CLUJ-NAPOCA WITHOUT STUDENTS: AN ESTIMATION OF THE GAP IN THE CITY’S ECONOMY*1

Abstract In the new context of COVID-19, the authors assess the economic gap of the absence of students from Cluj-Napoca for one month due to the pandemic quarantine and isolation. The economic gap is presented in terms of estimated expenditures that did not occur, on an average monthly basis. The estimations are based on the survey data carried among Babeș-Bolyai University students that assessed students’ expenditure for the year of 2015 in Cluj-Napoca. Our results suggest that due to the absence of the student population from Cluj-Napoca, around 33.4 million Euros is the expected amount of loss per month through spending that does not occur.


Introduction
In March 2020, due to the ongoing pandemic of the coronavirus disease 2019 (COVID- 19), universities across Romania suspended face-to-face instruction. Most academic activities moved online, most important activities -teaching and students' interaction with the professors and the administrative staff -being carried on exclusively online. This unprecedented situation cancelled students' on-site presence at the university, and so the need for the students to live in the university city has also become optional.
The absence of students from the university city becomes visible where usually the ratio of students to city residents is high. In Romania, the largest university centers are Bucharest, Cluj-Napoca, Iași, and Timișoara, however the highest ratio of students to residents is in Cluj-Napoca (Figure 1). 20  In terms of economic impact, the visitors of students enrolled at Babeș-Bolyai University alone spent an estimated 7.5 million Euros in the city of Cluj-Napoca in the year 2015 (Chircă and Lazăr, 2019). And since only 9.3% of those visitors spent a night in a hotel or hostel during their stay in Cluj-Napoca, the type of tourism they perform remains unexplored and mostly untraceable in the official tourism data (Chircă and Lazăr, 2019).
The university has a significant impact on the economic life of its host city: a large number of people who are not residents shop locally, salaries are paid to academics, a significant number of students from other places decide to stay after graduating becoming city residents (Pastor, Pérez and de Guevara, 2013). In economic impact studies, these are known as short-term and demand-side effects. As a large-scale consumer of inputs such as labor, goods and services, and generator of outputs: skills, know-how, and local attractiveness, the university is a major factor in metropolitan economic development (Felsenstein, 1996(Felsenstein, , p. 1566. Therefore, estimating students' expenditure is subject to universities' local economic impact studies. It is also explored in studies on the economic struggle to obtain higher education or simply on the cost of living for students. The complexity of the economic impact of the universities has on one hand generated a diversity of economic impact studies, but on the other hand discouraged such attempts (Pellenbarg, 2005). Studies linking university spending to the local economy can be traced back to 1949 (Florax, 1992): Tully, A.M., 'The Economic Contribution of Rutgers University to the Welfare of the City of New Brunswick during the Fiscal Year 1947Year -1948 Compared to the results of a survey conducted in 2016 to measure the expenditure of the Babeș-Bolyai University (BBU) students in 2015, we can estimate a gap in the local economy of Cluj-Napoca caused by one month of coronavirus (COVID-19) pandemic quarantine in 2020, as the quarantine completely changed the ways in which the university functions. For that, the students' expenditure assessed for 2015 was adjusted with the annual inflation rate, with the number of BBU students enrolled in January 2020 (a total of 31,570), and with the total number of students studying in all universities in Cluj-Napoca.
The paper consists of three sections. The next section explores the conceptual aspects of some studies on economic impact of universities in which student expenditure issues are analyzed. The third section presents the methodology and the main results of the survey carried out among BBU students to assess their expenditure in the reference year 2015 as part of a broader study on the BBU impact on the local economy of Cluj-Napoca. Finally, the estimations of BBU student expenditure are adjusted to quantify total student expenditure in Cluj-Napoca for one month in 2020.

Theoretical framework
Universities leave marks on the local community, some of them tangible, others intangible, overt or covert as they are caused by factors both internal and external to the academic environment (Chircă and Lazăr, 2018). Warsh (2006) argues that it is enough to look at any map to observe that cities hosting universities have remained on top or reinvented themselves after decline, inferring that knowledge is a powerful factor of production. The implications of hosting a major university campus by a community, such as the economic value and contribution of the university and its impact on the socio-economic development of the community are well-discussed in the literature (Armstrong, 1993;Bleaney et al., 1992 Universities, the traditional providers of human resources and knowledge, are critical socio-economic development actors (Dzisah and Etzkowitz, 2008, p. 101). Etzkowitz (2008) places universities in the Triple Helix (university -industrygovernment) -an interaction that is key to innovation in an increasingly knowl-edge-based society. Universities are assigned the role of catalysts of local and regional economy development increasingly more often (Steinnes, 1987).
Considerable efforts to understand the contributions of universities to the functioning of regional economies can be found in regional studies focused on their roles (Drucker and Goldstein, 2007;Goldstein and Renault, 2004;Florax, 1992). The 2011 EU Guide 'Connecting Universities to Regional Growth' calls for an active engagement of public authorities to involve universities in cooperation with research centers, businesses and other civil society actors, in regional innovation strategies for smart specialization (Hahn and Vassiliou, 2011).
Due to the rapid growth of the higher education sector in the United States in the 1960s, the first official methodology to research the local impact of a university on income and employment appeared in 1971. Caffey and Isaacs (1971) designed it for the American Council on Education as a template methodology to quantify the shortterm economic impact on local economy of the demand-side effects -the economic impact in a certain year (or fiscal year) of the university and its related spending.
More than twenty years later, Goldstein, Maier and Lueger (1995) proposed eight distinctive university outputs that can have an impact on economic development. Such varied and complex outputs of the higher education led to numerous discussions either on improving the ACE methodology (Bluestone, 1993;Brown and Heaney, 1997;Elliott, Levin and Meisel, 1988), or on creating a better classification of the impact (Beck et al., 1995;Leslie and Slaughter, 1992), or even on proposing a different methodology, such as the Ryan short cut method (Ryan and Malgieri, 1981). Schubert and Kroll (2016) summarized the complexity of the university factors of impact as inputs, outputs, first order effects and second order effects. The clear distinction between long-term, knowledge-based supply-side, and the short-term, expenditure-based demand-side, of the same inputs is very helpful ( Figure 2) because they allow for a complex and nuanced assessment of the economic contribution of higher education institutions.
Siegfried, Sanderson and McHenry (2007) examined 138 studies covering 241 institutions and presented their reliability limitations: the selection of multipliers for indirect and induced impacts; the measuring methods they used, double counting of data, as well as unclear counterfactual scenarios and delimitation of the areas of impact.
More recently in Europe, Pastor, Pérez and de Guevara (2013) improved Caffrey and Isaacs' methodology by adding stochastic processes for assumptions with uncertainty. Thus, by using the requirements specified by Siegfried, Sanderson and McHenry (2007), the methodology based on Monte Carlo simulations allows for a clearer estimation of students' expenditure through a survey design that distinguishes between students based on where they live and between types of spending, and by filtering the results through counterfactual scenario.
In Romania, the doctoral research of the author, 'Estimating the impact of the university on the local economy. Case study: Babeș-Bolyai University in Cluj-Napoca' (unpublished), addressed the complete lack of efforts to estimate the role, importance and impact of any Romanian universities on the economies of the cities where they are located. It is the first and it remains the only research to date of this kind in Romania. The authors use the results of that survey as a starting point in assessing the economic gap in Cluj-Napoca, caused by the absence of students for one month, imposed by the quarantine.

Students' expenditure findings
The methodology to quantify student expenditure is an important part of a broader study to estimate the economic impact of universities on hosting cities. It includes two main aspects: identifying the students as impact generating agents, and quantifying and estimating the total student spending.
When estimating the economic impact of the universities, four impact agents are identified: 1. The university -its own spending, minus employee salaries; 2. University employees -their salaries paid by the university and their spending; 3. Students -their spending in Cluj-Napoca, throughout their studies; 4. Visitors to Cluj-Napoca city when their visits are related to the university. This category targets students' visitors such as friends and family members, as well as visitors applying to college admissions.

Research methods
The assessment of BBU student spending was carried out through a sociological inquiry in 2016. The survey comprise questions aimed at quantifying students' spending during their residence in Cluj-Napoca, during the admission period, and the spending of their visitors.

Studied population
In December 2015, BBU reported 41,690 students of all categories attending its courses. From that total we need to exclude: (1) the 3,340 students who were studying in 15 other cities where BBU has faculties; (2) 4,069 postgraduates and teacher training programs; and (3) 3,648 students enrolled in distance learning or part time learning programs. Thus, our general population was 29,754 BBU students in Cluj-Napoca. The BBU statistics office provided the above data, mentioning that a student registered in two faculties would be double counted.

Research instrument
For 74 days, starting February 24 th , 2016, a voluntary non-probabilistic survey in the form of an online questionnaire was disseminated among BBU students. The questionnaire consisted of four parts: (1) socio-demographic questions; (2) questions about direct student expenditure on the duration of studying in Cluj-Napoca -divided into 24 different types of spending; (3) questions about student expenditure during admission period; and (4) questions about spending generated by students' visitors (family and friends).

Data analysis procedures
Based on results from a total of 1,159 valid questionnaires, a multiple linear regression using the backward elimination method was applied to identify the most significant independent variables that induce the expenditure dimensions. The total BBU student expenditure was then calculated based on weighted average from regression results. To assess the total student expenditure in 2020, the total BBU student expenditure estimated for the year of 2015 has been adjusted with: the annual inflation rates; the number of BBU students in different categories as reported by BBU for January 2020; and with the total number of students in all Cluj-Napoca's universities.

Sample characteristics
We gathered 1,225 answers from BBU students from a convenience sample, 1,159 of them passing the validity checks 1 . The sample covers students in all 21 faculties of the BBU, each program (undergraduate's, master's, PhD) and year of study, students coming from Bucharest (the capital city) and 40 out of 41 Romanian counties (NUTS 3 regions). The sample also includes foreign students and students from the Republic of Moldova -the latter have a special status among students from other countries. Table 1 contains descriptive statistics on the structure of the general population versus our sample, as the survey was a voluntary and non-probabilistic one. * Cluj-Napoca area will cover the city of Cluj-Napoca and the surroundings within 20km (university standard in terms of being eligible to receive dormitory). Table 2 presents the regression results, the dependent variable in the total average student expenditure using monthly data. The dummy coding technique was used for seven independent variables: program (undergraduate's, master's, PhD), year of study, tuition fee, gender, urban or rural, students' origin (from the Cluj-Napoca area or other), and their type of accommodation in Cluj-Napoca.

BBU students' variables that induce the expenditure dimensions
Out of 15 dummies variables of the seven categories, five variables are significant at 1%, one is significant at 5% and another at 10%. The results suggest that a student at a master's program spends annually on average 844 Euros 2 more than an undergraduate student, ceteris paribus. If the student opts for a rent instead dormitory, the annual cost will rise by 1,732 Euros. p-value (F-stat) 0.000 *** -significant at 1%, ** -significant at 5%, * -significant at 10%,

BBU total student expenditure for 2015
Combining BBU available data with our findings, we then estimated the total amount of spending for BBU students in the 2015 year. The total amount was obtained by using different mean values of the most significant variables (program, with or without a tuition fee, type of accommodation) showed by the regressions above. Where possible, in case of undergraduate students, we kept a separate category for students coming from Cluj-Napoca area. In calculating the mean values, we treated the missing values for eight categories 3 as no expenditure. Tables 3 and 4 show the annual expenditure of the BBU students based on different spending categories of students, and they are also divided by categories of goods and services. The first division is between undergraduates, master students and PhD students, as obtained from the BBU official data. Of 21,546 undergraduates, 17.8% are from Cluj-Napoca area and 37.7% are paying a tuition fee. We divided them into two groups: (1) coming from Cluj-Napoca area and (2) coming from other areas. Then we further distinguished between students paying a tuition fee from those who do not pay a tuition fee, assuming the same percentage of students paying a tuition fee exist in both groups. Undergraduates who reside outside Cluj-Napoca opt for a bigger variety of accommodations, therefore we further divided them in groups according to the type of their accommodation. The BBU data shows that 4,463 undergraduate students were granted a place in the university dormitories. Our findings suggest that 9.3% of these students pay a tuition fee, so we divided them accordingly. The remaining two groups based on accommodation type were also divided in: (1) those who rent or own a house and (2) those who live in other type of accommodation (with relatives, private dormitories).
We divided the master students only by the type of accommodation, and we did not apply any distinctions in the group of the PhD students due to lack of data and the limited size of our sample. The total BBU student expenditure amounts to over 141 million Euros for 2015. Implicitly, Tables 3 and 4 show the industries that benefit more from student expenditure and the estimate value of the direct impact of student expenditure into those economic sectors. As an example, the BBU students spent more than 5 million Euros on cigarettes in 2015, while other 5.5 million were spent on medical services, medications and dentists. The divisions in Table 3 were created to reflect the primary spending categories. Thus, almost 60% of the student expenditure is presented in Table 3. The results suggest that primary necessity products (foods, drinks, and hygiene) take more than 27% of the students' budget, while housing costs represent 23%. Spending for meals in cafeterias, bars, cafes and restaurants presented in Table 4 amount to 10.6% and could also be placed in the food and drinks category, among the primary ones shown in Table 3.

The total expenditure of Cluj-Napoca students in a month
1. Starting from the total expenditure of the BBU students in 2015, the following steps were taken to estimate the total expenditure of Cluj-Napoca students in a month in 2020.     We extended the research to the whole population of students in Cluj-Napoca by looking at the ratio of BBU students in the total student population of the city. Figure  5 below shows, for different years, the proportion of students in Cluj-Napoca's universities. It becomes visible that the BBU students have accounted for more than half of the student population in Cluj-Napoca in recent years.
The results show that the absence of student spending for one month in Cluj-Napoca generates an estimated loss of 33,390,572 EUR. We argue that this amount remains underestimated as in the initial estimations we excluded the spending of the distance learning and part-time students, postgraduates, and teachers in training. Also, the estimated 7.5 million EUR spent by the BBU students' visitors in Cluj-Napoca in 2015 (Chircă and Lazăr, 2019) can account for a monthly loss of 625,000 EUR.
Another underestimated factor is the ratio of the foreign students -the category with the highest spending budget -in other universities than the BBU. For example, their ratio is much higher in universities like 'Iuliu Hatieganu' University of Medicine and Pharmacy. According to the National Institute of Statistics (private correspondence with the Executive Director of County Statistical Direction Cluj July 14, 2020), in the recent years, around 33% of the University of Medicine and Pharmacy students were foreigners in comparison with only 3.5% enrolled in the BBU. On average, the proportion of foreigners in the total student population in Cluj-Napoca is around 7%.

Conclusions
This study shows that one month of lockdown have halted the regular university life with a large scale impact on the local economy of the city, suggesting that the ongoing pandemic of the coronavirus disease 2019 (COVID-19) is impacting heavily all economic sectors and is changing the economic framework of the city. Estimating the social-economic impact of the pandemic is a complex and difficult task, but we can anticipate some of the possible directions and future problems. One of the major possible outcomes is that at the rate of 33.4 million EUR loss per month, the industries and businesses that benefit most in a university city will face severe economic difficulties in the short and middle term.
We can also infer that if the pandemic continues in the long term, the economic sectors of the city impacted indirectly will also suffer a downfall. Therefore, further extending the research into student expenditure to indirect and induced impacts can reveal some of the possible outcomes and risks, helping universities as well local communities to adjust their long-term strategies to the new realities. Thus, estimating the student expenditure becomes relevant not only in measuring the impact of  the university on the local economy, but also as important information that universities, as well as the local authorities, need to have for themselves. Another possible direction of research that can be developed from this type of study is assessing the students' quality of life in Romania in general, not only in the context of the pandemic. In that respect, the 2015 study revealed that the total of the BBU student expenditure estimated for that year was almost twice (183%) the total university budget. This suggests that the burden of obtaining higher education is carried mainly by the students' families while the students are enrolled in state universities in Romania.
Cluj-Napoca has the biggest student ratio per capita and hosts the largest university in Romania, which makes it the most relevant case study for this type of research. Moreover, a very interesting particularity of this university center is that only around 25% of the BBU students in need for housing -those who come from beyond the 20km radius from Cluj-Napoca -are accommodated in university dormitories. Thus, the BBU student population is a very dynamic one in terms of their spending: they are important actors on the real estate market, especially in terms of renting, and most of the student expenditure in the other categories occurs outside the campus, impacting a large spectrum of economic sectors. Therefore, if the city of Cluj-Napoca loses all its student population due to the ongoing pandemic or for any other reason, we can expect the local economy to shrink by around 33.4 million EUR per month. As a valuable indicator of the general economic impact of universities on communities, as a particularly powerful indicator of the possible short, mid and long-term consequences of the pandemic on the university centers, as well as a potential indicator of the higher education burden and of the level of underfunding of the education system in Romania, estimating student expenditure should be viewed as a crucial task by the higher education and government institutions.