The COVID-19 pandemic has altered life in numerous ways, extracting a heavy toll on families, businesses, and society. Economic losses have been harsh, with ongoing job loss and business closures. During prior volatile periods, economic data was indicative of future trends, offering insight into potential inflection points within a downturn as well as the pace of a recovery. In this unprecedented time of COVID-19, once-reliable datasets used to gauge an economic recovery have been hampered or delayed by business restrictions and health safety precautions. Additionally, some standardized metrics may not provide an accurate reflection of current activity given the unique influences of the pandemic on the economy.
With traditional data less insightful during these unprecedented times, are there alternative measures that can help frame the pace of the recovery? The answer is, yes. Public agencies and private companies have stepped in to fill some of the void, using big data techniques and sophisticated modeling and sampling to provide insights and timely data on market segments. Analyzing these indicators can provide near real-time insights on the trajectory of the economic recovery from the COVID-19 pandemic as well as possible leading indicators for real estate performance. Beyond the near-term, many of these high frequency indicators will persist as useful tools in evaluating economic and real estate market performance.
Traditional Data for a New Era
During this unprecedented period and with local restrictions on business activities, it can be difficult to interpret traditional economic datasets to gauge the recovering economy. In some cases, data collection efforts have been complicated or delayed by restrictions on workplaces and a focus on health safety. In others, standard metrics may not provide an accurate reflection of activity given the unique influences on the economy. With some regions and industries able to fully reopen while others operate with capacity or other restrictions, ripple effects throughout the labor force and supply chain can be difficult to measure. Compounding these issues, consumer behavior has also been altered by enhanced unemployment benefits, stimulus payments, eviction moratoria, credit card and mortgage forbearance, and other government programs.
Challenges with Traditional Data
Standardized statistics can be very useful in analyzing economic conditions during normal periods. Traditional metrics, such as GDP, unemployment rate, and inflation, are informative and have long historical datasets that can be analyzed through different economic cycles. Yet in the midst of a pandemic, when shifts in the economy are determined by the course of a virus as well as government imposed restrictions on business activities, timeliness and inclusion of newer segments, such as gig workers, are perhaps more important than a long time series.
As an example of the data analysis challenges faced during the pandemic and recovery, the national unemployment rate is often a bellwether for economic performance. In normal economic periods, the unemployment rate, along with other releases of employment data, provides an accurate depiction of the state of the economy. Investors of all types utilize these economic data points to inform decisions, and stock and bond markets can rise or fall depending upon the data. Furthermore, one half of the Federal Reserve’s dual mandate is maximum employment, and monetary policy is influenced by the monthly unemployment reports.
After a record-shattering peak of 14.8% in April of 2020, the unemployment rate steadily improved to 6.3% in January 2021. A cursory glance would indicate this was a remarkable improvement, and a much quicker recovery than any prior recession. This improvement implied that nearly 17 million more people were employed in January than in April of last year. Furthermore, since 1970, the average unemployment rate is 6.2%, on par with the current level. This would appear to be an improving national economy with the unemployment rate headed in the right direction.
Yet, we know that many individuals remain unemployed, their chosen professions derailed by the pandemic, and unable to afford food and housing without assistance. And therefore some may assume that the level of unemployment is too high, particularly when compared with a national unemployment rate in the mid-3% range prior to the pandemic. Given the somewhat conflicting storylines, how should one view this unemployment rate?
New Data Sources
To help answer this question, we can turn to newly created data sources. In response to the difficulties of conducting surveys of households and businesses during the pandemic, the Census Bureau accelerated several experimental data products to fill the gap of relevant datasets during the pandemic. The Household Pulse Survey gathered data on the impact of the COVID-19 pandemic on people’s lives across the country at a weekly frequency. The survey utilized brief online question and response forms with text and email reminders, as opposed to the traditional surveying techniques of postal mail forms and in-person surveyors.
If we consider that one reason for tracking job loss is to provide supplemental income through unemployment insurance and other social measures, we can examine the new Census Household Pulse Survey for a comparable dataset. One of the questions within the survey asks respondents if they expect someone in their household to lose employment income within the next four weeks. The latest response from the beginning of February extrapolated to 57.8 million individuals, or roughly 23% of the population. During the last 12 weeks, the share that expected near-term employment income loss ranged between 23% and 31%, a narrow band.
The first survey, conducted at the end of April and beginning of May 2020, tabulated more than 96 million individuals expecting a loss of employment income within the following four weeks. So, while the share improved substantially since the onset of the pandemic-induced recession, roughly one-quarter of households expected some amount of near-term job loss throughout the last three months.
The lack of substantial improvement in this indicator even as the unemployment rate decreased indicates that job prospects remain poor for a large segment of the population and even though the unemployment rate equaled the historical average, further improvement is needed for the economy to expand.
How Technology Offers Real-Time Insights
Technology isn’t only allowing for different means of surveying individuals and businesses, but it can also be a direct source of data. Due to the evolution of technology and its prevalence in our modern lifestyles, there are several valuable sources of information that can be used in a myriad ways, including:
- Location tracking
- Search history
- Map directions
- Social media posts
Utilizing location data and map app searches, companies such as Apple and Google, can provide daily insights into changes in mobility patterns. These datasets can measure activity levels at essential and non-essential services, as well as workplaces and transit stops. As most office workers began to stay home in early 2020, transit patterns changed and visits to coffee shops and restaurants plummeted. One of the most important real estate questions during the pandemic is when will office workers return to their workplaces, particularly in gateway cities. We know that local economies and real estate markets ’won’t recover as long as downtown cores are void of office workers and foot traffic, but how can we determine when this recovery begins?
While it may be too soon to tell if office workers will return to central business district (CBD) towers permanently or continue to work remotely, mobility data allows insight into the degree to which employees are returning to workplaces. Analyzing Google mobility data for the United States, visits or activity at places of work remained on average 28% lower during the month of February compared with the previous year. While a substantial improvement from the initial months of the pandemic when activity declined by 40% to 50%, the average monthly activity remained somewhat stable during the last few months excluding holidays. One of the reasons for the stability is that most local economies reopened by the fall of 2020, and economies in several large cities, including those in California and New York, remain relatively restricted. As activity relative to the pre-pandemic periods begins to recover, the mobility data will provide a real-time indicator of this return of employees to workplaces and if these workers are visiting neighboring coffee shops and restaurants.
Interestingly, states that reopened economies earlier in 2020 didn’t recover as much activity as may be expected if the government restrictions were the only factor for reduced activity. Many individuals weren’t comfortable returning to workplaces full time and continued to work remotely when possible. Florida has constantly touted that it remained open for business, yet the combination of decreased travel reducing tourism-related business activity and remote working arrangements kept workplace activity 24% lower than it was in February 2020. Even without government restrictions, companies and individuals aren’t returning full-time to workplaces.
The mobility datasets can also provide insight into consumer activity. Even as retail businesses reopened in early and mid-summer in much of the country, shoppers were hesitant to return to stores in areas still impacted by elevated infection rates. Yet traditional government measures of retail sales indicated a strong recovery—7.4% in January compared with the previous year.
A Decline in Trips Outside the Home
Just as employees are often choosing to work remotely, individuals and households are electing to limit some shopping and entertainment trips. While shopping for necessities, including at grocery stores and pharmacies, remained common trips outside of the home, Google mobility data highlights a 25% decline in activity at retail, restaurants, and recreation locations in January 2020 to 2021. Some of this may be attributable to poor weather in parts of the country, but overall volume for these types of trips were lower in January and February 2021 than in 2020.
Rise in Online Shopping
While many retailers sell products through omnichannel platforms, smaller retailers and mom-and-pop stores don’t have the same ability to sell products online. The longer consumers stay away from physical storefronts, whether by choice or by local government restrictions, the more stores will shutter. While retail spending may in fact have increased in recent months, much of this purchasing activity transitioned to online shopping, often bypassing the many local stores and small businesses that occupy shopping centers. The government measure of retail sales was very positive, yet the increased purchases were not at local retailers but primarily through online shopping.
Similarly, data on restaurant reservations and dining activity provides an indication of whether consumers are returning to normal activities. With restaurants open for dining, albeit with capacity restrictions in some areas, the number of workers has rebounded to some degree. Analyzing only the recovery in restaurant employees doesn’t provide an accurate picture of the industry as many owners and managers are working line positions, for example, and takeout or curbside pickup require fewer staff than in-person dining. Anecdotal evidence points to fewer patrons at dining and drinking establishments, even though large shares of staff have been rehired in many parts of the country.
With sales tax receipts clouded by takeout and app-based delivery services, analysis of the recovery in restaurants can be murky. Restaurants with delivery options, fast food, and ghost kitchens may be doing relatively well, but many dine-in restaurants are barely surviving. This leads to media claims of tens of thousands of restaurants closing permanently. Yet, in many cases, struggling restaurants are persisting through the pandemic, at least for now. In the United States, approximately 80% of restaurants have reopened to some degree. Yet in February 2021, the count of diners was only roughly 53% of levels for February 2020, according to OpenTable, highlighting the consumers’ hesitancy to resume dining outside of their home.
Leveraging Data Sets to Understand Consumer Sentiment
The OpenTable data, as well as similar datasets, can offer real-time insight into consumer activity. Trends in dining activity will highlight the point at which more consumers feel comfortable returning to seated restaurants as well as any retrenchment should infection rates surge. Those with interests in the underlying real estate may analyze similar dining data to determine inflection points within dining trends in local markets.
As consumers become more comfortable making trips outside of the home and employees return to workplaces, mobility data can provide early insights into the emerging recovery. Analyzing these timely consumer datasets can shed additional light on household responses throughout the recovery cycle. While government statistics may track output or jobs, for example, related to reopening of industries, consumer data may provide an early indicator of how households perceive the relative safety of resuming normal activities, a leading indicator of when the economy may return to a more normal footing.
Innovation Has Its Advantages
The unique impacts of the pandemic have driven innovation in data collection and analysis. Many traditional surveys were forced to adapt to new methods of interviewing and in some cases adjusted methodologies to account, albeit perhaps temporarily, for issues such as respondents changing locations. Going forward, expect more timely data unencumbered by in-person surveys and postal mail responses released with greater frequencies.
The technology sector, unsurprisingly, was at the forefront of the rush to provide data and information to fill the void during the initial weeks of the pandemic and stay-at-home orders. While many companies amassed millions of terabytes of personal data, at times the utilization of these volumes of data for the public good were met with skepticism. During the pandemic, the potential utility of personal data was better understood by the general public as media reports highlighted mobility data, for example. As the uneven economic recovery continues, a growing range of data series can provide timely information on the pace of the recovery and which segments may be outperforming. Better and timelier data will lead to smarter investment decisions, particularly within the real estate sector.
This innovation in data collection and analysis will continue, long beyond the COVID-19 pandemic. Traditional economic indicators such as GDP, inflation, and unemployment will remain key indicators, but will be enhanced with modern data gathering techniques, including social media data mining, online surveys, financial transaction monitoring, and location tracking, allowing for more timely, and ultimately more accurate, data published with shorter intervals. New datasets will also begin to better capture information on groups such as gig workers and sole proprietorships, segments which often went untracked in some of the larger data series. Prudent investment due diligence should embrace some of the new datasets, incorporating the insights into analysis. Most of the datasets are publicly available in some form, and other options exist as well. Incorporating these real-time insights into investment analysis may offer an advantage and early indication of the economic recovery in local markets.
We're Here to Help
If you have questions about how these trends might impact your situation, please contact your Moss Adams professional.