Transport et problèmes environnementaux

Présentation au sujet: "Transport et problèmes environnementaux"— Transcription de la présentation:

1 Transport et problèmes environnementaux
Le travail à la maison et le télé-travail: vers une réduction de la demande en transport dans la région de Québec? Marius Thériault (ÉSAD) et Marie-Hélène Vandersmissen (Département de géographie) Centre de recherche en aménagement et développement Transport et problèmes environnementaux Que dit la recherche? Journée d’étude du CDAT, CRAD et l’IHQEDS Université Laval, 6 octobre 2006

2 Introduction Travail à la maison et télé-travail: travail à la maison plutôt qu’au lieu habituel avec les technologies d’information et de moyens de télécommunications Stratégie de réduction de la durée et du coût des déplacements, de la congestion routière, de la consommation d’énergie et de la pollution de l’air Sur le plan individuel, réduction du nombre de véhicules-kilomètres parcourus Réduction des coûts immobiliers pour les entreprises, amélioration de la qualité de vie pour les travailleurs et de la conciliation travail-famille - For the purpose of this study, telecommuting is defined as working at home rather than at the regular workplace during at least one day per period of two weeks. - Telecommuting has long been identified as a strategy for reducing travel time and cost, road congestion, energy consumption, and air pollution. Mokhtarian & Varma (1998) have established the benefits of telecommuting for persons, mainly in terms of reduced vehicle-miles travelled. - Recent literature reviews of telecommuting research were published by Shin et al. (2000) and Bailey & Kurland (2002). Telecommuting was developing in North- America during the last decades of the 20th century. It has been lauded in many different ways: decreasing real-estate costs for organizations, improving quality of life for workers, facilitating work-family balance, reducing air pollution and traffic congestion. At the end of the 20th century, there were more than 11 million telecommuters in the U.S. (Bailey & Kurland, 2002).

3 Mais impacts à long-terme à l’échelle du système urbain moins évidents
Télé-travail actuellement peu pratiqué dans les villes nord-américaines Ex.: Enquête longitudinale (10 ans), 218 fonctionnaires, Californie (Mokhtarian et al. 2004) Distances de navette (une direction) plus élevées pour les télé-travailleurs De moins en moins fréquent dans le temps Mais distances totales de navette inférieures pour les télé-travailleurs vs autres travailleurs However, the likelihood of long-term aggregated system-wide impacts of telecommuting seems less obvious, mainly in regard to the small amounts of telecommuting occurring today in North-American cities. In a recent research based on a longitudinal survey (10 years) of 218 California State employees, Mokhtarian et al. (2004) find that one-way commute distances were higher for telecommuters, that average telecommuting practice was lowering over time, but that quarterly total commute distances were lower for telecommuters that for others. It is therefore difficult to assess the actual impact and/or motivation of telecommuting.

4 Motivations pour le télé-travail peu claires
Facteurs liés au travail (volonté employeur) Employés (satisfaction, productivité, contraintes familiales?) Confusion dans la définition du télé-travail Travailleurs autonomes vs employés réguliers Flou relié aux états intermédiaires et grandes différences dans les caractéristiques démographiques des télé-travailleurs selon les études Genre, âge, statut professionnel (et pouvoir décisionnel), régime d’emploi, revenu, etc - Employees’ motivations for teleworking are unclear (Bailey & Kurland, 2002), as well as the effect of work-related factors like managers’ willingness to promote it. From the workers point of view, is it related to satisfaction and productivity? Is it related to family constrains? Is it a way for reducing inconvenience of long trips for parents of younger families who want access to home ownership while they should go farther away on the outskirts? Is it a threat for productivity? Those are issues of paramount importance for understanding socio-economic mechanisms behind this developing phenomenon (and reluctance to promote it) in order to assess its potential impact on reducing travel demand. - However, defining telecommuting is rather difficult. One has to distinguish selfemployed business owners, home-based workers and contractors from full-fledged employees. This leads to fuzzy overlapping between several intermediate states, strong variations of demographic characteristics among studies, probably also linked to peculiarities in survey methods and to the various ways questions are asked (Pratt, 2000). - Olszewski and Mokhtarian (1994) report a large proportion of mid-level professionals in their study involving State of California’s employees. In the U.S., Bailey and Kurland (2002) report on differences linked to gender: fulltime employees who telecommute are more likely to be male, slightly younger and making higher income, while part-time employees which are telecommuting, are doing that more informally, being predominantly female, older, and earning less. - Several previous studies suggest that job suitability is certainly of paramount importance to establish who is more likely to telecommute. However, job suitability must be complemented by status and power that are intertwined with occupation types and freedom of planning tasks. In the literature, both professional and clerical positions are deemed suitable. However, clerical workers having less control over their time schedule, they may encounter greater difficulties to convince their supervisors of the appropriateness of telecommuting.

5 Pourtant, le télé-travail =
alternative la plus prometteuse aux déplacements résidence-travail bonne stratégie pour réduire la demande en transport (surtout en heure de pointe) Donc, nécessité de mieux comprendre ce comportement et ses fondements socio-économiques pour Vérifier l’impact potentiel sur la réduction de la demande de déplacement Mise en place de politiques publiques pour favoriser son développement À l’échelle des grandes, moyennes et petites villes -According to Yen (2000), telecommuting is the most promising substitute of work trips, eventually providing a good strategy for reducing transportation demand. However, building public policies to promote its development implies a better understanding of this behaviour and of its socio-economic foundations, differentiating large, mid-size and small cities, because their traffic conditions are very different. This paper aims at furthering knowledge on telecommuting in QMA.

6 Objectif et hypothèses
Spatialiser les différences comportementales du télé-travail selon les genres et les professions pour la région métropolitaine de Québec (RMQ) Effets liés à l’âge, à la structure du ménage, à la motorisation, aux distances de navette et à la localisation dans la ville

7 Hypothèses (H1) Parce qu’elles détiennent des emplois avec moins de capacité décisionnelle, la probabilité de télé-travailler est plus faible pour les femmes que pour les hommes (H2) Parce qu’ils sont plus attirés par les nouvelles technologies, les jeunes adoptent plus facilement ce nouveau comportement (H3) Le statut professionnel des travailleurs est un facteur déterminant de leur capacité/volonté à télé-travailler (H4) le lieu d’emploi actuel joue un rôle significatif sur la proportion de télé-travailleurs (H5) Parce que les travailleurs souhaitent réduire les inconvénients liés aux déplacements, l’éloignement du lieu de résidence augmente la propension au télé-travail - With a population of roughly 683,000 inhabitants (2001 Canadian Census), the Quebec Metropolitan Area covers 3,154 square kilometres and is characterized by an extensive road network which greatly facilitates motorized movements and prevents most congestion problems. By 2000, the average household income stood at $50,230 ($58,630 for Canada as a whole) while per capita income was $27,939 ($29,769). The work force is devoted to retail business, services, government, administration, education, insurance and health care. Having few manufacturing activities, it is the type of urban area where telecommuting could develop rapidly. - For the purpose of this research, we will use sporadic home-working as a proxy for telecommuting. Following previous findings in the literature, this research aims at testing the following hypotheses for QMA: – (H1) Holding less power-oriented jobs in larger proportions, women are less likely to telecommute (or to be allowed to work at home) than men. – (H2) Being more technologically-literate, younger people are more likely to adopt this new behaviour. – (H3) Professional status of workers is of paramount importance in their ability/willingness to telecommute. – (H4) Actual location of the workplace plays a significant role on the proportion of telecommuters. – (H5) Because people are willing to reduce travel inconvenience, remote home location increases the likeliness of telecommuting

8 Méthodologie Données Analyses statistiques
Enquête origine destination 2001 (MTQ et RTC) personnes, ménages, déplacements Jour typique de la semaine (lundi-vendredi) Localisation précise des lieux de résidence et de travail Caractéristiques des personnes, des ménages et des déplacements Analyses statistiques Tableaux croisés et statistiques du 2 Modèle de régression logistique binomiale de la propension à travailler à la maison Var. dépendante = travailleurs travaillant à la maison au moins 1 fois/2 semaines (1) vs autres travailleurs (0) - From September 18th to December 17th of 2001, the Ministry of Transport of Quebec (MTQ) and the Quebec City Transit Authority (RTC) were conducting a large OD survey in the QMA. - It was involving 68,121 persons (about 8% of total population) living in 27,839 households and reporting 174,243 trips they made to reach activity places during a typical week day (Monday to Friday). - Each household has its home located on a map using street addresses. - Each person belongs to a household and is characterized by his/her age, gender, occupation (worker, student, retired, unemployed, etc.), and ownership of transportation resources: car, driver license, bus pass. - An original programming done by the authors was used in order to assign a role to each member of every household: lone adult, child (less than 16 years or less than 21 years and still at school), adult living in couple (husband, wife, father and mother), and adult living in multiple adults households. - Each workplace was located with methods yielding very accurate spatial references (pointing to a specific building or at least to the city block).

9 Map 1 shows location of workplaces in conjunction with the historical development of neighbourhoods. The initial CBD (La Cité) is complemented with secondary centres in the Sainte-Foy-Sillery, Les Rivières, Valcartier and Desjardins districts. Distribution spatiale des lieux d’emploi selon l’âge du quartier, RMQ, 2001

10 Distribution spatiale des travailleurs, RMQ, 2001
Modes de transport 73,3% automobile 13,2 % autobus 11,4% marche - Map 2 displays residential distribution of workers among older and newer neighbourhoods showing the widespread dispersion of home places. -In the QMA, trips using private cars (driver or passenger) predominate over other transportation means (73.3% cars, 13.2% bus networks, 11.4% walk). With a high density motorway network (21.7 kilometres per 100,000 inhabitants), the QMA is very well-equipped, a fact promoting car trips. Distribution spatiale des travailleurs, RMQ, 2001

11 Travailleurs travailleurs (temps plein et partiel, étudiants exclus) avec un lieu de travail habituel 952 travailleurs à la maison (5/5 jours; 3,3%) Notre définition des télé-travailleurs: travailleurs avec lieu habituel de travail autre que lieu de résidence / travail à domicile 1 jour / 2 semaines Employés salariés Hommes d’affaires travaillant surtout à la maison Exclue: Personnes travaillant au domicile sans aucun autre lieu de travail groupés - Among the 68,121 persons responding the 2001 OD survey, 32,455 were workers (part-time and full-time – students are excluded even if they hold a parttime job). Of them, 30,084 were reporting a specific workplace, 952 of them being home-based workers. -For the purpose of this paper, only full-time workers having a fixed non-home workplace were retained for analysing telecommuting behaviour (27,086 persons). Every worker was asked to disclose the frequency of homeworking he/she was experiencing during the preceding weeks. Answers were later aggregated into six categories (Table 1). - Following definitions of Choo et al. (2005), there are various types of teleworkers: (1) salaried employees of an organization, generally called telecommuters, (2) primary home-based business workers, and (3) people working at home without any other regular workplace, called home workers. It is near to impossible to distinguish between categories 1 and 2 using the 2001 OD survey for QMA; they are therefore agglomerated for the purpose of this research. People working on the road without fixed location or working at home on a regular basis are excluded from this study.

12 Rôle du travailleur dans le ménage
Regroupement des types d’emploi (auto-déclarés) selon pouvoir décisionnel, contrôle des tâches et de l’horaire de travail 1. Directeurs, propriétaires, travailleurs à leur compte 2. Superviseurs et professionnels 3. Employés hautement qualifiés 4. Employés non-spécialisés 5. Travailleurs qualifiés 6. Travailleurs non-qualifiés 7. Autres travailleurs Rôle du travailleur dans le ménage Localisation dans l’arrondissement ou municipalité -Each worker was asked to describe his/her type of job in his/her own terms. These job descriptors were later aggregated to categories building a hierarchy of decreasing power and control of duties and scheduling. - These categories distinguish groups of professions with similar level of authority. They are: (1) heads of hierarchy (managers; owners; self-employed persons), (2) supervisors and professionals (foremen; professors; doctors; dentists; lawyers; engineers; other professionals), (3) highly qualified employees (technicians; accountants; etc.), (4) unskilled employees (office clerks; administration employees; etc.); (5) qualified workers (plumbers, electricians, etc.); (6) unskilled workers (manoeuvres; blue-collars; salespersons; etc.), and (7) other workers (E.g. militaries). - Persons were also qualified according to their specific role within their household: lone persons, spouses (husbands/wives), lone parents, and other adults living in multiple adults or multi-generational households. - The number of cars per household was asked during the OD survey. - Finally, each workplace was assigned to a neighbourhood and to the municipality where it belongs.

13 Tableau 1 Fréquence du télé-travail, RMQ, 2001
- Cross-tables relating each attribute workers and the six categories of home work were built and a Chi-Square statistics was computed in order to test for the significance of relationships. - Table 1 relates frequency of home work to genders considering full-time and part-time workers. It shows a significant difference among genders. Men are more likely to telecommute 1 to 4 days per month. Women are more prone to work at home on a regular basis (home workers). 8.3% Pearson Chi-Square: 35.9 (p<0.000) Source : MTQ-RTC Enquête OD-2001 (sans expansion) Compilations réalisées par les auteurs

14 Tableau 2 Fréquence du télé-travail selon l’âge
2.8% 7.3% 8.8% 8.7% 9.8% 9.2% 9.6% En rouge, % total des télétravailleurs (1 jour/2sem +1 jour/sem jours/sem +4 jours/sem) Diminution des fréquences d’occurrences de 1 jour par 2 sem à 4 jours/sem (impact sur la demande en trabsport ?) Telecommuting patterns show significant relationships with age (Table 4 – younger workers seldom work at home and the proportion of telecommuters increases with age – about 16% among the years old and 27% among the elderly). Pearson Chi-Square: (p<0.000) Source : MTQ-RTC Enquête OD-2001 (sans expansion) Compilations réalisées par les auteurs

15 Tableau 3 Fréquence du télé-travail selon le statut professionnel
16.4% 17.1% 8.1% 33.8% 12.5% 28.7% 13.3% 19.5% 4.8% Relationships with professional statuses (Table 5) show that proportion of telecommuters is strongly related to qualifications and decisional status of the person, yielding higher levels of telecommuting for managers, self-employed persons, professors and lawyers than for office clerks, technicians and non-qualified workers. This last relationship is very strong suggesting that job empowerment (especially ability to control time schedule and to put priorities on duties) is of paramount importance for the development of telecommuting Source : MTQ-RTC Enquête OD-2001 (sans expansion) Compilations réalisées par les auteurs

16 Tableau 3 Fréquence du télé-travail selon le statut professionnel (suite)
12.4% 5.1% 2.5% 1.5% Relationships with professional statuses (Table 5) show that proportion of telecommuters is strongly related to qualifications and decisional status of the person, yielding higher levels of telecommuting for managers, self-employed persons, professors and lawyers than for office clerks, technicians and non-qualified workers. This last relationship is very strong suggesting that job empowerment (especially ability to control time schedule and to put priorities on duties) is of paramount importance for the development of telecommuting 1.0% 4.1% Pearson Chi-Square: (p<0.000)

17 Tableau 4 Fréquence du télé-travail selon le lieu de travail
5,3% 9,9% 6,7% 10,7% 8,6% Significant differences appear also when considering workplace locations within the city (Table 4). People working near the city centre (La Cité and Sainte-Foy-Sillery) are more willing than others to consider telecommuting. (9.8 and 10.7% respectively compared to 6.7% for Les Rivièresor 6.5% for Beauport People working in the suburban (Haute-Saint-Charles: 9.9%, Laurentien: 8.0%, Chutes-de-la-Chaudière West: 10.5%) and rural areas (7.7%) show higher levels of home-based working. Compared to 1.6% for La Cité or 2.2% pour Sillery-Ste-Foy 6,5% 7,7% Source : MTQ-RTC Enquête OD-2001 (sans expansion) Compilations réalisées par les auteurs

18 Tableau 4 Fréquence du télé-travail selon le lieu de travail (suite)
7,3% 6,2% 4,4% 8,0% 8,5% - People working in the suburban (Haute-Saint-Charles: 9.9%, Laurentien: 8.0%, Chutes-de-la-Chaudière West: 10.5%) and rural areas (7.7%) show higher levels of home-based working. Compared to 1.6% for La Cité or 2.2% pour Sillery-Ste-Foy 8,3% Source : MTQ-RTC Enquête OD-2001 (sans expansion) Compilations réalisées par les auteurs

19 Autres résultats significatifs des tableaux de contingence
Les télé-travailleurs achètent moins de laisser-passez d’autobus Les détenteurs de permis de conduire télé-travaillent dans de plus grandes proportions Les parents seuls pratiquent le télé-travail dans de plus grandes proportions que les parents traditionnels La disponibilité d’une automobile influence positivement le télé-travail; le travail basé au domicile est plus fréquent dans les ménages non-motorisés People working at home are less willing to buy a bus pass : 3.4% for no pass vs 0.8% for a pass Driver license owners are telecommuting in somewhat larger proportions: 8.5% vs 4.9% for no driver license Lone parents (having higher family constraints) are experiencing some level of telecommuting in larger proportions than husband/wife: 11.7% vs 9% Finally, car availability (at the household level) impacts positively on telecommuting (Table 8); home-based work being more prevalent for people living in non-motorized household: 4.7% (vs 3.0% for two or more cars hhds)

20 Tableau 5 Modèle de régression logistique binomiale du télé-travail
- Building contingency tables can be useful to explore relationship. - However, to test the actual significance of a relationship, one has to control for concomitant links because any Chi-Square relationship may arise from the effect of co-factors (E.g. a relationship with neighbourhoods can result from spatial segregation among professions and age groups, based on their average income). - This can be done using regression techniques in order to assess the marginal effect of one attribute while controlling the effect of co-factors. Table 9 presents a logistic binomial model of the propensity to telecommute. Dependent variable is binary: 1 for workers having a fixed workplace and working at home at least one day per period of two weeks; 0 for any other worker. Independent variables are related to personal attributes (gender, age, role in the household), access to transportation resources (driver license, bus pass, cars held by the household), professional status, location of workplace, distances from home to work and to central axis in the agglomeration (from Old Quebec to Laurier Shopping Centre). Significant relationships are bold. The significant relationships with access to transportation resources found in contingency tables disappear when considering locations, distances and professional statuses. Therefore, we may conclude that access to transportation means is not instrumental in the decision to telecommute, and that dual relationships comes from structural effects (e.g. households in remote areas are motorised; most highly qualified workers do have a driver license, etc.). Significant relationships hold with gender (ceteris paribus and in line with H1, women are far less likely to telecommute than men, meaning that there is gender effect on top of professional status), age (younger workers, 15 to 24 years old, are far less prone than others to telecommute; telecommuting is more prevalent among older workers – H2 is then rejected), household status (lone parents are more likely to commute than lone persons; married persons do not differ from lone persons; workers living in multiple adults households are far less likely to telecommute). Significance of coefficients: *<.05;**<.01; ***<.001

21 Tableau 5 Modèle de régression logistique binomiale
du télé-travail (suite) Professional status is certainly the most important attribute that should be used to model propensity to telecommute (Wald statistic is 1636 with 14 degrees of freedom). To ease comparison among professions, the reference is put on managers (16.4% of them are telecommuting – Table 5). Having to control other employees, foremen are twice less likely to telework (odds ratio is ..301). Among the highly qualified professionals, professors and lawyers are the best candidates for telecommuting; odds are respectively at 2.9 and 1.9. Powerless categories of employees (technicians, unskilled employees, qualified and unskilled workers) are far less likely to telecommute, their odds ratios are drastically decreasing when lowering their level of education. Except for some peculiarities linked to the specific nature of duties, the relationship with empowerment seems instrumental, thus strongly supporting H3. However, some peculiarities of women duties in the labour market appear when we use gender as interaction variable self-employed women are more prone to adopt telecommuting behaviour when their male counterpart do not; forewomen behave very differently than foremen (2.632 versus 0.301); the same comment apply for professors (there is no significant differences among gender among this group). More important is the inversion of odds ratios for qualified and unqualified workers. Women of those powerless categories are far more willing than their male colleagues for considering telecommuting (Odds ratios at more than 4). It seems that the relationships postulated by H3 and H1 still hold, but are far more complex than preliminary expectations. That is congruent with previous findings of Bailey & Kurland (2002, p. 386): “Olson & Primps (1984) report that clerical workers lost full-time permanent status, medical benefits, and vacation when they converted to telework, and their already low autonomy became further restricted. Professionals in that study, by contrast, were offered teleworking arrangements more in line with a job enrichment perspective, such that their autonomy, already high, expanded by working at home.” Significance of coefficients: *<.05;**<.01; ***<.001

22 Tableau 5 Modèle de régression logistique binomiale
du télé-travail (fin) Location of work places has a significant effect. The reference is put on rural areas. Odds of adopting telecommuting are higher when workplace is close to the city centre. New suburbs and agglomeration outskirts do significantly differ from rural areas in their proportions of teleworkers. This decreasing gradient of telecommuting when going apart from city centre is in line with H4. Finally, increasing Euclidean distance between home and workplace imply higher proportion of telecommuters (probability increases at a rhythm of nearly 1% for each Km of distance). However, distance between home and central axis of the agglomeration do has exactly the reverse effect, in clear contradiction with H5 but the odds are not significant Nevertheless, the odds ratio of this last distance is probably underestimated because it is multi-collinear with location of workplaces (most of them being on the central axis) and distance from home to work. - When controlling other co-factors, spatial structure in telecommuting displays more clearly, while remaining very complex. That is in-line with earlier findings by Thériault et al. (2005) in this same region. Results indicate that telecommuters are marginally trying to reduce inconvenience of travel (supporting H5), but this is only significant for low-frequency telecommuters (odds ratio of for each kilometre of Euclidean distance from home to workplace). However, this effect is counterbalanced by a reverse effect when going far away from main labour market (odds ratio of for each kilometre separating the home from the cental axis). This implies that there is some directional bias or anisotropy imbedded in the spatial structures related to telecommuting behaviour in the QMA. This is in-line with finding of Vandersmissen et al. (2003) reporting directional structures in travel times in the same region, probably related to the configuration of motorways. Significance of coefficients: *<.05;**<.01; ***<.001

23 Conclusion et perspectives
Modéliser le télé-travail n’est pas chose facile Niveau individuel = négociations entre employé et son employeur/superviseur Résultats vs Hypothèses (H1) Parce qu’elles détiennent des emplois avec moins de capacité décisionnelle, la probabilité de télé-travailler est plus faible pour les femmes que pour les hommes Confirmée en partie: car les femmes avec emplois à faible capacité décisionnelle télé-travaillent plus que leurs collègues masculins (H2) Parce qu’ils sont plus attirés par les nouvelles technologies, les jeunes adoptent plus facilement ce nouveau comportement Rejetée: les travailleurs plus âgés télé-travaillent dans de plus grandes proportions Preliminary results indicate that modelling telecommuting is not a simple task. At the individual level, one has to consider specific conditions involved in the negotiation between an employee, his/her supervisor and his/her employer. However, this study demonstrates that large OD surveys can yield interesting results at a city-wide level. Both models are highly significant, most coefficients are in line with expectations. For this particular research, we were testing five hypotheses: (H1) Occupying less power-oriented jobs in larger proportions, women are less likely to telework than men is partly supported by empirical data. However the relationship is not so simple because, ceteris paribus, less empowered women are teleworking more than their male colleagues. Is it their choice or an indication of degradation on their working conditions? (H2) Younger people are more likely to adopt this new behaviour is absolutely not supported by the facts and is rejected. In QMA, trends indicate that older workers are working at home in larger proportions.

24 (H3) Le statut professionnel des travailleurs est un facteur déterminant de leur capacité/volonté à télé-travailler Confirmée : mais inclut un effet lié au genre (H4) le lieu d’emploi actuel joue un rôle significatif sur le nombre de télé-travailleurs Confirmée: les lieux d’emplois centraux augmentent la probabilité d’adopter le télé-travail (mais effet directionnel ?) (H5) Parce que les travailleurs souhaitent réduire les inconvénients liés aux déplacements, le fait d’éloigner son lieu de résidence augmente la probabilité de télé-travailler Non-confirmée: le télé-travail diminue lorsque la distance entre la résidence et l’axe central augmente (H3) Professional status of workers is of paramount importance in their ability/willingness to telecommute is supported by facts, but should be interpreted with due care because it includes a gender effect, it is closely related by the exact nature of duties and empowerment, as well as to the will of organisations to favour this kind of labour relationship. (H4) Actual location of the workplace (type of neighbourhood) plays a role on the proportion of teleworkers is supported by empirical results. Even for a low-congestion agglomeration like QMA, it seems that centrally located workplaces increase the propensity to adopt telecommuting behaviour. Because our models are controlling for professional status, it seems reasonable to conclude that the marginal effect of this location factor is rooted in reality. However findings of Table 10 indicate that the odds are affected by a directional effect. Further research is needed in order to clarify the meaning of this finding. (H5) Remote home location increases the likeliness of telecommuting is absolutely not supported by empirical investigation; the gradient in the proportion of telecommuter is decreasing when distance of home from city centre is increasing.

25 Effet sur la demande en transport à Québec ?
88,4% des travailleurs ne pratiquent jamais le télé-travail 8,3% des travailleurs le pratiquent au moins 1 fois par deux semaines Plus de la moitié (4,9%) 1 fois par deux semaines seulement 3,4% 1 fois par semaine ou plus 3,3% travaillent à domicile en tout temps

26 Limites et perspectives
Absence d’information sur le revenu Pas de distinction entre travailleurs à temps plein et à temps partiel (à faire) Distances routières et durées à la place des distances euclidiennes (à faire) Meilleure compréhension de l’effet spatial Identification de l’employeur Type d’employeur (à faire) Améliorer la définition opérationnelle du télé-travail Mais question retirée de l’enquête OD 2006 Comment faire suivi et analyse l’influence sur la demande en transport et congestion routière ????? Due to peculiarities of data, this study has some limitations. Firstly, we do not have indication of the level of income of workers. Secondly, full-time and part-time employment regime should be distinguished; that will be done in a later step. Thirdly, Euclidean distances could be replaced by road distances or travel time; we expect that using more precise assessment will provide better understanding of directional bias, eventually leading to an experiment similar to that of Vandersmissen et al. (2003). Fourthly, taking into consideration the identification of the employer can improve the analytical capacities of our model; we have this information for a sample of workers in our database. Finally, the exact nature of the question about telecommuting should be improved, in line with recommendations of Joanne Pratt (2000) in order to enhance results for future OD surveys within QMA. Cependant, on peut se questionner sur l’intérêt des organismes publics (MTQ et RTC) envers le télé-travail puisque la question (posée pour la première fois en 2001) a été enlevée du questionnaire de 2006! Donc, ce sera très difficile de faire un suivi de ce phénomène et d’arriver un jour à estimer son impact sur la demande en tranport, la congestion routière etc.

27 Merci! Questions ?