Toys for Carpet Knights: Urban Travel Behaviour and ...Sykora, 2007). As such, urban sprawl soon became manifest at all levels in Romania, thereby triggering investigations at local,

Urbanism Toys for carpet knights: Urban travel behaviour and attitudesin the city of Cluj • C. Toşa, A. Mitrea

39

TOYS FOR CARPET KNIGHTS: URBAN TRAVELBEHAVIOUR AND ATTITUDES IN THE CITY OF CLUJ

Cristian TOŞAPh.D, School of Civil Engineering, Technical University of Cluj-Napoca,

Cluj, e-mail: [email protected]

Andrei MITREAPh.D, School of Urban Planning, ‘Ion Mincu’ University of Architecture

and Urban Planning, Bucharest, e-mail: [email protected]

Abstract. This is an exercise in studying travel behaviour among morningcommuters within the metropolitan area of Cluj, Romania, as well as theirattitudes towards urban transport. We have designed and conducted anon-line survey amongst two major occupational groups, namely privateand state sector employees. This survey concentrated upon the social andeconomic characteristics of these two commuter groups, while taking intoaccount existing structural and functional patterns within the builtenvironment. We have received answers that point to a noticeabledifference between private and state sector employees, but the overallresults are deeply worrying: Travel behaviour in Cluj is surprisinglyimmune to common sense. It is, however, deeply indebted to personalcomfort. Hence, considering current development patterns affectingRomanian cities, this research exercise critically examines travel behaviourand calls for a paradigm shift, which would hopefully lead to moreeffective transport policies.

Key words: commuting, occupational groups, urban transport, onlinesurvey, Romanian cities.

1. IntroductionIt is now more than a quarter centurysince Romania started its cumbersometransformation process towards amarket-oriented democracy (Pasti, 1995,2006; Gallagher, 2005, 2009). During thisentire period, Romanian cities were atthe heart of social and economic change.Thus, after an initial period of upheaval,they started experiencing almostcontinuous economic growth. One ofthe main consequences of this newlyfound financial prosperity was a

dramatic increase in car ownership,which, in turn, accelerated an incipienturban sprawl process that very soonbecame rampant (Toşa et al., 2015a;Soaita, 2013).

The city of Cluj has experienced thisphenomenon acutely over the pastdecade, with increasing real estatepressure, an intense blossoming ofeconomic activities and a gradualimprovement in transportationinfrastructure being the main factors

• Urbanism. Arhitectură. Construcţii • Vol. 9 • Nr. 1 • 2017

40

behind its fair share of urban sprawl(Suditu et al., 2010).

1.1. BackgroundHence, even though Romania’s accessionto the EU meant increasing theeffectiveness of the road transportationsystem on a national level (Robinson andStanciu, 2003; Robinson, 2006), the well-known ‘predict and provide’ policies,such as road widening and traffic flowoptimisation, have led to anecdotal trafficcongestion and associated environmentalexternalities, especially within cities (Becaand Cadar, 2013; Toşa et al., 2015b).

It is against this background that allmajor Romanian municipalities werecompelled to conduct surveys and todevelop sustainable urban mobility plans(SUMP), in order to be able to absorb EUfunds, thereby following the EuropeanUnion (EU) regulation concerning theUrban Mobility Package. The constituentmunicipalities of the Cluj-NapocaMetropolitan Area (CNMA) complied,and duly commissioned a comprehensiveSustainable Urban Mobility Plan (SUMP–CN) (Ove Arup and Partners, 2015).

This plan was released for public debatein November 2015. Unsurprisingly, itsmain priorities were accessibility, drivenby social equity, and environmentallysustainable economic efficiency.Glamorous wordings indeed, as befitsplanning documents issued these days.Perusing the list of recommendationswithin the plan, we became curiouswhether this was merely empty rhetoric,or whether the plan had addressedmatters related to mobility properly.

However, in order to decide upon therelevance of these recommendations, weneed a preliminary exercise. We thereforefind it timely to conduct a qualitative

analysis of travel behaviour patterns andto examine attitudes among differentcommuter groups. This exercise mightalso prove beneficial to fellowresearchers, who regard the case of Clujas a particularly rewarding field ofenquiry. We shall then return to a criticalappraisal of the SUMP–CNrecommendations in a future article.

This paper is structured as follows: Theremaining paragraphs within this sectioncontain a brief, albeit critical, historicalprofile of Cluj, which mainly concentratesupon matters related to the built-up area,the housing stock and the evolution ofthe public transport system. Althoughrarely mentioned, these few facts providevaluable insights into currentdevelopment trends, thereby renderingthem more transparent.

The ensuing section is methodological innature. It explains the research strategy,the data collection process and theanalysis proper. It is followed by Section3, which presents the results we haveobtained, in terms of modal share, theirspatial underpinnings, as well as thebehavioural profile of morningcommuters.

The paper concludes with an account ofthe findings and their policy implications,while at the same time sketching avenuesfor future research.

1.2. An historical overviewDuring socialist times, cities acrossCentral and Eastern Europe werecompact (Hirt, 2013). Romania hasproved to be no exception to this rule,with the Systematisation Law of 1974drastically curbing urban growth(Olteanu, 2013). In the particular case ofCluj, the built-up contours became fixedvia the Government Ordinance 120/1976.


41

They were to remain unchanged until theyear 2000.

However, after the fall of the communistregime, urban sprawl became thedominant trait of urban development.Against a background of rapid social andeconomic changes, this process has led tochanges within both land-use patternsand transportation system dynamics (cf.Sykora, 2007).

As such, urban sprawl soon becamemanifest at all levels in Romania, therebytriggering investigations at local, regionaland national levels (cf. Suditu et al., 2010).Regrettably though, a consistent andcoherent transportation systemsperspective is currently lacking, withquantitative and qualitative studies beingrather scarce (Boitor et al., 2013).

1.2.1. Land-use and housing stockAt the dawn of the market-orienteddemocracy in 1990, the city of Cluj wasone of the main industrial centres inRomania. It also ranked only second afterBucharest as an academic and medicalcentre.

Its urban form displayed the followingquantitative profile: The most visibleexponents of the socialist housing stockwere the four storey high blocks of flats,which amounted to some 70% of theentire housing stock. Furthermore, thebuilt-up area in Cluj comprised 3,987hectares, i.e. more than two times lessthan it is today. Within city limits, thepublic transport system was the maincommuting mode for everyone (Mitrea,1998). All in all, this was an image of ahighly compact city.

These figures are relevant in pointing tothe hallmarks of socialist planning.Hence, the consumption of housing

within socialist states was restricted inaccordance with national standards,which set a maximum living space perperson and household, by taking intoaccount the prevailing householdstructure and size. More importantly, ahousehold could only own one dwellingand had limited choice over the type andsize of the units, their location and theirquality (Tsenkova, 2008). This rationingprocess led to high savings within thebuilt-up area.

This is also the main psychologicalreason why, after 1990, residents weredriven by the desire to switch fromapartments to single-family homes.Hence, the city of Cluj experienced areal estate boom, which became rampantbetween the years of 2005 and 2008. It isprecisely this phenomenon that led to asharp increase in per capitaconsumption of built-up area, whichspiked from 127.9 to 292.9 m2/person.As a consequence, the built-up area roseto 9,508 ha in 2014.

Furthermore, due to the huge disparity insize and potential between the city coreand the other components of themetropolitan system, migration of bothpopulation and economic activities fromthe core towards the hinterlandssustained this on-going process of urbansprawl (cf. Soaita, 2013).

Against this background, Fig. 1 displaysthe population percentile changes withinthe metropolitan area of Cluj, withrespect to the first population census afterthe regime change of 1989. Although aconstant population growth ratecharacterised the city throughout theperiod covered by the censuses, anintensive migration soon becamemanifest towards the hinterlands, alongthe Someş River.


42

Fig. 1. Population distribution over time withinCNMA

This uncontrolled urban expansionhindered local authorities in providingappropriate infrastructure and vitalservices to the migrating population,such as suitable road network, effectivehealth services, and resilient publictransportation (Ove Arup and Partners,2015). Unfortunately, this state of affairscontinues to this day.

1.2.2. Transportation systemThe socialist public transport system inCluj was geared towards connecting thehigh-density residential areas with theindustrial platforms. Buses were theonly vehicles operating within thepublic transport system until theintroduction of trolleybuses in 1959.Increasing demand led to theintroduction of a high capacity tramlinein 1987.

Severe fuel rationings and low carownership entailed an extensive use ofthe public transport system. Hence, theelongated radial road network of Cluj

proved sufficient for the amount ofexisting traffic.

However, after the fall of the communistregime, the situation changed rapidly.Suffice it to say that the car fleet almostquadrupled between 1990 and 2016,rising from about 54,000 vehicles in 1990,to more than 200,000, in 2016. Against thebackground of an undersized growth ofthe road network, which only increasedfrom 342 to 403 km during the sameperiod, this remarkable rise in carownership stifled the carrying capacity ofthe road network, while demanding anever-increasing number of parkingfacilities.

Translated into an econometricperspective, Fig. 2 plots the overlapbetween the evolution in gross domesticproduct (GDP) within Cluj, expressed inpurchasing power standard units (PPS)per inhabitant, and the concomitantdeterioration in road network capacity. Itis interesting to observe that the GDP(PPS/inhabitant) has reachedapproximately 65% of the Europeanaverage value in 2013. Its growthtranslated most noticeably into anunparalleled rise in both motorised andnon-motorised traffic. Hence, it is notdifficult to understand why the citysuffers from chronic parking deficiency.

Fig. 2. Economic growth and mobility indicators


43

Parking on the road kerbside naturallyfollowed, which, in turn, stifled the roadnetwork carrying capacity even further.In response, authorities counteracted thelow levels of service through roadwidening and traffic flow optimisation, atthe cost of both actual and perceivedquality of pedestrian and bicycle spaces.

It is only recently that pedestrian friendlyzones and bike-sharing system have beentaken into account, albeit very timidly,thereby transforming them into a pie inthe sky for concerned activists (ELTIS,n.d.).

2. MethodologyTravel behaviour research is usuallyscarce within post-socialist societies,with most available studies beingexploratory in nature (cf. Bole andGabrovec, 2014; Burgmanis, 2012;Pucher, 1999; Grime and Duke, 1996). Awider array of empirical studiesconcentrating upon the travel-to-workbehaviour of government employees isavailable for some developing countries,motivated, no doubt, by the rapideconomic growth and majordemographic changes affecting them(Amoh-Gyimah and Aidoo, 2013).

Irrespective of their focus on developingor developed countries, authors usuallydraw their conclusions from field studies,which generally use disaggregated datathat reflects travel behaviour patternsamongst different individuals (Comminsand Nolan, 2011). This data then entersconditional logistic regression models,which reveal underlying factors affectingmodal choice when commuting.Sometimes, the insights gained arecomplemented by qualitative research, inthe guise of surveys directed at differentcategories of users (cf. Grdzelishvili andSathre, 2011; Jensen, 1999).

We have adhered to this type of researchmethodology only to a limited extent,mostly because of budgetary constraints.Nonetheless, we do hope that ourfindings will spark a renewed researcheffort amongst other scholars.

In the following sections, we shallpresent our research strategy, the datacollection process and a general outlineof the analysis proper. As we delveddeeper into the subject matter, wequickly encountered vast swaths ofunchartered territory. It is probably nomystery by now that Romanian dataabout mobility is notoriously scarce.When available, it is usually of a poorquality. Within such a context, researchefforts have mostly remained shallow.

We must therefore stress at this particularpoint that our paper amounts to littlemore than a preliminary study. We shallreturn to this matter in the final section ofthe paper.

2.1. Research strategyFrom the very beginning, we were veryinterested in gaining a deeperunderstanding of the following paradox:While congestion levels within the citybecome less and less tolerable, there seemto be more and more people in carscrisscrossing the city at every hour of theday. Hence, we wanted a clear answer tothe following research question: Whatmotivates people to use their privatelyowned cars for commuting, instead ofrelying more upon the public transportsystem?

At the core of our study was a web-based questionnaire, aimed at privateand government employees, whichcomprised questions concerning travelbehaviour, social and demographiccharacteristics, as well as attitudes


44

towards travel and the environment.The main idea behind the questionnairewas simple enough: To see howemployees commuted during themorning rush hours, and to grasp theirreasons for choosing a particular modeof transport.

In academic parlance, we began lookingcloser at modal share, travel behaviourand attitudes. In order to test whetherthese attitudes were rational, we checkedthem against unbiased measurements.Hence, we mapped distances betweeneach respondent’s home and the nearestpublic transport stop on his or hercommuting route towards the workplace,as well as the Euclidean distancesbetween home and work for eachrespondent. The results provedremarkable.

2.2. Data collection and analysisPrior to delivering the questionnaire, weformulated official written requests tomost of the administrative andgovernment agencies within the city. Wealso sent e-mails to the large privatecompanies working in the city. Whilemost of the government institutionsapproved the distribution of thequestionnaire via the employees’ e-mailaddress, few private companiesresponded to our request.

We then administered the questionnairevia the SurveyGizmo on-line platform. Wechose this platform because it has thebenefit of allowing users to access thequestionnaire through an Internetaddress link, in order to report data. Inaddition, considering the fact thatrespondents might lose patience whilecompleting the questionnaire, and alsodue to a lack of habit in reportingpersonal data, we appreciated its addedbenefit of allowing them to stop and to

continue reporting data at some othertime.

Hence, after a period of 60 days, wereceived a total of 804 responses. Thecompleteness rate was around 58%,meaning that the remaining respondentsterminated the interview before the datareporting process was complete.

At the end of the 60-day period, westarted filtering the data sets andanalysing them. In addition, we preparedreported home and workplace addressesfor geocoding.

The instruments of choice for dataanalysis were the R programmingenvironment and the Libre Office Suite, aswe strongly favour an open source andfree working environment. Forvisualisation purposes, we used theQuantum Geographic Information System(QGIS) application. We also used the on-line Global Positioning System (GPS)visualizer tool, with a MapQuestApplication Programming Interface (API)developer key, in order to obtaingeographic coordinates for home andwork locations.

3. Results and discussionAfter cleaning and filtering the raw dataset, we were left with a final sample of200 state sector employees and 126private company employees. The mainreason for such a reduced sample isunder-reporting. Hence, among thecompleted questionnaires were 140respondents for which we lacked eitherhome or work addresses, or other socialand economic variables, like age, gender,car ownership or income, which weneeded in our analysis.

The general traits of the sample aredisplayed in Table 1. We can distinguish


45

between social characteristics, such asoccupational status, income intervals andpossession of a driver’s licence, anddemographic characteristics, such asgender, age and marital status. We haveconcentrated upon these indicators, as webelieve that they are relevant andrepresentative for our research. Thefollowing paragraphs will hopefullyshow why this is indeed so.

Looking at Table 1, we can observe thefollowing: State sector employees seem toexhibit a higher share of women, and are,on average, of a higher age. They do,however, have an overall lower income.In contrast, private company employeesare exhibiting almost opposite traits.Peculiar as they might seem at firstglance, these patterns agree with thelabour market composition and itsevolution across post-socialist states(Sootla and Lääne, 2013).

Unfortunately, it seems that theadvantages of male employment stillremain manifest, after having beenfostered by the socialist regimes undervarious guises (cf. Prodan andManolescu, 2006; Silova and Magno,2004). There is also an additionalsocialist legacy worth mentioning here:The share of state sector employees with

a higher age average implies almostinsignificant labour dynamics within thissector.

Moreover, the age structure within oursample suggests that population ageingwill definitely have an impact upon theperformance the transport systems, aswell as upon their future developments(Coughlin and Tompkins, 2009; Ryserand Halseth, 2012; EC, 2014).

Focusing upon the drivers within oursample, we learn from government datafor the year 2009 that, out of 6.063 milliondrivers, approximately 70.3% are men.Now, in the county of Cluj, where aboutone third of the population owns adriver’s licence, women account for46.3% of the drivers. Our data shows alower percentage of persons having adriver’s licence among state sectoremployees, probably due to the highershare of women within the sample.

In contrast, highly paid professionalsworking in the private sector exhibithigher shares of a driver’s licenceownership. At the same time, theirhouseholds reveal a higher car ownershiprate, 1.15 cars per household, comparedto 1.09 cars per household in the case oftheir state employee counterparts.

Table 1. Socio-demographic characteristics of the sampleCharacteristic Details State (%) Private (%)

Occupation Respondents 61.35 38.65Gender Female 74.00 59.52

<25 3.00 15.8726…45 47.50 78.5745…65 43.50 4.76Age intervals

>65 6.00 0.79Marital Status Married 67.50 50.00Drivers licence Car 67.00 83.33

900-1500 29.00 10.321501-2500 48.50 14.292501-3000 0.00 8.73Income intervals (in RON)

>3000 22.50 66.67


46

In short, it is plain to see that privatesector employees are generally wealthierand that, consequently, they own, onaverage, more than one car perhousehold, thereby needing moredriver’s licences. This much is fairlyevident from simply looking at Table 1.Things become more interesting,however, when studying modal share.

3.1. Modal shareWe then looked at modal share from ageneral perspective, i.e. taking intoaccount the following three majortransportation categories: Non-motorised,including walking and cycling, privatemotor vehicles and public transport.

For the two occupational groupsmentioned above, as well as for aweighted average for the whole sample,commuting patterns are revealed in Fig.3. Here, we can notice that privateemployees are more active and moremobile, even though we have no data onthe complete set of daily activities thatthey might perform, besides theircommuting pattern.

Fig. 3. Sample modal share

We further assumed that modal share isindicative for each respondent’scommuting history. Hence, Figs. 4a and4b show the time spans during whichrespondents used their preferredcommuting mode. This we calculated by

transforming modal share preferences foreach time period into per cents, therebyobtaining the relative modal share foreach period, for both state and privatesector employees. Note that summing upthe percentages for the three transportmodes used by all employees during oneperiod always returns a value of 100%.

Fig. 4a. Historical mode use (%) - state

Fig. 4b. Historical mode use (%) - private

Now let us consider the modal share overtime: For state sector employees,commuting by public transport seemed tobe the norm until ten years or so ago,when the use of private cars caught up. Itis interesting to note, however, thatduring the year of the survey (2015),public transport surpassed the use ofprivately owned cars anew. We cannotpinpoint the exact reasons for this shift in


47

preference, as prices for petrol remainedrelatively constant during 2015.

There are, however, two possibleexplanations for this particular state ofaffairs: It is either an improvement inpublic transport quality, or the fact thatgrowing congestion affects fuel efficiencyas well as commuting quality. If we takeinto account that state sector employeesare generally older and paid less thantheir privately employed counterparts,then savings might be more likely toaffect modal choice.

When it comes to private sectoremployees, however, the picture is farclearer: The privately owned car is thenorm. Again, if we take a closer look atthe year when the survey was conducted(2015), it is the public transport that nowis, hopefully, beginning to catch up.However, it is still early days for us toascertain whether this is a consistenttrend or merely an accident. If it is indeednot a mere accident, then we might havethe chance, for the first time in Cluj, tocontradict the old adage that individualcommuting habits hardly change overtime, unless additional taxation applies.

3.2. Spatial characteristicsSocial and economic variables, spatialcharacteristics, and modal featuresrepresent exogenous variables that affectmodal choice for commuting. We notedearlier that there are important shares ofstate sector employees using primarilynon-motorised and public transport forcommuting, in contrast to private sectoremployees, who use their own cars.

In this section, we explore the spatialpatterns of commuting, such as Euclidiandistance from home to work, as well asthe proximity of public transport stopsnear the origin and destination points for

morning commuters. Initially, we hadreason to believe that the modal splitdifferences observed among state andprivate sector employees are caused bysuch spatial considerations.

As mentioned earlier, we geocoded bothhome and work location for eachrespondent. It was therefore possible tocalculate the walking distance betweeneach home and the nearest bus, trolleybusor tram stop, which enabled eachcommuter to reach his or her workdestination. For this purpose, we haveemployed the ‘Directions’ feature of theGoogle Maps platform. The final resultsare displayed in Fig. 5, and they arestriking.

Fig. 5. Walking time to PT stop

More than half of the respondents fromeach occupational group within thesample live within less than sevenminutes away from their relevant publictransport stop. Furthermore, it is plain tosee that variations in walking time to thenearest stop between the two categoriesare insignificant. There is thereforeabsolutely no reason to believe thatdistance towards the nearest bus stop isaffecting modal choice for commuting.


48

Another measure to influence modechoice for commuting is the relativespatial location of home and work(Limtanakool et al., 2006; Cervero, 2002).The reason for using such a measure isthat we are currently unaware of thedistances travelled by employees in eachof the available modes. To furthercomplicate matters, we also lackinformation concerning the level ofservice on each of the transportationnetwork links, which, in turn, mightinfluence the amount of travel time spenton commuting.

Fig. 6 thus plots the Euclidean distancebetween home and workplace for eachoccupational group. We obtained thesedistances through QGIS, by using thegeocoded addresses for home andworkplaces. Essentially, private sectoremployees live farther away from theirworkplace, most probably due to theresidential migration of young people (cf.Beige and Axhausen, 2008). Thissupposition is supported by theexpanding housing stock within themetropolitan area of Cluj, as revealed bythe time dependent spatial distribution ofpopulation in Fig. 1.

Fig. 6. Distance between home and work location

However, most of these variations intravel to work distance are small betweenthe two occupational groups. Moreover,after reaching the 5-km limit, they tend todisappear altogether. We can thereforeascertain that travel-to-work distanceshardly influence modal choice. In short,neither the distance to the nearest stop,nor the overall distance to the workplaceis at stake here.

Pondering upon the meaning of thesefindings, we were able to find twointerpretations for them: Firstly, if wetake into consideration the historicdevelopment of Cluj, we encounterspatial concentrations of publicinstitutions within, or quite near, the citycentre. Following from this observation,we can infer that state sector employeesenjoy the benefits of overlapping publictransport routes, and, implicitly, a higheraccessibility of their workplaces. Andsecondly, one might infer that new realestate developments offer a moregenerous parking supply, therebyencouraging the use of a privately ownedcar.

Nonetheless, one should not lose sight ofthe fact that the city of Cluj is a relativelyaccessible city, displaying highconcentrations of both workplaces andamenities within a densely packed area.We are therefore far more inclined toattribute a private sector employee’spropensity towards favouring theprivately owned car when commuting toa question of social status.

In simple terms, we believe that, ingeneral, the wealthier an employee is, themore likely it is for him or her to favourhis or her own car over the publictransport system. However, given thecurrent congestion levels within the citycentre, as well as the daily traffic jams


49

facing commuters travelling to thesuburbs of Gilău and Floreşti, in the west,as well those travelling to Apahida andDezmir, in the east, we consider suchattitudes to be highly questionable, to putit mildly.

3.3. AttitudesIt is now time to consider the attitudesproper. Hartgen (1974) stated that themode choice mechanism is determinedprimarily by situational constraints, suchas car ownership and income, and onlysecondarily by the quality of alternativemodes. In this section, we analyse thecommuters’ attitudes towards the modesavailable within the choice set, byexamining the characteristics thatrespondents value.

The respondents were asked whichfeatures they value most when choosingtheir commuting mode. Data wasreported on comfort, commuting timeand cost. We aggregated state and privatesector employees on each of their chosencommuting mode. By considering Figs. 7aand 7b, we notice the propensity of carcommuters for appreciating comfort,irrespective of their particularoccupational group.

Differences in attitude arise, however,when analysing commuting time andassociated cost. As non-motorisedcommuting is hardly influenced by thelevel of service within the transportationnetwork, it is more relevant to comparethe public transport system withprivately owned cars.

On the one hand, state sector employeesassociate public transport mainly withlower costs and, to a lesser degree, withshorter travel times. Thus, it seems likelythat state sector employees associatelower costs primarily with transportation

expenditure, which is often a consistentpart of the total income of an individual,especially if wages are relatively low (cf.Sugiarto et al., 2014).

The shorter travel times within this sectorare explained by the average of 1.21transfers needed to reach the workplace,compared to the average of 1.52 transfersneeded by private sector employees. Toput this somewhat artificial figure intocontext, let us state that 79.5% of stateemployees commute to work without anytransfer, compared to 55.6% of privatecompany employees. However, thisproportion is reversed in the case ofcommuters who need one transfer, with20% of state sector employees needing asingle transfer, compared to 38.1% ofprivate sector employees.

These figures buttress our earlierobservation that state sector employeeshave the benefit of working in or near thecity centre, where the degree ofaccessibility is somewhat higher.Therefore, from a strictly temporalperspective, it is wiser for a state sectoremployee to commute by publictransport, due to the connectivity of mostof the public transport lines with the citycentre.

Fig. 7a. State employees commuting attitudes

Moving on, we observe that privatesector employees value first and foremost


50

comfort and shorter travel times. Traveltimes, in this care, are partly explained bythe number of transfers needed to reachthe workplace. More transfers implyhigher prices for the monthly publictransport pass, which private employeesneed in order to commute.

Fig. 7b. Private company employees commutingattitudes

One would expect, however, that privatesector employees would value comfortmore highly. In fact, only about a quarterof respondents within this category chosecomfort as the main reason forcommuting by car. In contrast, more thana third of the state sector employeesmade the same choice.

Surprisingly enough, we believe that thisobservation is linked to the modal sharediscussed earlier, especially if weconsider that over the past year, few ofthe state sector employees used their carfor commuting. Hence, decreasingcomfort might actually be related togrowing congestion levels, whichinherently diminishes the comfort of thedriver, by causing longer travel times andimplicit stress.

Unfortunately, it seems that thisdiminishing comfort is not strong enoughyet, in order to tip the balance in favourof the public transport system. It might

however explain the growing number ofpublic transport users amongst theprivate sector employees during the pastyear (cf. Fig. 4b). Furthermore, it is clearthat both categories have become moresensitive to the transportation systemperformance, as it directly affectsdeparture for work times, as well as theoverall duration of commuting (cf.Pendyala et al., 2002).

4. ConclusionsThis was a first exercise in revealingtravel behaviour and attitudes ofmorning commuters in city of Cluj. Inorder to gain a comparative perspective,we studied both state and private sectoremployees. By identifying modal splitdifferences, we further explored theparticularities of each of the occupationalgroups.

An historical overview of commutingmodes revealed an extensive, but slightlydecreasing, use of public transportamong state sector employees. At thesame time, the use of the privately ownedcar amongst private sector employees isincreasing. This is the first message of ourpaper.

We then plotted the accessibility of publictransport stops against each respondent’sresidence, and we concluded that morethan half of respondents, irrespective oftheir occupational group, lived less thanseven minutes away from the neareststop servicing public transport towardstheir workplace. In other words, thedistance between home and the neareststop does not influence modal sharesignificantly. The same applies to the(Euclidean) distance between home andwork. Hence, distance proved not to bean issue with using the public transportsystem more intensively. This is oursecond message.


51

However, in order to be completelyhonest about the limitations of our study,we did not have any informationwhatsoever concerning the performanceof the transportation network. It remainsfor future research to delve deeper intothis matter. For the time being, suffice itto say that direct observation of variousroutes showed us that the publictransport is currently more thanacceptable in quality, albeit with ampleroom for improvement.

Attitudes seem to be the key to betterexplain the prevailing modal share. Weconcentrated upon questions related tofeatures such as comfort, travel time andtravel cost for each mode of commuting.The results were somewhat surprisingfrom a social and environmentalperspective, although completely in linewith our intuitive reasoning: Hence,while state sector employees value thelower travel costs and the shorter traveltimes offered by the public transportsystem, private sector employees seem tofavour comfort above all else.

This is our third and final message: Itseems that private sector employees,although being younger on average andearning higher wages, compared to theirstate sector counterparts, are less prone touse the public transport system, andhence contribute significantly to currentcongestion and pollution problems. Thisis a worrying conclusion, but we believeit should always be kept at hand whendrafting public policies.

Again, to qualify the previous conclusion,we need to stress that, indeed, some ofthe new real estate developments, whichmainly house office spaces, do not offeran easy connection to the public transportnetwork. However, this is not the case forthe vast majority of workplaces.

Furthermore, we were able to unearth aninteresting trend, which evolved solelyover the past year: Private sectoremployees have begun to use publictransport more intensively. Although stillsceptical in this respect, we are curious tosee whether this becomes a substantialtrend aimed at toppling car use as themost favoured mode of commuting. Thereasons behind this incipient shift inperspective are still unclear to us for thetime being, but we believe that theymight have to do with intolerablecongestion levels.

To sum up, the imperfect and oftenrigged transition to the market economyhas initially caused confusion, andsubsequently brought about an unevenand an unfair urban development withinthe city and its metropolitan area. Hence,pedestrian as it might seem at first glance,we need to constantly remind ourselvesthat modern-day cities are the result ofyesterday’s planning decisions.Generally, most of these decisions haveproved to be unwise and unprofessional.As one of their main consequences, adominating logic obedient to the motorcar has pervaded planning decisions. Theeffects of such planning practices are stillwith us today.

In short, the intention of this researchpaper was to question that status quo. Asbefits a preliminary study such as thispresent one, we have also drafted a list ofopen questions, aimed at guiding furtheravenues of research:

Firstly, we have arrived at the conclusionthat relations between modal choice, thetransportation network characteristicsand the functional as well as spatialarchitecture of the city are poorlyunderstood in the case of Cluj. This lackof understanding necessarily translates,


52

in our opinion, into poorly designed andpoorly detailed transport policies.Furthermore, until these matters are laidto rest, targeted policies, such as mobilitymanagement for private companies(telecommuting), parking management orcongestion charging in downtown areas,as well as public transport networkperformance assessment per currentneeds and commuting patterns remainvoid of strength and applicability.

Secondly, the complexity of mode choicebehaviour requires further studies indiscrete choice analysis, in order to obtaina comprehensive review of the relativeinfluence of exogenous variables on thestudied outcomes. In simpler terms, weacknowledge the clear limitations of ourstudy, but we intend to remedy them at alater stage, as we get hold of more data.Hence, once these connections are foundand explored, policymakers could thenconcentrate solely upon designinginclusive accessibility and equity policiesfor the future, as well as transport andland-use policies to encouragesustainable commuting behaviour and topromote compact urban forms inemerging urban areas. Until then, currentpolicy recommendations remain largelyempty rhetoric.

Thirdly, considering the findings of thisstudy, we may conclude that thedevelopment patterns developing in Clujhave brought about social and economicsegregation among commuters. Theextensive use of privately owned cars hasgrave repercussions upon the level ofservice and perceived quality of othertransport modes. Undoubtedly, this is aquestion of responsible citizenship, bothsocial and environmental, which,unfortunately, is largely absent fromwhat we gather.

Moreover, we feel that the municipality,as well as the academic environment, forthat matter, is lacking a responsive andprofessional team of researchers andpolicy makers. We therefore stronglybelieve that more profound and systemicaction is urgently needed in thisdirection. This paper can be read as anincipient commitment to such anendeavour.

AcknowledgementsWe are highly indebted to ProfessorVasile Mitrea for his generous commentsand suggestions pertaining to the urbanhistory of Cluj. They helped us frame ourarguments far clearer.

REFERENCES

Amoh-Gyimah R., Aidoo E. N. (2013), Mode ofTransport to Work by GovernmentEmployees in the Kumasi Metropolis,Ghana, Journal of Transport Geography31: 35-43.

Beca M., Cadar R. (2013), An Assessment Modelof Urban Noise Performed ThroughSIMITR Subsystem, Romanian Journalof Acoustics and Vibration 10(2): 115-118.

Beige S., Axhausen K. W. (2012),Interdependencies between Turning Pointsin Life and Long-term Mobility Decisions,Transportation 39(4): 857-872.

Boitor M., Antov D., Iliescu M., Antso I., Mäe R.(2013), Sustainable Urban TransportPlanning, Romanian Journal ofTransport Infrastructure 2(1): 39–50

Bole D., Gabrovec M. (2014), Geography of DailyMobilities in Post-Socialist EuropeanCountries: Evidence from Slovenia, in:Burrell K., Hörschelmann K. (Editor),Mobilities in Socialist and Post-SocialistStates, Palgrave Macmillan UK, pp. 217-239.

Burgmanis G. (2012), Children’s Everyday SchoolTravel and Mode Choice in a Post SocialistCity: The Case of Riga, Latvia,International Proceedings of EconomicsDevelopment and Research 31: 47-53.

Cervero R. (2002), Built Environments and ModeChoice: Toward a Normative Framework,


53

Transportation Research Part D:Transport and Environment 7(4): 265-284.

Commins N., Nolan A. (2011), The Determinantsof Mode of Transport to Work in the GreaterDublin Area, Transport Policy 18(1): 259-268.

Coughlin J. F., Tompkins C. J. (2009),Demographics, Destiny, and Anticipatingthe Future of the Transportation System,Public Works Management & Policy13(4): 284-287.

ELTIS. (n.d.). The Critical Mass Expands the BikeCulture in Bucharest (Romania),http://www.eltis.org/discover/news/critical-mass-expands-bike-culture-bucharest-romania-0

European Commission (2014), The 2015 AgeingReport: Underlying Assumptions andProjection Methodologies, http://ec.europa.eu/economy_finance/publications/european_economy/2014/pdf/ee8_en.pdf

Gallagher T. (2005), Theft of a Nation. Romaniasince Communism, C. Hurst & Co.,London, United Kingdom.

Gallagher T. (2009), Romania and the EuropeanUnion — How the Weak Vanquished theStrong, Manchester University Press,Manchester, United Kingdom.

Grdzelishvili I., Sathre R. (2011), Understandingthe Urban Travel Attitudes and Behavior ofTbilisi residents, Transport policy 18(1):38-45.

Grime K., Duke V. (1996), Urban TransportChoice: A Post-Communist Dilemma,Regional Studies 30(8): 791-795.

Hartgen D. T. (1974), Attitudinal and SituationalVariables Influencing Urban Mode Choice:Some Empirical Findings, Transportation3(4): 377-392.

Hirt S. (2013), Whatever Happened to the (post)Socialist City?, Cities 32: S29-S38.

Jensen M. (1999), Passion and Heart inTransport — A Sociological Analysis onTransport behaviour, Transport Policy6(1): 19-33.

Limtanakool N., Dijst M., Schwanen T. (2006),The Influence of SocioeconomicCharacteristics, Land Use and Travel TimeConsiderations on Mode Choice forMedium-and Longer-Distance Trips,Journal of Transport Geography 14(5):327-341.

Mitrea V. (1998), Housing in Cluj until the Year2006 [in Romanian]. DoctoralDissertation, ‘Ion Mincu’ University of

Architecture and Urban Planning,Bucharest, Romania.

Sykora L. (2006), Urban Development, Policy andPlanning in the Czech Republic andPrague, in: Altrock U., Guntner S.,Huning S., Peters D., Burrell K.,Hörschelmann K. (Eds.), Spatial Planningand Urban Development in the New EUMember States. From Adjustment toReinvention, Ashgate PublishingLimited, UK, pp. 113-140.

Sootla G., Lääne S. (2013), Public AdministrationDevelopments and Practices in Estonia, in:Liebert S., Condrey S. E., Goncharov D.(Eds.), Public Administration in Post-Communist Countries: Former SovietUnion, Central and Eastern Europe, andMongolia, CRC Press, pp. 161-190.

Olteanu M. M. (2013), Legislation and UrbanPlanning in Socialist Romania [inRomanian], in: Soare F. S. (Editor),Politics and Society in the Ceauşescu Era,Editura Polirom, Iaşi, Romania, pp. 17-33.

Ove Arup & Partners (2016), Cluj-NapocaSustainable Urban Mobility Plan 2016 –2030,http://www.primariaclujnapoca.ro/userfiles/files/Plan%20mobilitate%20Cluj%20Napoca.pdf

Pasti V. (1995), Romania in Transition. The Fallinto the Future [in Romanian], EdituraNemira, Bucureşti, Romania.

Pasti V. (2006), The New Romanian Capitalism [inRomanian], Editura Polirom, Iaşi,Romania.

Pendyala R. M., Yamamoto T., Kitamura R.(2002), On the Formulation of Time-SpacePrisms to Model Constraints on PersonalActivity-Travel Engagement,Transportation 29(1): 73-94.

Pucher J. (1999), The Transformation of UrbanTransport in the Czech Republic, 1988–1998, Transport Policy 6(4): 225-236.

Prodan A., Manolescu I. (2006), GenderDiscrimination in Romania, Journal ofOrganizational Change Management19(6): 766 – 771.

Robinson R. (2006), A Perspective on Road SectorRestructuring in Developing andTransitional Countries. PublicAdministration and Development 26(3):265-278.

Robinson R., Stanciu M. (2003), Romanian Modelfor Road Sector Restructuring,Transportation Research Record 1848:10-18.


54

Ryser L., Halseth G. (2012), Resolving MobilityConstraints Impeding Rural Seniors’Access to Regionalized Services, Journal ofAging & Social Policy 24(3): 328-344.

Silova I., Magno C. (2004), Gender EquityUnmasked: Democracy, Gender, andEducation in Central/Southeastern Europeand the Former Soviet Union, ComparativeEducation Review 48(4): 417-442.

Soaita A. M. (2013), Romanian Suburban Housing:Home Improvement through Owner-building, Urban Studies 50(10): 2084-2101.

Suditu B., Ginavar A., Muică A., Iordăchescu C.,Vârdol A., Ghinea B. (2010), UrbanSprawl Characteristics and Typologies inRomania, Human Geographies —Journal of Studies and Research inHuman Geography 4(2): 79-87.

Sugiarto S., Miwa T., Sato H., Morikawa T.(2014), Transportation ExpenditureFrontier Models in Jakarta MetropolitanArea, Procedia - Social and BehavioralSciences 138: 148-158.

Toşa C., Miwa T., Morikawa T. (2015a),Modelling and Forecasting CarOwnership in Romania’s Counties UsingBass Diffusion Model, in: EuropeanTransport Conference 2015, AET.

Toşa C., Antov D., Köllő G., Rõuk H.,Rannala M. (2015b). A Methodology forModelling Traffic Related Emissions inSuburban Areas, Transport 30(1): 80-87.

Tsenkova S. (2009), Housing Policy Reforms inPost-Socialist Europe: Lost in Transition,Physica-Verlag, Heidelberg, Germany.

Received: 12 December 2016 • Revised: 9 January 2017 • Accepted: 16 January 2017

Article distributed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License (CC BY-NC-ND)

Documents

Toys for Carpet Knights: Urban Travel Behaviour and ...Sykora, 2007). As such, urban sprawl soon became manifest at all levels in Romania, thereby triggering investigations at local,