Hotel Demand and Real Personal Income: Inextricably Linked!

By Jack Corgel, Ph. D. and Brett Edgerton

Executive Summary

  • The BEA reports personal income as an aggregate of employee compensation, proprietors’ income, rental income, personal income on receipts from assets, and personal current transfer receipts. Employee compensation (i.e., wages and salaries) now constitutes just over one half of this measure.
  • Personal income reflects the potential for individuals to engage in travel and serves as a general indicator for judging the financial well-being of businesses to support travel.
  • Since 2010, real personal income has more accurately matched hotel demand growth than other metrics including employment and GDP. CBRE Econometric Advisors forecasts above long-term average growth in personal income—a signal of continued strong hotel demand.
  • Years of experience at CBRE Hotels’ Americas Research shows that personal income is more reliable for forecasting hotel demand particularly across local markets than GDP, consumer expenditures, and disposable income because of inherent differences between how each economic measure is calculated and the difficulity of each to be forecast.
  • Upper-price hotels rebounded quickly from the Great Recession due to personal income growth among those in the upper one-half of all incomes while lower priced hotels, whose demand relies more heavily on wage earners employed in industries experiencing the highest unemployment and income losses during the recession, recovered years later.
  • Lodging market investors and managers should review periodic reports of personal income as a fundamental indicator of the underlying strength of hotel demand. Personal income can be monitored at the national and local levels.

I.    Introduction   

This article expands on the idea that market participants and analysts should pay close attention to periodic reports of personal income (PI) as the primary indicator of underlying lodging demand strength (weakness). While other economic measures correlate well with hotel demand across the U.S. market, among chain scales and locations, and for MSAs, our sixteen years of experience forecasting hotel demand provides a basis for the conclusion that PI, specifically for modeling purposes real personal income (RPI), dominates other economic measures for reliability and consistency. The U.S. and most cities are now experiencing robust RPI growth, thus despite signals of downside risks emanating from global economic weakness and financial market volatility, hotel markets across the nation are poised for demand growth in line with long-run averages despite record ADR levels in the industry. Importantly, macroeconomic forecasters expect RPIs to rise between three and four percent over the next few years.

Travelers will make trips and hence require lodging if they ‘have the money’ and are not afraid to travel; either for fear of their security such as after 9/11 or fear for their economic security such as during the financial crisis. In the absence of elevated travel risk concerns, RPI levels and growth provide time-tested guidance about the financial well-being of household and business travelers.

Why is RPI such a reliable and consistent indicator of hotel demand? How can hotel managers and investors use this knowledge to make more informed decisions? We provide answers to these questions. In the section that follows, we demonstrate the reliability and consistency of the relationship between RPI and hotel demand holding other factors constant. In the next section, we decompose the U.S. Department of Commerce, Bureau of Economic Analysis (BEA) definition of PI. Wages and salaries, perhaps the most closely watched income measure and logical hotel demand driver, constitute just over one-half of total personal income and have been declining for decades relative to other income sources. Section Three reviews various measures of aggregate income and expenditures, such as GDP, personal consumption expenditures (PCE), and disposable income for their applicability to forecasting hotel demand relative to RPI. The concluding section provides our views on what managers and investors should look for when examining periodic reports of personal income.

II.    The Personal Income and Hotel Demand Relationship

Over the past two decades of forecasting hotel markets, we found RPI to be the most reliable variable for explaining variation in hotel demand throughout the cycle. Currently, hotels are exhibiting strong financial performance coincident with employment gains, income growth, and other economic indicators. Rising from the depths of the Great Recession, however, employment and GDP stagnated while PI grew, particularly among individuals positioned in the top one-half of the U.S. income spectrum. The hotel sector of the economy historically ‘lags in recovery.’ Yet during the recent recovery, there appeared a clear bifurcation among upper-priced and lower-priced chain scales. Corgel and Woodworth (2012) show that upper-price hotels rebounded quickly from the Great Recession due to PI growth among those in the upper one-half of all incomes while lower priced hotels, whose demand relies more heavily on wage earners employed in industries experiencing the highest unemployment and income losses during the recession, recovered years later. The V-shaped rebound out of the Great Recession for all but the lowest price hotels manifested in the hotel sector actually ‘leading in recovery.’

Exhibit 1 presents CAGRs from 2010 to present for U.S. hotel demand, RPI, real GDP, and total employment. The same data are provided for several large MSAs. These data show that in every market RPI CAGRs equaled or exceeded those for real GDP and employment, in some cases by more than 100 basis points. Also, the RPI CAGRs match hotel demand CAGRs more closely than other measures. Employment experienced strong growth during the past two years, while RPI growth has been consistently strong since 2010. 

Hotel demand modeling is the central activity at CBRE Hotels’ Americas Research for generating Hotel Horizons® forecasts. This activity involves estimating long-run, stable relationships between economic measures and the number of rooms sold (i.e., hotel demand) by applying accepted econometric methods to data covering several cyclical phases. In repeated tests of our national, chain scale, location, and city demand models, RPI outperforms other economic variables. At the national level GDP works almost as well as RPI in explaining hotel demand, but at the local level metropolitan RPI dominates metropolitan GDP.

II.1  Show the Money: Five Categories that Comprise Personal Income

Dissecting the BEA’s PI measure into its components lends insight as to why RPI is the dominant indicator of hotel demand strength. The BEA derives PI from five categories of income: employee compensation, proprietors’ income, rental income, personal income on receipts from assets, and personal current transfer receipts. Exhibit 2 presents the five PI categories as well as our view about how each relates to hotel demand. The aggregate PI measure informs about the amount that can be applied to tax payments and consumption while holding wealth constant. Each category reflects on individuals’ gains from their productivity.
Wage stagnation since the Great Recession has been a much discussed, but somewhat misunderstood, topic. Furth (2015) presents evidence that wage growth rates during the past few years are not substantially different than historical rates. Also, recent nominal wage growth, while hovering around one percent, must be considered in the context of very low inflation, recording year-over-year increases of less than one percent during most months since 2013. The other categories captured in the PI data have overridden these weak wage gains as indicated by the 2.7 percent CAGR for RPI from 2010 through 2015 (see Exhibit 1).1  

Exhibit 3 presents the mix of PI categories from 1970. Even as far back in time as 1970, wages and salaries only comprised 60 percent of PI. During the subsequent 45 year period wages and salaries slid to slightly over 50 percent of the total, and as shown in the exhibit all the other components of PI proportionally grew. This indicates that lodging demand has increasingly becoming a function of non-wage and salary income sources.

IV.    Why Not GDP, PCE, and Disposable Income for Hotel Demand Forecasting?

In the previous section, we examine the categories of income that comprise the BEA measure of PI. This metric is one among many governmental agency measures tracking incomes within the U.S. Most notably, the BEA produces GDP and its component part PCE for aggregate income and consumer spending. The U.S. GDP consists of the four general output categories in Equation (1):

GDP = C + I + G + NX          (1)
where (C) is PCE, (I) is business investment, (G) is government spending, and (NX) represents imports minus exports, or net exports.

The main reason GDP is not as reliable an indicator of hotel demand as PI is that the four components of GDP sometimes unexpectedly move in opposite directions thus masking the strength of the two components that have the most meaning to hotel demand, (C) and (I). As shown in Exhibit 4, this is what happened during several quarters since the Great Recession. Also, from 2006 to early 2008 GDP grew as the result of net exports and government spending, yet hotel demand was flat or negative as businesses cut back on investments and consumers began to spend less.

The correlation between PI and PCE is quite high so PCE may explain variation in hotel demand about as well as PI in many forecasting models. Yet, PCE measures actual consumption expenditures while PI informs about potential consumption expenditures. Forecasts of historical expenditures, if used to drive hotel demand forecasts, may not pick up shifts in expenditure patterns whereby PI avoids this problem.

IV.2 Why Personal Income and Not Disposable or Discretionary Income?

Personal income (PI), disposable income (D1I) and discretionary income (D2I) are related in the following ways:

D1I = PI – direct taxes          (2)

where direct taxes include federal and state income taxes while indirect taxes, not reflected in Equation (2), include value add taxes, sales taxes, and employer contributions to social security. Also,

D2I = D1I – typical expenses to maintain a certain standard of living2.          (3)

We trust forecasts of the aggregate PI while having less faith in future estimates of direct taxes and certain specific living expense deductions. This is in part the result of the fact that federal and state tax law changes impact D1I and are not easily forecast. Likewise, D2I is challenging to predict given the high level of volatility in food and utility prices.

Exhibit 5 compares the correlations between changes in PI and D2I, PCE, and GDP since 1965 using trailing five year data. While the long-term series are all highly correlated, during certain periods, particularly since 2000, these economic variables followed somewhat different paths. The deviation may be partially due to the fact that technology changes are difficult to incorporate into GDP measurement (Brynjolfsson and Saunders 2009) since consumers potentially are consuming more but not necessarily engaging in transaction that can be tracked (e.g., people watch YouTube instructional videos at no cost, whereas historically this consumption would require purchases). Similarly, energy price shocks in the 1970s and 2000s caused correlations between RPI and disposable income to diverge as incomes rose, while living expenses increased more quickly, thus reducing disposable income.

The divergent correlations among economic variables imply that their ability to predict hotel demand may change over time. In fact, there is evidence such a shift has occurred since 2010. Exhibit 6 shows the correlations among the alternative economic variables. We find that the association between changes in hotel demand and RPI remained steady in recent years, while real GDP, Disposable Income, and GDP have proven less consistent.

In summary, we have the most confidence in PI to inform about hotel demand over alternative economic measures because it provides reliable signals about consumers’ and businesses’ abilities to spend and avoids some troublesome forecasting issues. Furthermore, recent history suggests the statistical relationship between hotel demand and other measures, such as GDP, has weakened relative to that of PI. Because we are interested in the link between future hotel demand and economic variables that relate to potential consumption, RPI dominates alternative measures of income on conceptual as well as empirical grounds.

V.    Interpretations for Hotel Managers and Investors

Hotel investors and managers must continuously monitor relevant indicators of current and future financial performance to make informed decisions. Monthly reports from STR, Inc. on individual hotel market shares within competitive sets have an important role in this process. Economic data on national and local market demand and supply conditions also need to be consulted for signals of changes in market conditions that have not yet manifested in market share reports. In this regard, Hotel Horizons® forecasts and research studies from CBRE Hotels’ Americas Research become valuable.

The reporting on hotel markets by CBRE Hotels’ Americas Research can be even more valuable to investors and managers if the process of converting economic data into hotel performance measures is transparent and verifiable. Given the lead times for development decisions and construction, the consequences of upcoming hotel openings and closings can be ascertained. Understanding the implications of changes on the demand side of the market, however come with a higher degree of difficulty. Our long-term experience in hotel market forecasting leads us to the conclusion that, despite claims of hotel demand being the product of forces measured with a broad array of economic variables, variation in the number of rooms sold across hotel markets and over time can be explained quite well with a couple of broad economic indicators. Specifically PI, and to a lesser extent employment, subsume the relevant information contained in most other variables.

The total PI received by individuals not only incorporates essential information about the potential for individuals to spend on travel but also serves as a general economic indicator of employer health and the strength of sectors such as real estate (i.e., rental income is a component of PI). Consequently, PI serves as an indicator of both future leisure and business travel opportunity. During 2016, real incomes are forecast to rise by 3.4 percent according to CBRE Econometric Advisors and then by 3.8 percent in 2017, both exceeding the long-run average PI growth rate of 2.4 percent since 1988. Most cities also are also forecast to see robust income growth in the next two years. If these projections prove accurate, economic fundaments should help alleviate fears of the downside risks currently presented by slowing global growth and financial market volatility. Whether the outlook is positive or not, PI behaves as the “canary in the coal mine” for when hotel demand will rise or fall as the two are inextricably linked.
1 The post-Great Recession RPI growth is well below long-run (i.e., 1980– 2007) rate of approximately six percent.
2 Rent or mortgage payments, utilities, insurance, medical, title, transportation, property maintenance, child support, food and sundries.


Brynjolfsson, Erik, and Adam Saunders (2009). "What the GDP Gets Wrong (Why Managers Should Care)." MIT Sloan Management Review.

Bureau of Economic Advisors (2015), Concepts and Methods of US National Income and Product Accounts. Bureau of Economic Advisors, December.

Corgel, J.B., and R.M. Woodworth. (2012). Why hotels? Economy weakens but hotels remain relatively strong – what gives? And what might give? Cornell Hospitality Quarterly, 53(4), 270-273.

Furth, S. (2015), Stagnant Wages: Fact or Fiction? Heritage Foundation, March 11.

Ruser, J., A. Pilot, and C. Nelson (2004). Alternative Measures of Household Income: BEA Personal Income, CPS Money Income, and Beyond. Federal Economic Statistics Advisory Committee.


Jack Corgel Brett Edgerton
Managing Director, CBRE Hotels’ Americas Research Economist
Professor of Real Estate at the Cornell University School of Hotel Administration CBRE Hotels’ Americas Research
Acknowledgments: Bram Gallagher, Jamie Lane, Robert Mandelbaum, and Mark Woodworth made helpful comments on drafts of this report.