Occasional StudiesThe Fiscal and Economic Impact of Walter E.Washington Convention Center ConferenceAttendees on the District of ColumbiaCurrent Trends and Prospects for Years 2020 to 2022Hamzah ShahBrown UniversityOffice of Revenue AnalysisOffice of the Chief Financial OfficerDistrict of Columbia GovernmentIssued December 2019

I.IntroductionBusiness travelers, as a segment of the tourism industry, are an increasingly importantcomponent of the national and local economies for four key reasons. First, there is the risingnumber of trade shows, conferences, business and networking events taking place wherebusiness professionals need to be present to meet clients and other business associates in person. 1Second, hotels often rely on business travel because it tends to be steadier year-round and morepredictable than leisure travel as many business events are organized well in advance. Third,many business travelers often visit the same area frequently. And fourthly, many businesstravelers tend to extend their business trip by one or more days in order to explore and enjoytheir business destination. 2 And because these trends are expected to continue into theforeseeable future amidst the growing national economy, the business travel industry is estimatedto grow nationally from 1.3 trillion in 2016 to 1.7 trillion in 2021.3In 2017, the District of Columbia attracted 22.8 million visitors to the city.4 Approximately, 41percent of these visitors came to the city primarily for business and 59 percent came to the cityprimarily for leisure. Being unable to account for all business travel into the District, this studyanalyzes the economic and fiscal impacts of conference attendees of the Walter E. WashingtonConvention Center (WEWCC). More specifically, this study assesses the effect of WEWCCconferences on annual city hotel tax revenue and projects convention center conferenceattendance and resultant citywide hotel spending for the coming years of 2020 to 2022.Convention center conference attendees are a small but important subset of business travelers tothe District of Columbia because of the essential role they play in the success of the 2.3 millionsquare-foot publicly financed WEWCC and the adjoining 14-story 1,175 room convention centerhotel. In 2017, over 455,000 conference attendees (the highest on record) visited the WEWCCand spent an estimated 116 million in hotel spending which in turn generated over 16.8 millionin hotel tax revenue (7.8 percent of total citywide hotel tax revenue). Total spending in the cityby WEWCC conference attendees could be at least 464 million according to the U.S. TravelAssociation. 5“How and Why Hotels Should Make Business Travelers Their Priority”. -make-business-travelers-their-priority2Research commissioned by fa-4df8-4210-a498db7a1b988f43Business Travel News ness-Travel-Spend-Through202142017 Visitor Statistics. Destination DC. /2017 washington dc visitor statistics - destination dc.pdf5These figures exclude air travel. root/document/Research FactSheet Domestic-Travel.pdf Travelers in 2018 tended to spend 25 percent of their travel budget on lodging. TheDistrict of Columbia is ranked the fifth most expensive city to visit for business travelers in les/9. Microsites/Corporate Travel Index/Corporate Travel Index 2019/US Diem/2 USPerDiemsRanked.pdf12

Compared to years 2005 to 2013, the average yearly conference attendance at the convention center foryears 2014 to 2016 was down 6.1 percent even as hotel spending rose by 24 percent. This result stemslargely from the growing role of medical conferences at the WEWCC. Whereas medicalconferences attendees used to account for 43.1 percent of all annual conference attendees at theconvention center, they now account for 50.9 percent of all attendees and are responsible for 68.8percent of hotel spending by all conference attendees. Given the national and local trends forbusiness travelers and the hospitality industry, the relatively new convention center hotel, and theexisting convention center conference bookings for years 2020 to 2022, it is possible that cityhotel spending by all conference attendees at the WEWCC can match or even exceed that of2017 in the coming years.II.DataThis study uses three types of data. First, hotel sales tax data for each hotel in the city are usedto assess the hotel activity on a monthly basis for years 2005 to 2016. Second, WEWCCconference booking data for large conferences for years 2005 to 2022 from Destination DC areused to relate all conference data (dates, number of conference days, number of registeredattendees, number of requested hotel rooms) to the monthly hotel tax data. 6 7 And third,citywide hotel data (citywide average daily hotel room rates (ADRs) and occupancy rates) fromSmith Travel Research (STR) are used for years 2005 to 2018. All hotel sales tax data, hotelroom rate data and hotel spending estimates are adjusted for inflation and put in terms of 2019dollars. The primary study period is from 2005 to 2016.Walter E. Washington Convention Center Conference DataFor years 2005-2016, 322,000 conference attendees on average visited the city annually8,contributing an estimated average of 7.1 million in hotel tax revenue per year. But, the annualaverage has been trending downward as shown by the red dotted trend line in Figure 1. Andwhile the number of annual conferences reported for the study period has also trended down, asshown by the red dotted trend line in Figure 2, STR reports that the citywide ADR 9 has beenincreasing at an average of 2.1 percent annually. For the study period, there has been 261conferences at the convention center and consequently 4 million hotel room nights in the cityrespectively. In 2016, there were 19 conferences and approximately 400,000 correspondinghotel room nights. These room nights accounted for a mere 4.6 percent of the city’s total roomdemand for the year (Figure 3) and an estimated 12.9 million in hotel tax revenue.For this study, conferences under analysis are of two types: citywide and miniwide. Citywide conferences are thoseheld at the WEWCC with 2,500 or more hotel rooms booked in the city on peak nights, and miniwide conferencesare those held at the WEWCC with 1,350-2,499 rooms hotel rooms booked in the city on peak nights.7Destination DC is a private, non-profit corporation with a membership of over 1,000 local businesses andorganizations that support the city travel and tourism sector. It is the contracting arm of Events DC. Theorganization is funded by a percentage of DC’s hotel occupancy tax, along with membership dues and co-operativemarketing fees. Destination DC serves as the lead organization to manage and market Washington, DC as a globalconvention, tourism and special events destination, with a special emphasis on the arts, cultural and historicalcommunities.8In terms of annual convention center conference attendees for this time period, there was a high of 389,402attendees in 2011 and a low of 228,400 in 2009.9STR, the industry standard for reporting average hotel rates, reports pre-tax room rates. These pre-tax rates wereincreased by 14.5% to reflect post-tax, nominal rates. Hereafter, ADRs will always represent post-tax levels.63

Figure 1. Yearly Convention Center Conference 0150,000100,00050,000-Figure 2. Annual Convention Center Conferences3025201510502005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016Figure 3. Summary of Conference Data4

Upon review of the conference data for all years, medical conferences are the only subset ofconferences that continues to be substantially large in all years in terms of reserved hotels roomnights. They alone account for almost half of all city hotel room nights attributable to conventioncenter attendees. Accordingly, all convention center conferences for all years were categorizedinto two broad categories: medical conferences and all other conferences.Monthly Hotel Room Rates & Conference SeasonalitySince the primary objective for this study is to estimate the annual hotel tax revenue for theDistrict of Columbia generated by conference attendees of the WEWCC for the study period, onekey variable is the operative hotel room rate for both medical conference hotel rooms and allother conference hotel rooms. The data available for this task is the citywide average monthlyhotel rate data collected by a monthly survey and provided by STR. Hotel rooms occupied byWEWCC conference attendees account for less than five percent of the city’s total demand forhotel rooms. However, these rooms are likely, in many cases, to command higher than citywideaverage room rates (even during peak months) because these conferences require thousands ofhotel rooms at very specific dates. And, many of these conferences tend to occur during thecity’s peak hotel periods (spring and fall) when hotel rates are highest and city hotel vacancyrates tend to be lowest.When we aggregate monthly citywide hotel room rate data to an annual level, it is likely that thismethodology significantly understates the annual average hotel rate paid by conference attendeesbecause it weights the months when the citywide hotel rates are lowest (December and January)equally with months when the citywide hotel rates are highest (e.g. April, May and October). Forexample, Octobers tend to account for 15 percent of all conference attendees on average, whileJanuarys only account for 1 percent (Figure 4). Because, we believe the applicable hotel roomrates for this analysis are higher than the citywide average monthly hotel rate provided by STR,we adjust the citywide average monthly hotel rate for the seasonality of conferences at theconvention center.Figure 4. Average Attendance by Month# of edicalOther5

On average, March, April, May, September, and October were the five priciest months in the cityin terms of hotel rates from 2005 to 2016 (Figure 5). STR’s ADRs for these five months was 258.46, compared to 212.19 for the rest of the year. 10Figure 5. Average STR Room Rates by Month (2005-2016) 300Nightly Room Rate 250 200 150 100 50 05 Priciest Months7 Cheapest MonthsOf the five months that tend to host the most medical conferences, four of them were among thefive priciest months (Table 1). In contrast, of the five months with the most all other conferences,only two of them were within the five priciest months. This illustrates that using citywidemonthly average hotel room rates may understate the economic and fiscal impact of conventioncenter conference attendees of which medical conference attendees are a major component.Table 1Top Months by Attendance(months matching up to the 5 priciest months are marked with an *4 of the 5 berMarch*2 of the 5 priciestmonthsThe Cherry Blossom season (the start of the spring tourist season), spring congressional legislative sessions (andthe resultant host of lobbyists, corporate executives and other special interests visiting the city), businessconferences and other spring business events tend to attract many visitors to the city in March, April, and May.Similarly, the fall congressional legislative sessions, business conferences and other fall business events help attractmany more visitors to the city in September and October. Using the unadjusted attendee statistics, 52.2 percent of allattendees tend to visit the city in March April, May, September and October, which includes 60.1 percent of allmedical attendees.106

For example, STR’s ADR on an annual basis for all years 2005 to 2016 is 231.47, and this iscalculated by averaging the monthly average citywide hotel rate for all 144 months. For thisstudy, STR’s ADRs are modified by weighting monthly ADRs by all WEWCC conferenceattendance and then separately by medical conference attendance data and the all otherconference data. In sum, weighted ADRs on an annual basis for all years 2005 to 2016 for allconferences attendees is 3.5 percent higher than the STR unweighted ADR, and the weightedADRs all years but only for all medical conferences attendees is 6.4 percent higher than the STRunweighted ADR (Table 2).Table 2. STR Room Rates Weighted by AttendanceTime PeriodSTR ADRWeighted ADR(All)Weighted ADR(Medical)Weighted ADR(Other)2005-2016 231.47 239.50( 3.5%) 246.27( 6.4%) 234.38( 1.3%)III.MethodologyThis study implements multivariate linear regression analysis to quantify the contribution ofconference attendees to the District’s monthly hotel tax revenue, with the model taking thefollowing form:𝑦𝑦𝑡𝑡 𝜂𝜂0 𝛼𝛼𝑐𝑐𝑡𝑡 𝛿𝛿𝑡𝑡 𝑚𝑚𝑚𝑚𝑚𝑚 12 𝛾𝛾𝑧𝑧𝑡𝑡 𝛽𝛽𝑤𝑤𝑡𝑡 𝜀𝜀𝑡𝑡where: 𝑦𝑦𝑡𝑡 represents hotel tax revenue during month t 𝜂𝜂0 is the intercept 𝛼𝛼 measures the average fiscal impact per attendee and 𝑐𝑐𝑡𝑡 represents the number ofattendees during month t δ𝑡𝑡 𝑚𝑚𝑚𝑚𝑚𝑚 12 is a seasonal adjustment for month t 𝛾𝛾 measures the average fiscal impact per point fluctuation in hotel industry-specific,regional CPI and 𝑧𝑧𝑡𝑡 represents hotel industry-specific, regional CPI for month t 𝛽𝛽 measures the average fiscal impact per point fluctuation in U.S. personal income and𝑤𝑤𝑡𝑡 represents U.S. personal income for month t 𝜀𝜀𝑡𝑡 is the error term for month tTo break down the effect of attendees as a whole into that of medical attendees and otherattendees, the model took a modified form of:𝑦𝑦𝑡𝑡 𝜂𝜂0 𝛼𝛼1 𝜃𝜃𝑡𝑡 𝛼𝛼2 𝜔𝜔𝑡𝑡 𝛿𝛿𝑡𝑡 𝑚𝑚𝑚𝑚𝑚𝑚 12 𝛾𝛾𝑧𝑧𝑡𝑡 𝛽𝛽𝑤𝑤𝑡𝑡 𝜀𝜀𝑡𝑡where: 𝛼𝛼1 𝜃𝜃𝑡𝑡 𝛼𝛼2 𝜔𝜔𝑡𝑡 replaces 𝛼𝛼𝑐𝑐𝑡𝑡 so thato 𝛼𝛼1 measures the average fiscal impact per medical attendee (𝜃𝜃𝑡𝑡 )o 𝛼𝛼2 measures the average fiscal impact per other attendee (𝜔𝜔𝑡𝑡 )7

IV.ResultsFor the set of regressions analyzing the citywide effects of the adjusted attendee data, thecoefficient of the key variable of interest measures the average fiscal impact per attendee. Itspecifies how many dollars an individual convention center attendee contributes to hotel taxrevenue per day. (Estimates for hotel spending by conference attendees was possible using monthlycitywide hotel sales tax data, which was only available for the years 2005 through 2016.) As shown inTable 3, an attendee contributed an average of 21.27 per day to hotel tax revenue for years 2005to 2016.Table 3. Regression Results: CitywideTime PeriodConference erCoefficient .98080.360.71910.9808Table 4 displays the estimated average per person per day hotel expenditure derived from theregression outputs shown in Table 3, as well as the estimated expenditures derived from theweighted STR averages shown in Table 2. 11Table 4. Hotel Expenditure: CitywideTime Period2005-20162014-2016Conference TypeAllMedicalAllMedicalEstimated Daily Hotel Expenditure (ADR)RegressionSTR Weighted 167.95 239.50 233.80 246.27 221.78 257.33 299.42 270.06Figure 6 shows the regression estimates for daily hotel expenditures (blue solid bars) from Table4 and compares them to estimated expenditures derived from the weighted STR averages (pinkhollow bars). 12 The regression estimates for medical attendees for 2005 to 2016 spent 39 percentmore than did all attendees on a hotel room per day on average ( 233.80 vs. 167.95), and from2014 to 2016 they spent 35 percent more ( 299.42 vs. 221.78). For the 2014 to 2016 timeperiod, all attendees and medical attendees spent 32 percent more ( 221.78 vs. 167.95) and 28percent more ( 299.42 vs. 233.80), respectively, on hotels than they did during the 2005 toExpenditure by “all other” attendees is not displayed, as their regression output is insignificant at the 𝛼𝛼 0.05significance level.12For the 2005-2016 period, the model (blue solid bars) explains on average 83 percent of the weighted STR data(pink hollow bars), and the explanatory power increases to an average of 99 percent when only considering the mostrecent years of 2014-2016. We conclude the regression models have substantial power in explaining the amount ofhotel tax revenue (and hotel expenditures) paid by convention center conferences attendees.118

2016 period. In essence, the figure shows the tendency for medical attendees to spend 35 percentor more than attendees from all other types of conferences at the convention center, and forattendees to spend more on a hotel room per day in years 2014 to 2016.Figure 6. Hotel Expenditure: CitywideEstimated total hotel spending for the 2005 to 2016 time period averaged 56.2 million annually( 8.1 million in hotel tax revenue), but 70.8 million ( 10.3 million in tax revenue) for the 2014to 2016 time period. 13 This increase in yearly hotel expenditure in the more recent years was inspite of average annual attendance decreasing to 319,020 in years 2014 to 2016 (Table 5).Table 5. Attendees and Hotel Expenditures for WEWCC Conferences 2005 to 2016Average Yearly l Other190,482156,547Total334,571319,020Average Yearly Hotel Expenditures( in millions)MedicalAll OtherTotal 33.7 22.5 56.2 48.7 22.1 70.8% Difference 12.8%-17.8%-4.6% 44.5%-1.8% 26.0%For medical conferences at the convention center, attendees spent an annual average of 48.7million ( 6.2 million in hotel tax revenue) in years 2014 to 2016. This annual average amountsuggests that hotel spending, as well as medical conference attendees, has been trended higher inmore recent years.13To calculate aggregate hotel spending by attendees, expenditure per attendee from the regression was multipliedby the number of attendees on record in the time period. For example, regression-calculated individual attendeeexpenditure of 167.95 for 2005-2016 was multiplied by a little over the 4 million attendees on record for the 12years, to calculate aggregate hotel spending generated by attendees as 674,295,065.9

Overall, medical attendees made up 43 percent of all attendees yet accounted for 60 percent ofhotel tax revenue/spending (Figure 7). But in the last three years of the study, medical attendanceand medical-specific revenue/spending, as shares, increased to 51 percent and 69 percent,respectively. This indicates that medical attendees are a growing share of total conferenceattendees at the convention center. And because the estimated average hotel room rate paid bythese attendees are on average at least 35 percent more than the estimated average hotel roomrate paid by all other conference attendees, the medical attendees account for 69 percent of allhotel expenditures paid by convention conference attendees in the latter years.Figure 7. Attendance and Hotel Tax Revenue/Spending as Shares: CitywideFor the study period, total attendance at the convention center trended downward in the recentyears of the study period on the order of 4.6 percent. But simultaneously, hotel spending by theseconference attendees is estimated to be 26 percent higher in recent years compared to earlieryears of the study period for two key reasons. First, while total convention center conferenceattendance is down 4.6 percent, medical conference attendance is up nearly 13 percent. Andsecond, medical conference attendees have been found to pay even higher hotel rates than allother convention center conference attendees.10

Looking AheadFigure 8 shows convention center conference attendance for time period: 2005 to 2019. The figure alsoshows the estimated attendance based on confirmed bookings for years 2020-2022. The linear trendlines are reflective of attendance trends for years 2005-2016 only. They are extrapolated out until2022 as points of reference. Table 6 summarizes 2017 to 2022 attendance relative to the trendlines.Figure 8. 2017-2022 Annual Attendance Relative to Trend500,000450,000# of 0100,00050,000-AllMedicalLinear (All)Linear (Medical)Table 6. 2017-2022 Annual Attendance Relative to TrendYearAllMedical2017 145k 62k2018 73k-35k2019*-25k-51kTotal 193k-24k% Difference from ProjectedTrend for 2017-201920.7%-6.0%2020* 9k 45k2021* 145k 62k2022* 73k-35kTotal 4k 45k% Difference from ProjectedTrend for 2020-20220.3%10.8%* Estimated attendance based on confirmed bookings11

For 2017 to 2019, overall attendance surpassed the study period trend by 20.7 percent, butmedical attendance fell 6 percent below trend. During this time period, the District saw fewer ofthe attendees who produced the greatest economic impacts for the city. But for years 2020 to2022, the outlook is positive in terms of attendance and economic and fiscal impact based onconfirmed conference bookings. Both total attendee and medical attendee figures are above the2005 to 2016 trend with medical attendees being significantly so. And while the number ofconfirmed conference attendees for 2022 may be concerning (Table 7), conference bookings arelikely to only increase given that there are more than 24 months to secure additional conferencebookings.Table 7. 2017-2022 Annual Hotel SpendingYearAnnualAttendanceEstimated AverageRoom RatedEstimatedHotel Spending2014-2016 a319,020 221.78 70,752,2562017455,486 b 254.50 115,920,0962018381,290 b 273.34 104,221,2252019281,458 c 293.19 82,520,5482020352,160 c 314.07 110,603,5822021311,995 c 336.44 104,968,1302022247,727 c 360.40 89,281,983a- Amounts are annual averagesb- actualc- projections based on confirmed bookingsd- Predicted rates account for the tax rate increases in 2018 to 14.8% and in 2019 to 14.95%.To estimate these attendees’ impact on the city’s hotel spending for years 2017 to 2019 and 2020to 2022, the coefficients of the primary regression model was used. 14 For years 2005 to 2016,there was an average of 7.0 medical conferences and 14.8 all other conferences per year.However, for years 2017 to 2022, there is an average of 7.2 medical conferences and 16.7 allother conferences per year. Additionally, in years 2005 to 2011 the average conference length(number of conference days per conference) was 3.4 days for all conferences and 3.6 days onlyfor medical conferences. However, for years 2012 to 2016 the average conference length was 3.8days for all conferences and 4.3 days. The current trend of slightly more conferences, longerconferences (Figure 9) and more attendees at the WEWCC is expected to increase the economicimpact to the city of future conferences.Estimates for hotel spending by conference attendees was possible using monthly citywide hotel sales tax data,which was only available for the years 2005 through 2016. To predict aggregate hotel spending, average individualhotel expenditure for each year from 2017-2022 was calculated and multiplied by attendance on record each of thoseyears. Daily expenditure per attendee was derived by applying a constant growth rate consistent with the model.From the model, daily expenditure per attendee increased 32 percent from 2005-2016 to 2014-2016. Taking 2010and 2015 as the respective centers of the two time periods, hotel expenditure increased 32 percent over 4.5 years.This equates to 7.12% yearly growth — the rate by which hotel rates were appraised each year.1412

Figure 9. Annual WEWCC Conferences Years 2005 to 2022Annual WEWCC Conferences353025201510502005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022AllV.MedicalConclusionThis study analyzed the economic and fiscal impacts of conference attendees of the Walter E.Washington Convention Center (WEWCC) for years 2005 to 2022. We consider convention centerconference attendees a small but important subset of business travelers to the District of Columbiabecause of the essential role they play in the success of the 2.3 million-square-foot publiclyfinanced WEWCC and the adjoining 14-story 1,175 room convention center hotel.Prior to 2017, the average yearly conference attendance at the convention center was trendingdown even as hotel spending by these fewer attendees was trending up. This result largely stemsfrom the growing role of medical conferences at the WEWCC. Whereas medical conferencesattendees used to account for 43.1 percent of all annual conference attendees at the conventioncenter, they now account for 50.9 percent of all attendees and are responsible for 68.8 percent ofhotel spending by conference attendees.In 2017, over 455,000 conference attendees (the highest on record) visited the WEWCC andspent an estimated 116 million in hotel spending which in turn generated over 16.8 million inhotel tax revenue (7.8 percent of total citywide hotel tax revenue). Given the national and localtrends for business travelers and the hospitality industry, the relatively new convention centerhotel, and the existing convention center conference bookings for years 2020 to 2022, it ispossible that city hotel spending (and other subsequent spending in the city) by all conferenceattendees in years 2020 to 2022 may more closely approach or even exceed that of 2017.13

to grow nationally from 1.3 trillion in 2016 to 1.7 trillion in 2021. 3. In 2017, the District of Columbia attracted 22.8 million visitors to the city. 4. Approximately, 41 percent of these visitors came to the city primarily for business and 59 percent came to the city primarily for leisure.