孙斌栋
, 尹春
Sun Bindong, Yin Chun
中图分类号: D669.3;TU984
文献标识码: A
文章编号: 1000-0690（2018）01-0041-08
通讯作者:
收稿日期: 2017-04-18
修回日期: 2017-08-13
网络出版日期: 2018-01-10
版权声明: 2018 《地理科学》编辑部 本文是开放获取期刊文献，在以下情况下可以自由使用：学术研究、学术交流、科研教学等，但不允许用于商业目的.
基金资助:
作者简介:
作者简介：孙斌栋（1970-）,男,河北阜平人,博士,教授,博士生导师,主要从事城市地理、城市规划和区域经济研究。E-mail: bdsun@re.ecnu.edu.cn
展开
摘要
基于上海市1 605个居民抽样样本,利用多层次线性回归模型,探讨了街道尺度人口密度对个体居民通勤时耗的影响及条件效应。在控制了个体社会经济属性、通勤方式和其他建成环境要素后发现,高人口密度总体上会增加居民通勤时耗,而且其作用是有条件的,主要受“设计”“地铁站可达性”“到就业中心距离”和“通勤方式”的调节。具体而言,提高街道路网密度和高速路密度,缩短居住地到就业中心距离,提高地铁站密度,鼓励居民采用地铁方式通勤,减少小汽车使用,都可以显著降低人口密度对通勤时耗的增加作用。因而,通过优化城市建成环境和调节人口密度来降低过长的通勤时耗是可行的。
关键词：
Abstract
Given that long commuting duration is considered as a crucial problem of urban transportation, urban planners and geographers explored the impacts of built environment on commuting duration in order to reduce long commuting duration and alleviate traffic congestion. However, the findings on the connection between population density and individual commuting duration are inconsistent. Some literature showed that population density and individual commuting duration have a negative correlation, while others found a positive correlation or an insignificant relationship. Secondly, previous literature on individual commuting duration paid little attention to the conditional effects of population density. Just as Jane Jacobs remarked at The Death and Life of Great American Cities, “No concentration of residents, however high it may be, is ‘sufficient’ if diversity is suppressed or thwarted by other insufficiencies”, the impact of population density on commuting duration may depend on the other built environmental attributes. Furthermore, previous studies focused mainly on the cases in developed countries and far less on developing countries especially China whose urban built environment is characterized by larger population size and higher density. Based on a sample of 1 605 individuals from the “the Yangtze River Delta Social Transformation Survey (FYRST)” Project in Shanghai in 2013, this study examines the impacts and conditional effects of population density at the sub-district spatial level on individual commuting duration by using hierarchical models. Controlling individual socioeconomic attributes, commuting modes, and other built environmental attributes including diversity, design, accessibility to transit and distance to the job center, the random intercept model estimations show that individual commuting duration is positively associated with population density at the sub-district spatial level. One possible reason is the traffic congestion resulting from extremely high density and the job-housing imbalance. Moreover, this effect of population density on commuting duration is conditional on the “design”, the “accessibility to metro”, the “distance to the job center”, and the “commuting mode”. Specifically, higher road network density and highway density, shorter distance to the job center, higher metro stations density, and encouraging commuting by metro instead of driving could diminish the duration increasing effect of population density. But the expected conditional effect of “diversity” is not significant and not supported by this article. Thus, it is feasible to reduce commuting duration and traffic congestion by optimizing built environment and adjusting population density. Firstly, it’s necessary to point out that the advocated concept of compact development should orient at reasonable high density instead of over-concentration. Secondly, planners should pay more attention to encouraging job-housing balance, including both quantitative and structural balance. Last but not least, it is important to expand the underground space to encourage commuting by metro and to construct the highway to shunt the traffic flow for the sake of shortening commuting duration and improving traffic efficiency.
Keywords：
随着中国城市化的推进和社会经济的高速发展,以通勤时耗增加为标志的城市交通拥堵问题日益凸显。城市地理学家和城市规划工作者提出新城市主义、精明增长等理论,力图从优化城市土地利用空间结构和建成环境的角度来改善城市交通环境,促进职住平衡,缩短通勤距离和时耗,缓解城市交通问题。
人口密度作为建成环境要素中最重要的指标,其对通勤时耗的影响却仍未得到一致结论[1]。西方学者普遍发现,高人口密度与平均通勤时耗或个体通勤时耗呈负向关联[2,3,4]。其原因被认为是,在高人口密度地区,居民更可能在附近找到工作[5]。有学者对北京市通勤者的研究也发现高人口密度与通勤时耗显著负相关[6,7]。另一方面,也有西方学者认为,过度密集会导致交通集中化,造成交通拥堵,延长通勤时耗[8,9],一些学者对广州和上海的实证研究支持了这一观点[10,11]。也有学者认为人口密度对通勤时耗的影响可能是非线性的,如存在“U”型关系[12]。还有学者对北京市通勤者的通勤时耗研究发现,人口密度对通勤时耗的作用并不显著,其原因被认为是过度集中产生的拥堵效应与由高密度带来的目的地邻近效应相抵消[13]。
已有文献在探讨人口密度对通勤时耗的影响时,对人口密度的条件效应缺乏关注。早在1961年,雅各布斯在《美国大城市的死与生》就认为,人口密度对个体出行行为的作用,受到其他因素很大的影响,“一个住宅区的住宅密度不管有多高,如果因为其他方面因素的不到位而阻碍或压制了多样性,那么密度再高也起不了什么作用”;“尽管这样,……如果没有这个条件（高密度）,促进多样性产生的其他因素也就不会有什么作用”[14]。Handy指出,“恐怕不是密度,可能是伴随密度的其他因素对通勤行为产生作用”[15]。一些学者也认为,人口密度本身的作用可能不大,而是人口密度代理的变量（如公交可达性）或与人口密度相伴随的其他因素在起作用[16, 17]。但有研究发现,在控制了其他因素后,密度本身对出行模式就有重要影响[18]。由于对密度本身作用的不确定,已有文献在尝试解释人口密度对个体通勤行为的作用机制时,主要认为密度的作用是依赖于建成环境的其他因素,而不仅仅是其本身的作用,如高密度周边伴随设施邻近而更可能实现职住平衡,利于步行、公交的环境而促进非机动化的出行方式[6,19],或认为高密度带来的交通拥堵延长了通勤时耗[10,11,20]。
可以发现,目前人口密度与个体通勤时耗的关系尚没有明确的结论,且其条件效应被忽略,这可能会影响规划的成效。再者,中国城市与西方国家城市建成环境特征具有明显差异,尤其体现在人口密度上。因而,通过对中国城市人口密度与居民通勤时耗关系的进一步探讨,检验人口密度对通勤时耗的作用是否受其他建成环境要素的影响,有助于为中国城市规划实践提供科学决策依据。
数据来自2013年“长三角地区社会变迁调查”（FYRST-2013）上海调查数据。该调查由复旦大学社会科学数据研究中心主持,以1980~1989年出生的一代人（简称80后）为跟踪主体,采用分层、多级概率抽样设计,在上海市范围内抽取40个街道（镇）80个居（村）委会3 311个家庭,正式入户调查完成问卷2 774份,共回收有效问卷2 367份[21]。80后群体经历了单位制度的解体,正处在人生的重要转折阶段,这部分人群的通勤行为更能反映建成环境对新兴一代居民的普遍影响。第六次人口普查数据也表明,上海约1/3的家庭中有80后群体,因此这部分样本具有相当的代表性。本文研究区域为上海市辖区,剔除样本缺失值和崇明岛样本后,本文的工作样本为1 605人,其所在街道（镇）见图1。
图1 抽样街道（镇）及街道（镇）中心点
Fig. 1 The distribution of sample and the center of streets in 2013
居民通勤时耗主要受街道建成环境、个体通勤方式和个体社会经济属性影响（图2）。由于居民和街道是一种嵌套关系,即居民居住在街道中,故采用多层次模型进行分析。本文主要关注高层次人口密度变量对低层次的个体通勤时耗的直接影响和条件效应,将个体层变量作为模型的控制变量纳入模型,旨在得出一般性的均质结论。在控制了个体层变量后,认为建成环境要素对居民通勤时耗在各个街道的影响斜率是一致的,其差别只体现在截距上,故仅考虑随机截距模型,模型设置如下：
式中,Cij代表居住于j街道i个体的通勤时耗,是固定效应部分与随机效应部分之和。第一个括号中的各成分是模型的固定效应部分,B0j代表第j个街道的建成环境变量,Mij代表居住于j街道i个体的通勤方式变量,Sij代表居住于j街道i个体的社会经济属性变量,β是回归系数。其中,β0代表总截距（总体平均值）,是固定参数;β1、β2、β3分别是建成环境变量、通勤方式变量、社会经济属性变量的回归系数。第二个括号中的各成分是模型的随机效应部分,其中,μ0j代表第j个街道的截距到总截距β0的距离,是街道层次的随机变量,反映了组间变异的大小;εij代表居住于j街道i个体到j街道截距的偏离,也就是常见的随机误差项,反映了组内变异的大小。
图2 研究分析框架
①是人口密度对个体通勤时耗的主效应;②~⑥分别是人口密度在多样性、设计、公共交通可达性、到就业中心距离和个体通勤方式条件下对个体通勤时耗的条件效应
Fig. 2 The analytical framework of this study
1.3.1 个体层变量
模型因变量为居民个体通勤时耗。在数据处理中,为了防止异常值影响研究结果,对该变量进行2.5%双侧缩尾处理,并在进行回归时,对其取对数。在工作样本中,平均通勤时耗为49.64 min,与百度“我的2014年上班路”上海市平均的51 min通勤时耗十分接近[22],可以认为研究所用的“80后”样本基本可以反映上海市居民普遍的通勤时耗现状。
通勤方式包括步行、自行车、电动车、公交车（含单位班车）、地铁和小汽车（含出租车）6种方式。使用公交车和地铁方式通勤的比重高达54.71%,说明样本主要采取公共交通方式通勤;除步行外,自行车、电动车、公交车和小汽车4类路面通勤方式,占到了64.05%,说明居民通勤活动主要发生在路面。从单趟单方向平均通勤时耗来看,步行最短,为17.9 min,其后依次是电动车（24.2 min）、自行车（28.1 min）、小汽车（35.1 min）,而公交车和地铁的平均通勤时耗最长,分别为65.7 min和71.6 min,远高于其他通勤方式。值得一提的是,在理论上,通勤方式和通勤时耗可能存在双向因果,因为不同通勤方式因其速度不同会影响通勤时耗,而通勤时耗也会促使居民对其通勤方式进行调整[23]。但本文采用的是截面数据,所汇报的通勤时耗一定是其当前采用的通勤方式的结果,故不存在双向因果问题。
个体社会经济属性包括性别、年龄、户口、受教育年限、年收入、已婚或同居、孩子数、租房。研究样本中,男女比例相当,平均年龄在28岁,有上海户口占78%,平均受教育年限为14 a,接近专科学历,年收入平均在6.2万元,已婚或同居居民占52%,还没有子女的居民占61%,22%的居民租房居住,这些都基本符合上海市“80后”的社会经济属性现状。在进行回归分析时,对年收入加1后取对数。
1.3.2 街道层变量
样本分布在上海市37个街道,建成环境变量除了核心关注的人口密度外,还包括土地混合度、路网密度、高速路密度、公交站密度、地铁站密度、到最近就业中心距离。研究样本街道平均人口密度1.87万人/km2,平均土地混合度为0.55,路网密度平均为4.30 km/km2,高速路密度平均为0.94 km/km2,公交站密度均值是19.36个/km2,地铁站密度均值是0.37个/km2,平均到就业中心距离8.32 km。
人口密度依据街道常住人口数量除以街道行政面积计算得到,由于街道尺度常住人口数据只在普查年份统计,故采用2010年上海市第六次人口普查数据。
多样性通过土地混合度反映。土地混合度（MIX）反映了街道内设施多样性的情况,其表达式[24]为：
式中,Nk是街道内设施k的数量,N是街道所有设施的数量,值越大说明土地混合度越大。11类土地利用设施包括了政府机构、医院、学校科研院所、宾馆酒店、公司企业、公园广场、餐饮娱乐、商业大厦、银行、零售行业和其他设施[25]。
设计包括了路网密度和高速路密度两个指标,分别将街道内主要道路里程、高速路里程除以街道行政面积计算得到。公共交通可达性通过公交站密度和地铁站密度反映,分别将街道内公交站点个数和地铁站点个数除以街道行政面积计算得到。到就业中心距离用街道几何中心点（图1）到最近就业中心的直线距离表示,反映居民跨街道通勤可能的通勤距离。就业中心包括外滩街道、潍坊新村街道、沪东新村街道、安亭镇、新成路街道、颛桥镇、江川路街道、莘庄镇、松江工业区、石湖荡镇、廊下镇,共11个就业中心[26]。
在进行回归分析时,对所有街道建成环境变量都取对数,最小值存在0值的变量,加1后取对数。
零模型结果显示,街道之间的变异值为0.150（t=3.846）,十分显著,说明街道层次对解释个体通勤时耗具有重要意义。零模型组内相关系数（ICC）为19.6%,属于高强度关联,即认为个体通勤时耗中约有19.6%的变异来自于街道层次,其余80.4%来自于个体层次。由此可见,个体层和街道层因素都是解释居民通勤时耗的重要因素,虽然个体层更重要一些,但街道层因素的作用仍不能忽视。此外,研究样本在每个街道平均为43.4个,满足多层次模型对样本量的需求。
表1显示了多层次线性回归模型的分析结果。模型1是基础模型,模型2~7分别加入密度与其他建成环境及通勤方式的交互项以检验条件效应,土地混合度、公交站密度和公交车通勤方式分别与密度的交互项都没有通过显著性检验,限于篇幅在这里没有报告其结果。
表1 多层次线性回归模型结果（样本量：1605）
Table 1 The estimation results of multilevel models (sample size: 1605)
| 模型1 | 模型2 | 模型3 | 模型4 | 模型5 | 模型6 | 模型7 | |
|---|---|---|---|---|---|---|---|
| 建成环境 | |||||||
| 人口密度 | 0.143* | 0.317** | 0.236*** | 0.238*** | -0.057 | 0.175** | 0.132* |
| 土地混合度 | 0.019 | -0.025 | 0.157 | -0.169 | 0.203 | 0.065 | 0.030 |
| 路网密度 | 0.076 | 0.069 | 0.014 | 0.102 | 0.093 | 0.088 | 0.079 |
| 高速路密度 | -0.004 | -0.055 | 0.148 | -0.080 | -0.061 | -0.013 | -0.006 |
| 公交站密度 | -0.026 | -0.092 | -0.033 | -0.088 | -0.047 | -0.040 | -0.029 |
| 地铁站密度 | -0.241 | -0.077 | -0.196 | 0.500 | -0.127 | -0.216 | -0.242 |
| 到最近就业中心距离 | 0.101** | 0.116*** | 0.099** | 0.124*** | 0.140*** | 0.110*** | 0.102** |
| 条件效应 | |||||||
| 人口密度×路网密度 | -0.114* | ||||||
| 人口密度×高速路密度 | -0.214** | ||||||
| 人口密度×地铁站密度 | -0.605** | ||||||
| 人口密度×到最近就业中心距离 | 0.094* | ||||||
| 人口密度×地铁方式通勤 | -0.149*** | ||||||
| 人口密度×小汽车方式通勤 | 0.079** | ||||||
| 通勤方式（以步行为参照） | |||||||
| 自行车 | 0.532*** | 0.533*** | 0.531*** | 0.539*** | 0.532*** | 0.539*** | 0.529*** |
| 电动车 | 0.482*** | 0.480*** | 0.480*** | 0.489*** | 0.479*** | 0.497*** | 0.468*** |
| 公交车 | 1.241*** | 1.241*** | 1.240*** | 1.243*** | 1.242*** | 1.245*** | 1.241*** |
| 地铁 | 1.330*** | 1.331*** | 1.331*** | 1.328*** | 1.329*** | 1.433*** | 1.339*** |
| 小汽车 | 0.614*** | 0.613*** | 0.612*** | 0.620*** | 0.615*** | 0.628*** | 0.638*** |
| 个体社会经济属性 | |||||||
| 男性 | 0.037 | 0.037 | 0.037 | 0.035 | 0.037 | 0.043 | 0.040 |
| 年龄 | 0.020*** | 0.020*** | 0.020*** | 0.020*** | 0.020*** | 0.019*** | 0.019*** |
| 上海户口 | 0.068 | 0.069 | 0.064 | 0.071 | 0.069 | 0.064 | 0.070 |
| 受教育年限 | 0.028*** | 0.028*** | 0.028*** | 0.027*** | 0.028*** | 0.026*** | 0.029*** |
| 年收入 | 0.003 | 0.003 | 0.003 | 0.003 | 0.004 | 0.003 | 0.003 |
| 已婚或同居 | -0.005 | -0.006 | -0.007 | -0.007 | -0.006 | -0.000 | -0.006 |
| 孩子数 | -0.016 | -0.017 | -0.018 | -0.014 | -0.015 | -0.010 | -0.015 |
| 租房 | 0.017 | 0.016 | 0.011 | 0.020 | 0.022 | 0.014 | 0.020 |
| 常数项 | 1.429*** | 1.649*** | 1.412*** | 1.595*** | 1.299*** | 1.459*** | 1.424*** |
| 街道间变异 | -1.821*** | -1.871*** | -1.919*** | -1.943*** | -1.886*** | -1.904*** | -1.831*** |
| 街道内变异 | -0.428*** | -0.428*** | -0.428*** | -0.428*** | -0.428*** | -0.430*** | -0.429*** |
| Log likelihood | -1613.710 | -1612.218 | -1610.881 | -1610.788 | -1612.003 | -1608.375 | -1611.505 |
| chi2 | 823.007 | 835.405 | 847.793 | 852.438 | 838.650 | 852.458 | 830.918 |
| ICC | 0.058 | 0.053 | 0.048 | 0.046 | 0.051 | 0.050 | 0.057 |
模型1是基础模型的回归结果,在控制了通勤方式和社会经济属性后,街道建成环境中人口密度和到最近就业中心距离均呈正显著,这证明了中国城市人口密度对通勤时耗的影响规律确实不同于西方。可能的原因有两个方面：首先,街道人口密度高,具有通勤需求的人口基数大,在通勤高峰期时更可能造成拥堵,延长通勤时耗[10]。其次,高人口密度也可能在通勤高峰时段带来大量非通勤目的的出行活动,如通学、就医、外出办事等,这些都可能增加道路拥挤情况,延长通勤时耗。在实证分析中,还尝试加入了人口密度的二次项探讨人口密度对通勤时耗可能存在的非线性关系,但并不显著。居住地到最近就业中心距离与通勤时耗呈正相关,说明距离就业中心越远,居民通勤时耗越长。
模型2和模型3检验了“设计”对人口密度作用的条件效应,在模型1的基础上分别加入了人口密度与路网密度的交互项、人口密度与高速路密度的交互项,两者都通过了显著性检验且系数符号都为负,说明路网密度和高速路密度都可以降低人口密度对通勤时耗的增加作用。在土地较少用于道路交通的中国城市背景下,路网密度提高了道路通达性水平,有助于提高交通速度[27,28];当局部路段拥堵时,居民可以根据实时路况灵活选择更便捷的通勤路线,使通勤人口在路面上分布更分散,降低高人口密度带来的无序拥堵,缩短通勤时耗。高速路作为立体交通,其修建开拓了更多的可利用空间,其最显著的特点就是车速高,可以使居民更快到达目的地;同时,高速路密度的提高,也能分流一部分路面交通,这些在一定程度上都能降低高人口密度的对通勤时耗的增加效应。
模型4检验了“公共交通可达性”对人口密度作用的条件效应,在模型1的基础上加入了人口密度与地铁站密度的交互项,系数符号为负,说明人口密度对通勤时耗的增加作用会因地铁站密度的提高而降低。原因可能是,地铁将大量通勤个体引入地下空间,分担了路面人口密度,进而降低了高密度带来的高通勤时耗。检验“公共交通可达性”时,还尝试在模型1基础上加入人口密度与公交站密度的交互项,其符号为负,虽然符合预期,但并不显著。可能是由于目前上海并没有真正实现公交优先,且公交车因其自身车辆性质（如车速较慢、绕路等）而处于竞争劣势,在通勤高峰时间,其速度甚至不及自行车,因而不能对人口密度产生条件效应。
模型5检验了“到就业中心距离”对人口密度作用的条件效应,在模型1的基础上加入了人口密度与到最近就业中心距离的交互项,结果呈正显著,但此时,人口密度一次项不显著,进行对中处理后,仍不显著。这说明,人口密度的主效应被人口密度与到最近就业中心距离的交互项所承担,即人口密度作用的前提是职住距离的存在,随着到最近就业中心距离的增加,人口密度对通勤时耗的增加效应越强。可能的原因是,职住距离的增加使一定地域空间的通勤量增加,进而导致通勤密度上升,造成通勤时耗的增加。这样的结果在一定程度上也支持了雅各布斯和Handy的观点[14,15],即控制了人口密度发挥作用的条件后,人口密度本身的独立作用会减弱。
模型6和模型7检验了“通勤方式”条件,在模型1的基础上分别加入了人口密度与地铁方式通勤的交互项、人口密度与小汽车方式通勤的交互项,两者都通过了显著性检验,前者为负,后者为正。这说明,使用地铁方式通勤可以减少地面交通流而降低人口密度对通勤时耗的增加作用,而使用小汽车方式通勤则会加剧道路拥挤,强化人口密度对通勤时耗的增加作用。本文还检验了人口密度和公交车方式通勤的交互项,但没有通过显著性检验。
此外,在模型1的基础上尝试加入了人口密度与土地混合度的交互项,以检验“多样性”对人口密度作用的条件效应,但并没有通过显著性检验,且系数符号为正。由此可见,“多样性”条件并不被支持。可能的原因是,随着土地混合度的增加,设施的便利和邻近也可能增加通勤时非通勤目的的出行人口;同时,在高土地混合度的街道,居民出行更可能是多目的出行链,通勤目的地只是其中的一个组成部分,这也可能增加通勤密度和居民通勤时耗[29]。因此,“多样性”条件还有待进一步验证。
就个体社会经济属性而言,年龄和受教育年限与通勤时耗均呈正显著,说明80后群体（24~35岁）随着年龄和受教育年限的增加,通勤时耗也随之增加。西方研究表明年龄与通勤距离呈倒U型关系,即在35岁之前,随着年龄增加,通勤距离逐渐提高[30],而通勤距离与通勤时耗存在明显的正向关联,因而西方这一研究结果跟本文结论是一致的。有关中国城市的研究也发现,30岁左右的居民通勤时耗最长,并认为这可能是缺乏一定工作经验和物质基础造成的[31, 32],随着年龄的增加,在事业上升期的80后群体更可能为寻求更好的岗位和收入,承担更长的通勤时耗。
关于受教育年限,西方研究同样表明,受教育水平越高的居民对于收入要求也高,因而为了寻求更高的收入,可能在更大的范围内寻找适合自己的工作;另一方面,一般性工作对教育水平要求不高,在空间上分布更为普遍,而要求职工有较高教育水平的工作相对较少,因而在空间上分布更为分散,这也增加了高教育水平居民的通勤时耗[20]。
较高的人口密度是中国城市建成环境有别于西方的重要特征之一,其与个体通勤时耗的关系尚存争议,已有研究也忽视了人口密度的作用可能依赖于其他建成环境要素和个体通勤方式的作用。本文基于上海个体样本的实证结果发现： 中国城市高人口密度总体上增加居民通勤时耗; 中国城市人口密度对通勤时耗的作用是有条件的,主要依赖于“设计” “地铁站可达性”“到就业中心距离”和“通勤方式”,具体而言,提高街道路网密度和高速路密度,缩短居住地到就业中心距离,提高地铁站密度,鼓励居民采用地铁方式通勤,减少小汽车使用,都可以显著降低人口密度对通勤时耗的增加作用。
鉴于通勤时耗对于人口密度及其条件效应变化的敏感,通过优化城市建成环境和调节人口密度来减轻拥堵带来的通勤时耗过长是可行的。因此,要把城市空间规划与交通问题解决统筹考虑。首先,要明确高密度发展是促进紧凑而不是过度集中,要保持适度规模,人口密度过大的城市应适机向多中心转型。其次,尽管实现完全的职住平衡有很大难度,但职住平衡仍有助于缓解长距离通勤带来的高通勤时耗,在具体实施中应注意数量平衡和结构平衡并重。第三,坚持加大轨道交通和高速路建设力度,鼓励地铁方式通勤,在当前国家内需不足,急需投资刺激的大背景下,建设轨道交通和高速路正当时机。
本研究为人口密度与通勤时耗的关系提供了新的证据,并首次检验了人口密度对居民通勤时耗作用的条件效应,是深入理解中国建成环境与个体通勤时耗关系的有益尝试。本文基于上海市“80后”问卷调查数据分析,结论的推广性还需要更多的证据支撑。
The authors have declared that no competing interests exist.
| [1] |
城市通勤时耗的空间结构影响因素——基于文献的研究与启示 [J].Impacts of urban spatial structure on urban commuting duration: based on literature review . |
| [2] |
Commuting trends in US cities in the 1990s [J].https://doi.org/10.1177/0739456X09331549 URL [本文引用: 1] 摘要
This article explores changes in commuting trends seen in cities in the United States through the 1990s. The authors focus on factors that might explain the increase in commuting times that was experienced in the 1990s after prior decades of relative stability. They discuss traditional demographic variables (population growth and densities) and transportation variables (e.g. road capacity and transit use), but determine that their overall impact was small. They propose that the increase in commuting times was affected by strong income growth in the late 1990s. This increase in income is associated with an increase in non-work vehicle miles traveled; these affect commuting times because many non-work trips take place in peak hours.
|
| [3] |
职住均衡能够缓解交通拥堵吗?——基于GIS缓冲区方法的上海实证研究 [J].https://doi.org/10.16361/j.upf.201705011 URL [本文引用: 1] 摘要
职住失衡被认为是城市交通拥堵恶化和通勤时耗增加的重要原因之一,但也不乏相反观点.在理论方面,赞成者和质疑者都提出了各自的依据,在实证研究方面也得出了不同的结果.这可能源于不同研究在对象、方法和数据上的差异.其中,职住均衡测度的准确性是影响研究结果可信度的重要方面,国内已有文献都是基于既定的行政空间单元测度职住均衡关系,空间单元面积不等、边界固定的性质严重影响了测度的准确性.为了克服这一问题,采用GIS缓冲区技术,基于问卷调查数据,以上海外环高架线以内的102个街道为研究样本,准确地测度每个缓冲区职住均衡程度,并分析了与通勤时耗的相关性.结果表明,职住均衡确实是影响通勤时耗的显著因素;当职住比率在1.05-1.2左右时,就业者和居住者的通勤时耗最短.因而,城市规划政策中应坚持职住均衡的原则,合理布局居住和就业用地.
https://doi.org/10.16361/j.upf.201705011 URL [本文引用: 1] 摘要
职住失衡被认为是城市交通拥堵恶化和通勤时耗增加的重要原因之一,但也不乏相反观点.在理论方面,赞成者和质疑者都提出了各自的依据,在实证研究方面也得出了不同的结果.这可能源于不同研究在对象、方法和数据上的差异.其中,职住均衡测度的准确性是影响研究结果可信度的重要方面,国内已有文献都是基于既定的行政空间单元测度职住均衡关系,空间单元面积不等、边界固定的性质严重影响了测度的准确性.为了克服这一问题,采用GIS缓冲区技术,基于问卷调查数据,以上海外环高架线以内的102个街道为研究样本,准确地测度每个缓冲区职住均衡程度,并分析了与通勤时耗的相关性.结果表明,职住均衡确实是影响通勤时耗的显著因素;当职住比率在1.05-1.2左右时,就业者和居住者的通勤时耗最短.因而,城市规划政策中应坚持职住均衡的原则,合理布局居住和就业用地.
|
| [4] |
Urban land uses, socio-demographic attributes and commuting: A multilevel modeling approach [J].https://doi.org/10.1016/j.apgeog.2011.02.001 URL [本文引用: 1] 摘要
Studies on the interrelationship between land use and travel behavior are often at some aggregate geographic unit such as the census tract, traffic analysis zone (TAZ), or the zip code level, and results have been inconclusive. This paper uses a multi-level modeling approach to examine the combined effects of land use types and socio-demographics (including both individual and neighborhood attributes) on commuting. Data used in this research include the Baton Rouge Personal Transportation Survey (BRPTS) of individual households and the Census Transportation Planning Package (CTPP) at the TAZ level. Land uses include agricultural, commercial/office and residential of various densities. A localized job-housing balance ratio is computed to measure job accessibility and also to capture the degree of mixed land uses around each residential location, and its effect on commuting behavior is examined. Commuting is measured by both travel distance and time in order to capture the spatial separation of residence and workplace as well as road conditions. The results indicate that land use types are significant in explaining commuting time, and socio-demographic characteristics of both individuals and neighborhoods play an important role in shaping individuals鈥 commuting behavior.Highlights? This paper uses a multilevel modeling approach to explain individual commuting. ? Models examine the combined effects of land use and socio-demographics of individuals and neighborhoods. ? Land use measures include land use types, proximity to a high-performing school and job-housing balance. ? Individual attributes include gender, race, income, education, presence of children, job status, and multi-worker household. ? 14 neighborhood socio-demographic variables are consolidated into factors.
|
| [5] |
The impact of metropolitan structure on commute behavior in the Netherlands: A multilevel approach [J].https://doi.org/10.1111/j.1468-2257.2004.00251.x URL [本文引用: 1] 摘要
This paper documents the investigation of the impact of metropolitan structureon the commute behavior of urban residents in the Netherlands. Not only has the impact of monocentrismversus polycentrism been analyzed, but the influence of metropolitan density and sizehas also been considered, together with the ratio of employment to population and the growth ofthe population and employment. Furthermore, data are used at a variety of levels of analysisranging from the individual worker to the metropolitan region rather than being drawn from aggregatelevel statistics alone. Multilevel regression modeling is applied to take account of the interdependenciesamong these levels of aggregation. With regard to mode choice, the results indicatethat the probability of driving an auto to work is lower in employment-rich metropolitan regions,and rises as the number of jobs per resident has grown strongly. Furthermore, women in mostpolycentric regions are less likely to commute as an auto driver. All else being equal, commutedistances and times for auto drivers are longer in most polycentric regions than in monocentricurban areas. In addition, commute time as an auto driver rises with metropolitan size, whereascommute distance depends on employment density and the growth of the number of jobs per resident.The investigation shows that metropolitan structure, although significantly influencingcommute patterns, explains only a small part of the variation of individuals鈥 commute behavior
|
| [6] |
Impact of the jobs-housing balance on urban commuting in Beijing in the transformation era [J].https://doi.org/10.1016/j.jtrangeo.2009.09.008 URL [本文引用: 2] 摘要
This empirical study examines the impact of the jobs-housing balance on individual commuting time in Beijing in the period of transformation of the Chinese economy and society. The results of the analysis show that the jobs-housing balance has a statistically significant association with a worker’s commuting time when the factors of transport accessibility, population density and worker’s socioeconomic characteristics are controlled. The higher the jobs-housing balance, the shorter the worker’s commuting time. The finding suggests that the jobs-housing balance still has significant implications for commuting time, although the recent market-oriented reforms in housing are changing the jobs-housing balance in the danwei system that prevailed in the socialist era. As the housing markets are imperfect, with strong government intervention in Beijing, the finding implies that the co-location hypothesis – which believes development management would create ‘barriers’ to a jobs-housing balance and increase commuting time – needs to be rethought before it can be generalized and applied to China’s cities. The results of the analysis also show that the workers living in danwei housing still have shorter commuting time. The finding indicates that the housing marketization is likely to induce a local jobs-housing imbalance and thereby increase commuting time. In this sense, a deterioration in the jobs-housing imbalance and increased commuting time in Beijing may owe much to the adoption of market-based housing supply.
|
| [7] |
The impact of the built environment on individual workers' commuting behavior in Beijing [J].https://doi.org/10.1080/15568318.2012.692173 URL [本文引用: 1] 摘要
The increasing emission of transport-related pollutants has become a key issue in relation to climate change mitigation and the improvement of air quality in China's cities. This article aims to examine the effects of changes in the built environment on transportation by examining the case of Beijing. Looking at household survey data, the analysis found that individual workers鈥 commuting behavior (concerning travel destination, mode choice and travel time) is significantly related to some aspects of the built environment when socioeconomic and demographic characteristics are taken into account. There are obvious differences in the effects of the built environment on commuting across income groups, occupations and industries.
|
| [8] |
The paradox of intensification [J].https://doi.org/10.1016/j.tranpol.2010.05.007 URL [本文引用: 1] |
| [9] |
Car use in Netherlands daily urban systems: Does polycentrism result in lower commute times? [J].https://doi.org/10.2747/0272-3638.24.5.410 URL [本文引用: 1] 摘要
The effect on travel time of the decentralization of land use is an unresolved issue in the literature on travel behavior. Here we report our testing of the contention that commute times are lower in polycentric than in monocentric urban systems with data from the 1998 Netherlands National Travel Survey. Multilevel regression analysis indicates that the role of urban form variables in the explanation of commute times is limited in comparison with micro-level personal and household characteristics. Car commute times are shown to be higher in most poly-centric systems. In addition, car commute times tend to increase with higher residential densities and with the size of daily urban systems (DUS). The results suggest that the relocation of residences and employment in the Netherlands has not by and large led to lower commute times. One reason for this might be the strong spatial planning tradition in the Netherlands.
|
| [10] |
Spatial-temporal characteristics and factors influencing commuting activities of middle-class residents in Guangzhou City, China [J].https://doi.org/10.1007/s11769-016-0806-1 URL [本文引用: 3] 摘要
The middle class in metropolitan Chinese cities has become an important social group. With the rapid development of urbanization and constant advancement of suburbanization, the middle class has increasingly come to influence city traffic. Research into middle-class commuting activities thus has practical significance for improving traffic congestion and reducing the commuting burden in metropolitan cities. Based on a dataset formed by 816 completed surveys, this paper analyzes the commuting mode, time and distance of middle-class residents in Guangzhou City using the descriptive statistical method. The results indicate that private cars are the main commuting mode, followed by public transport. Meanwhile, middle-class residents mainly undertake medium-short time and medium- short distance commuting. The study subsequently uses multilevel logistic regression and multiple linear regression models to analyze the factors that influence commuting mode choice, time and distance. The gender, age, number of family cars, housing source and jobs-housing balance are the most important factors influencing commuting mode choice; housing, population density, jobs-housing balance and commuting mode significantly affect commuting time; and transport accessibility, jobs-housing balance and commuting mode are the notable factors affecting commuting distance. Finally, this paper analyzes what is affecting the commuting activities of middle-class residents and determines the differences in commuting activity characteristics and influence factors between middle-class and ordinary residents. Policy suggestions to improve urban planning and urban management are also proposed.
|
| [11] |
城市建成环境对居民通勤行为的影响——上海实证研究[D] .Impact of urban built environment on the residential commuting behavior: the case of Shanghai . |
| [12] |
Density and the journey to work [J].https://doi.org/10.1111/grow.1997.28.issue-2 URL [本文引用: 1] 摘要
This study analyzes the effect of accessibility to jobs and houses at both the home and work ends of trips on commuting duration for respondents to a household travel survey in metro-politan Washington, DC. A model is constructed to estimate the effects of demographics and relative location on the journey to work. Analysis finds that residences in job-rich areas and workplaces in housing-rich areas are associated with shorter commutes. An implication of this study is that, by balancing accessibility, the suburbanization of jobs maintains stability in commuting durations despite rising congestion, increasing trip lengths, and increased work and non-work trip making. 0 1998 Elsevier Science Ltd. All rights reserved
|
| [13] |
Transport impacts of clustered development in Beijing: Compact development versus overconcentration [J].https://doi.org/10.1177/0042098011410336 URL [本文引用: 1] 摘要
This research aims to inform the compact city discussion with a case study of Beijing, where urban planning has emphasised clustered suburban development in the past half-century. It uses three decades of census data to describe Beijing’s spatial development trajectory and a household survey to assess its transport impacts. The research reveals an overconcentration of urban activities as a result of the featureless expansion of the central built-up area and the absorption of the suburban clusters; and, a lengthened commuting time stemming from the observed spatial development pattern. Beijing’s experience adds to the existing literature by informing the search for good city forms in urban areas of high density. It is essential to differentiate compact development from overconcentration when combating sprawling development. Developing and maintaining suburban nodal characteristics around public transit can reduce travel in high-density urban areas.
|
| [14] |
The Death and Life of Great American Cities . |
| [15] |
Travel behavior issues related to neo-traditional developments-a review of the research [C]// |
| [16] |
Travel behavior as function of accessibility, land use mixing, and land use balance: evidence from San Francisco Bay Area [J].https://doi.org/10.3141/1607-16 URL [本文引用: 1] 摘要
No abstract provided.
|
| [17] |
Transportation-land-use interaction: empirical findings in North America, and their implications for modeling [J].https://doi.org/10.1016/S1361-9209(99)00036-X URL [本文引用: 1] 摘要
The last two decades have witnessed the implementation of various policies based on land-use to address problems stemming from automobile ownership and use. There are, however, questions with respect to the efficacy of such policies. This paper therefore reviews the literature on empirical studies of the transportation–land-use interaction with the objective of identifying the current state of knowledge concerning the interactions, particularly with respect to the impact land-use policies are likely to have on the system. The focus is on studies conducted in North America. The results are mixed; some studies conclude that urban densities, traditional neighborhood design schemes, and land-use mix have an impact on auto ownership and use. Other studies find the impact of such variables to be at best marginal. Gaps in our understanding of the interaction are identified. These are found to be primarily the result of data limitations and methodological weaknesses. A detailed discussion of the implications of the findings for the development and application of integrated transportation–land-use models is provided, with the recommendation that without such an integrated approach to analyzing the transportation–land-use interaction, any study of impacts of urban form on travel behavior is likely to yield erroneous results.
|
| [18] |
Role of the built environment on mode choice decisions: additional evidence on the impact of density [J].https://doi.org/10.1007/s11116-007-9153-5 URL [本文引用: 1] |
| [19] |
Walking, obesity and urban design in Chinese neighborhoods [J].https://doi.org/10.1016/j.ypmed.2014.10.002 URL PMID: 25451324 [本文引用: 1] 摘要
Abstract OBJECTIVE: We examined the connections (1) between the design of the built environment and walking, (2) between the design of the built environment and obesity, and (3) between walking and obesity and income in urban settings in China. METHODS: Six neighborhoods with different built environment characteristics, located in the Chinese cities of Shanghai and Hangzhou, were studied. Data on walking and other physical activity and obesity levels from 1070 residents were collected through a street intercept survey conducted in 2013. Built environment features of 527 street segments were documented using the Irvine-Minnesota Inventory-China (IMI-C) environmental audit. Data were analyzed using the State of Place64 Index. RESULTS: Walking rates, household income and Body Mass Index (BMI) were related; neighborhoods with a higher State of Place64 Index were associated with higher rates of walking. CONCLUSION: This study began to establish an evidence base for the association of built environment features with walking in the context of Chinese urban design. Findings confirmed that the associations between "walkable" built environment features and walking established in existing research in other countries, also held true in the case of Chinese neighborhoods. Copyright 08 2014 Elsevier Inc. All rights reserved.
|
| [20] |
Car use in Netherlands daily urban systems: does polycentrism result in lower commute times? [J].https://doi.org/10.2747/0272-3638.24.5.410 URL [本文引用: 2] 摘要
The effect on travel time of the decentralization of land use is an unresolved issue in the literature on travel behavior. Here we report our testing of the contention that commute times are lower in polycentric than in monocentric urban systems with data from the 1998 Netherlands National Travel Survey. Multilevel regression analysis indicates that the role of urban form variables in the explanation of commute times is limited in comparison with micro-level personal and household characteristics. Car commute times are shown to be higher in most poly-centric systems. In addition, car commute times tend to increase with higher residential densities and with the size of daily urban systems (DUS). The results suggest that the relocation of residences and employment in the Netherlands has not by and large led to lower commute times. One reason for this might be the strong spatial planning tradition in the Netherlands.
|
| [21] |
复旦大学社会科学数据研究中心.“80后”的世界——长三角社会变迁调查结果发布 [EB/OL].The world of "the post-80s"—the release conference of Yangtze River Delta Social Transformation Survey . |
| [22] |
中国人才网.2015全国50城市上班族通勤距离及用时排行出炉 [EB/OL]. 中国人才网,China Rencai. |
| [23] |
Built environmental impacts on commuting mode choice and distance: evidence from shanghai [J].https://doi.org/10.1016/j.trd.2016.06.001 URL [本文引用: 1] 摘要
The impacts of the built environment characteristics in residential neighborhoods on commuting behavior are explored in the literature. Scant evidence, however, is provided to scrutinize the role of the built environment characteristics at job locations. Studies also overlooked the potential error correlations between commuting mode and commuting distance due to the unobserved factors that influence both variables. We examined the impacts of the built environment characteristics at both residential and job locations on commuting mode and distance, by applying a discrete-continuous copula-based model on 857 workers in Shanghai. In contrast with studies of Western countries, we showed residential built environment characteristics are more influential on commute behavior than the built environment characteristics at job locations. This suggests the importance of local specificity in policymaking process. We also found the proportion of four-way intersections, road density, and population density in residential areas are negatively associated with driving probability, with elasticity amounts of 611.00, 610.23, and 610.08, respectively. Hence, dense and pedestrian- and cyclist-oriented development help to reduce travel distance and encourage walking, biking, and transit modes of travel.
|
| [24] |
Spatial Dimensions of Urban Commuting: A Review of Major Issues and Their Implications for Future Geographic Research [J].https://doi.org/10.1111/j.0033-0124.2004.05602002.x URL [本文引用: 1] 摘要
Modern urban transportation systems continuously challenge, and are challenged by, the changing nature of 21st-century travel demand. Today, congestion is the norm in cities of the United States, and researchers and practitioners are seeking solutions to these problems. Urban commuting is identified as contributing to the suboptimal performance of transportation systems. This paper offers a review and critique of recent research on urban commuting, emphasizing geographical dimensions of this topic. Three broad areas of research related to urban commuting are discussed. These include (1) urban sustainability, (2) land use, and (3) geographic information systems (GIS). Major themes are examined in an effort to elicit thought on future geographic research. At the conclusion of the paper, summary remarks are provided and avenues for research are outlined.
|
| [25] |
居民户外活动影响因素的实证研究 [J].https://doi.org/10.3969/j.issn.1006-3862.2014.09.009 URL [本文引用: 1] 摘要
促进居民规律性的户外活动是提高全民身体素质的重要途径,并有利于节约能源、增加社区活力。研究以河北省石家庄市500余个家庭样本和500余个居民个体的调查为依据,采用定序回归方法,从个人、家庭和小区建成环境三个维度分析影响居民户外活动的各类因素。研究发现:居民性别、年龄、户外活动态度,家庭收入、家庭结构,小区周边路网密度、服务设施混合度,小区建筑密度和容积率等,与居民户外活动强度有显著相关性。为此,建议:重视户外活动的宣传教育,提高全民健康水平;提倡高密度、小网格的道路系统设计,增强土地利用的混合度;合理控制建筑密度和容积率,提高小区户外活动空间的便利性和舒适性。
article-title>An empirical study of the factors associated with residents’ outdoor activities .https://doi.org/10.3969/j.issn.1006-3862.2014.09.009 URL [本文引用: 1] 摘要
促进居民规律性的户外活动是提高全民身体素质的重要途径,并有利于节约能源、增加社区活力。研究以河北省石家庄市500余个家庭样本和500余个居民个体的调查为依据,采用定序回归方法,从个人、家庭和小区建成环境三个维度分析影响居民户外活动的各类因素。研究发现:居民性别、年龄、户外活动态度,家庭收入、家庭结构,小区周边路网密度、服务设施混合度,小区建筑密度和容积率等,与居民户外活动强度有显著相关性。为此,建议:重视户外活动的宣传教育,提高全民健康水平;提倡高密度、小网格的道路系统设计,增强土地利用的混合度;合理控制建筑密度和容积率,提高小区户外活动空间的便利性和舒适性。
|
| [26] |
上海都市区就业-人口空间结构演化特征 [J].Spatial distribution and structure evolution of employment and population in Shanghai Metropolitan Area . |
| [27] |
Shaping carpool policies under rapid motorization: the case of Chinese cities [J].https://doi.org/10.1016/j.tranpol.2011.03.005 URL [本文引用: 1] 摘要
Rapid motorization and fuel cost hike over the past few years have made carpool a new mode of travel in Chinese cities. But transportation policy makers have been rather ambivalent, if not indifferent, about carpool. Unlike cities in highly motorized societies, little is known about carpooling behavior in emerging economies such as China. This paper provides an initial discussion of carpooling in China by exploring a series of questions. What are the current practice and issues of carpool in Chinese cities? How do carpools in China compare with those in the motorized Western cities? Can carpools help Chinese cities mitigate the negative impacts of rapid motorization? Are foreign policies such as High-Occupancy-Vehicle (HOV) lanes transferable to China? Acknowledging the social benefits of voluntary carpooling, this paper argues: (1) bus lanes may be a better choice than HOV lanes when converting general motor vehicle lanes; (2) policies subsidizing carpoolers cannot be justified on either efficiency or equity grounds because a marginal carpooler is more likely transitioning from a transit user or non-motorized traveler than from a driver. Policy suggestions are proposed to Chinese decision makers.
|
| [28] |
Urban spatial structure and commute duration: an empirical study of China [J].https://doi.org/10.1080/15568318.2015.1042175 URL [本文引用: 1] 摘要
Urban traffic is embedded in and fundamentally shaped by the spatial pattern of urban land use, such as city size, density, extent of polycentricity, and the relationship between employment and residential locations. Previous evidence, mainly from European and American cities, suggests that the duration of commute trips increases with city size and the spatial separation between jobs and housing. On the other hand, the influences of density and polycentricity are less clear. Using data from 164 cities in China, this study empirically analyzes the relationship between city average commute duration and multiple dimensions of urban spatial structure. Controlling for economic, demographic, and infrastructure characteristics, the authors find that commute duration correlates positively with city size and jobs鈥揾ousing separation but negatively with density and polycentricity. As one of the earliest studies on commute cost in the rapidly urbanizing and motorizing Chinese cities, this study can help Chinese decision makers improve urban economic and environmental efficiency through spatial planning and policy making. Specifically, compact, mixed-use, and polycentric spatial development may ease the burden of commute, and thus substitute for unnecessary infrastructure investment and energy consumption during a period of rapid urban expansion in China.
|
| [29] |
Analysis of travel time and mode choice shift for non-work stops in commuting: Case study of Beijing, China [J].https://doi.org/10.1007/s11116-016-9749-8 URL [本文引用: 1] 摘要
This paper investigates travel time attributed to non-work stops in multi-purpose commuting trips. Travel time is explained by socio-demographics and spatial attributes, mode shift is also included to
|
| [30] |
Factors influencing commute distance: a case study of Toronto’s commuter shed [J].https://doi.org/10.1016/j.jtrangeo.2011.10.005 URL [本文引用: 1] 摘要
This study examines factors that influence commute distance within the commuter shed of Toronto, Canada. The investigation uses data drawn from the 2006 Census of Canada Master File (20% sample). The findings from two multiple linear regression models indicate that being a recent migrant and/or living in the most accessible rural areas of Toronto’s commuter shed substantially increases commute distance. At the same time, employment type, sex, and age, among other factors, are also found to significantly influence commute distance.
|
| [31] |
西部大城市居民通勤特征及其与城市空间结构的关系研究——以成都市为例 [J].The characteristics of resident commuting and its relationship with urban spatial structure in large cities of western china: A case study of Chengdu . |
| [32] |
上海城市建成环境对居民通勤方式选择的影响 [J].https://doi.org/10.11821/dlxb201510010 URL [本文引用: 1] 摘要
伴随中国快速城市化与机动化进程,私人汽车拥有量不断增长,由此引起的交通拥堵和环境问题已成为制约中国城市可持续发展的难题.基于上海市区的居民通勤问卷调查数据,采用多项Logit模型检验了街道尺度城市建成环境对于居民通勤方式选择的影响,结果表明,在控制了其他因素后,提高居住地的人口密度、土地利用混合度与十字路口比重,可以减少小汽车通勤方式的选择,而就业地建成环境对居民通勤方式选择影响相对较弱;建成环境对通勤方式选择的影响会因个体的社会经济异质性而不同.这些结论为通过优化土地利用规划来优化居民通勤结构的城市交通和城市规划政策提供了启示.
Impact of urban built environment on residential choice of commuting mode in Shanghai .https://doi.org/10.11821/dlxb201510010 URL [本文引用: 1] 摘要
伴随中国快速城市化与机动化进程,私人汽车拥有量不断增长,由此引起的交通拥堵和环境问题已成为制约中国城市可持续发展的难题.基于上海市区的居民通勤问卷调查数据,采用多项Logit模型检验了街道尺度城市建成环境对于居民通勤方式选择的影响,结果表明,在控制了其他因素后,提高居住地的人口密度、土地利用混合度与十字路口比重,可以减少小汽车通勤方式的选择,而就业地建成环境对居民通勤方式选择影响相对较弱;建成环境对通勤方式选择的影响会因个体的社会经济异质性而不同.这些结论为通过优化土地利用规划来优化居民通勤结构的城市交通和城市规划政策提供了启示.
|
/
| 〈 |
|
〉 |
