams are constantly reviewing quality assurance reports throughout each quarter.  They are very detailed in their inspection of the data, diligently examining the nuances of each data set (skilled nursing, actual rates, etc.). Although each data set is unique, we’ve put together a common quality control process that ensures the accuracy of all data before it is reported to our clients.” said Raney.

Research Manager Brian Connolly’s group is responsible for initial data collection and quality assurance. Many sources provide data constantly, while others provide periodic updates. Some data sources are one-time downloads. In the case of ProMatura’s contribution, Connolly says, “ProMatura is calling the same properties in the same markets every quarter. They’ve been able to build relationships with these communities and have built a lot of trust with the people they’re talking to provide the detailed data points that NIC requires for its reporting. They’re speaking the same language. Every quarter they call about 13,000 properties. This past quarter they finished 96% of those calls, meaning they’ve entered all the relevant data reporting into our survey instrument and submitted it for our review and publishing. Again, should any data they report fall outside our thresholds, it is flagged for review. We review those records every day, as they come in.

“ProMatura has worked with us for over ten years. They understand how this data is used in the industry, and how important it is to everyone out there that it is accurate and complete. We have an excellent partnership with them. They also manage data that some of our corporate contacts enter into workbooks, which they then enter into our system, which then can be reviewed for quality along with all the other data.

“We also get data directly from corporate partners at the property level, which is now a somewhat automated process. There’s a separate QA process for this type of electronic data submission. The systematic computerized process makes it much easier for these partners to submit their data to us and reduces the strain on our own resources as well.”

All property data is collected by Connolly’s group, and combed and checked for accuracy. He explained the purpose of NIC’s multiple checks prior to analysis: “The point is to have some redundancy in the way we review all the data. It means we have numerous eyes on every data point, with numerous checks throughout each quarter. The thresholds we use are based on historical trends for that property. We can use them to benchmark the data we’re collecting and make sure there are no surprises sitting so far out of the trend that we need to look at it more closely and clean it up. After the collection period ends, in the data finalization process we run all of our checks repeatedly to make sure the data is clean before it is turned over to our analysts for reporting.”

Data collection is far from the end of the story. Next week’s post will delve into the analysis, quality assurance, and reporting efforts of NIC’s data group.

The Labor Department reported that there were a significant number of jobs added in January–304,000 and well above the consensus expectation of 172,000. This was the 100th consecutive month of job growth.  December was revised down to 222,000 from 312,000 and November was revised up to 196,000 from 176,000.

The 800,000 Federal workers going unpaid during the partial federal government shutdown were still counted as employed and as a result federal government payroll estimates were largely unchanged in January. In contrast, the uptick in the unemployment rate may have been influenced by the furloughed employees due to definitional differences in the household survey which is used to estimate the unemployment rate.

Indeed, the unemployment rate increased 0.1 percentage points from 3.9% in December and from a near-50 year low of 3.7% in November to 4.0% in January. Nevertheless, the jobless rate remains well below the rate of what is generally believed to be the “natural rate of unemployment” of 4.5%, which suggests that upward pressure on wage rates will continue. Further indications that this is in fact starting to occur were released in the report. Average hourly earnings for all employees on private nonfarm payrolls rose in January by three cents to $27.56. Over the past 12 months, average hourly earnings have increased by 85 cents, or 3.2%.

A broader measure of unemployment, which includes those who are working part time but would prefer full-time jobs and those that they have given up searching—the U-6 unemployment rate—rose to 8.1% in January from 7.6% in December.

In January, employment in health care rose by 42,000. In the past year, health care has added 368,000 jobs.

The labor force participation rate, which is a measure of the share of working age people who are employed or looking for work rose 0.1 percentage point to 63.2% in January, still very low but up from its cyclical low of 62.3% in 2015. The low rate at least partially reflecting the effects of an aging population.

Adding further evidence to the strong job market are unemployment insurance claims which recently reached a near 50 year.

In an announcement by Federal Reserve Chairman Powell this week, the Fed indicated that they would not be raising interest rates in the immediate future due to concerns about global economic growth and limited inflation. That said, the jobs report shows that the labor market remains robust. Further indications of strength in the economy could cause the Fed to raise rates sooner. In December, the Federal Reserve raised the fed funds rate by 25 basis points to a range of 2.25% to 2.50%. That marked the fourth increase in 2018. The Fed has now raised rates by a quarter percentage point nine times since late 2015, after keeping them near zero for seven years.

Expanding on a recent NIC Blog Post that detailed care segment occupancy across the NIC MAP^® 99 Primary and Secondary Markets within Continuing Care Retirement Communities (CCRCs, also referred to as life plan communities) compared with those in non-CCRC freestanding or combined communities, the second installment of this two-part blog post examines the regional occupancy performance of independent living, assisted living, memory care and nursing care segments within and across entrance fee and rental CCRCs.

The following narrative describes 3Q2018 CCRC occupancy aggregated from the NIC MAP Primary and Secondary Markets—99 of the nation’s largest core-based statistical areas (CBSAs), broken out across eight geographical regions. (Note that NIC MAP recently released data for 4Q2018). For the analysis, occupancy data for CCRCs, both entrance fee (EF) and rental payment CCRCs was considered. CCRC segments inventory was comprised of 63.3% entrance fee and 36.7% rental payment models in 3Q2018. The occupancy rate used was the “all occupancy” rate which includes properties still in lease-up as well as stabilized properties.

Entrance Fee CCRCs vs. Rental CCRCs

Entrance fee CCRCs generally had higher occupancy than rental CCRCs, with some variability by care segment and by geographic region. Entrance fee CCRC occupancy was 3.3 percentage points higher than rental CCRCs in the third quarter of 2018 (92.1% vs. 88.7%). Across regions, entrance fee CCRCs led rental CCRCs by 2.8 percentage points in the independent living segment (93.0% vs. 90.2%), 1.6 percentage points in the assisted living segment (91.9% vs. 90.3%), 3.8 points in the memory care segment (90.9% vs. 87.2%), and by 2.5 percentage points in the nursing care segment (89.3% vs. 86.8%). Note: the referenced percentage point differences were rounded.

• Independent living segment occupancy was the highest in the Mid-Atlantic region for both entrance fee CCRCs (95.1%) and rental CCRCs (93.2%). The largest differences in independent living segment occupancy between entrance fee and rental CCRCs was reported in the Mountain region (8.4 percentage points). Independent living segment occupancy was slightly higher in rental CCRCs than entrance fee CCRCs in the Southeast region (by a one percentage point difference).
• Assisted living segment occupancy was the highest in the Northeast for rental CCRCs (93.7%) and the West North Central and Northeast regions for entrance fee CCRCs (93.4% and 93.3%). Rental CCRCs had higher assisted living segment occupancy than entrance fee CCRCs in the Southwest, Southeast and Northeast regions. The largest differences in assisted living segment occupancy between entrance fee and rental CCRCs was reported in the Mountain region (10.9 percentage points).
• Memory care segment occupancy was highest in the East North Central and Mid-Atlantic for entrance fee CCRCs (94.4% and 93.4%) and the Northeast and Southwest for rental CCRCs (94.0% and 93.6%) but was particularly low for rental CCRCs in the Mountain region (69.1%). The largest differences in occupancy between entrance fee and rental CCRCs was reported in the Mountain region (17.3 percentage points) and Pacific region (11.9 percentage points), where entrance fee CCRC occupancy was higher than rental CCRCs, and in the Southwest, where rental CCRC occupancy was 11.0 percentage points higher than that for entrance fee CCRCs.
• Nursing care segment occupancy was highest in the Northeast for both entrance fee and rental CCRCs (92.9% and 91.8%). Occupancy was particularly low in the Mountain region for rental CCRCs (76.4%). Rental CCRCs had higher nursing care segment occupancy than entrance fee CCRCs in the Pacific and Southeast regions. The largest difference in nursing care segment occupancy between entrance fee and rental CCRCs was noted in the Mountain region (11.2 percentage points).

Higher occupancy in entrance fee CCRCs may be due in part to higher rates of resident turnover in rental-model CCRCs, according to data from ASHA’s State of Seniors Housing, 2018 database. While communities with rental CCRC payment plans offer monthly or annual leases for housing and services, entrance fee CCRCs include contract stipulations such as refund percentage and timing of refund, which may help retain residents who put down a sizable up-front fee to reside in a CCRC. An additional reason for higher occupancy rates in entrance fee CCRCs may be due in part to the fact that the majority of entrance fee CCRCs are nonprofit organizations (72.6%). According to NIC MAP, nonprofit CCRC occupancy in the Primary and Secondary Markets has consistently exceeded that of for-profit CCRCs by more than four percentage points since 2Q15.

However, product differences are unlikely to tell the whole story. Further analysis is needed to more fully explain CCRC segment occupancy performance by payment plan and by region. In addition to local economic drivers, income levels, inflation-adjusted purchasing power, and penetration rates, other contributing factors that could be explored include inventory growth patterns, consumer acceptance of the entrance fee model, and the health of the residential housing market—specifically as it relates to entrance fee CCRC occupancy performance since often CCRCs target upfront fees to local home sales prices near or around the median to enable consumers to fund a move with a near one-to-one exchange.

(Double-click the image below to enlarge.)