Factors Influencing Residential Location Choice towards Mixed Land-Use Development: An Empirical Evidence from Pakistan

Abstract

This study is aimed to determine the MLU development factors by executing a Delphi Method (DM). The MLU factors can contribute to the land-use development process in the thickly populated urban centers of developing countries. This is the first study of its type conducted to clarify MLU development factors in urbanized settlements of Sindh province, Pakistan. Karachi and Hyderabad are thickly populated cities in Pakistan where mixed land-use (MLU) development prevails over the years. The DM was attempted in two rounds focusing on the opinion of urban development specialists and academic experts. The experts initially provided a set of forty-two factors identified from the literature. These factors were arranged in a Likert-based questionnaire and determined through the coefficient variation. The prominent factors were identified as household savings, travel costs and low rent, nearby household items and shared utility services, economic vitality, variety in purchasing grocery and shopping items, demographic change and social poverty, accessibility to nearby public services, reduction in commuting time and easy access to restaurants. This proposed research recommends policy implications focusing on identified key parameters of MLU development, such as low carbon exposure, livable environment, and planned municipal system.

1. Introduction

The mixed land-use (MLU) development needs time in congested urban centers. MLU is one of the modern planning paradigms, which has its origin in Jane Jacob’s critique that “Fine grain mixing of diverse uses create a vibrant and successful neighborhood”. MLU is a key feature in many of today’s urban development concepts, such as smart growth, transit-oriented development, walkable, compact cities, etc. [1]. The physical extent of cities has accelerated over the past century, and more than 50% of the global population now lives in urban areas [2]. This rapidly increasing urban population creates chaos for development authorities. If urban areas expand horizontally, it is difficult for development authorities to provide utilities in scattered areas. In sprinkled urban areas, there is also a lack of economic opportunities, reliance on the private mode of transport, and social isolation for the residents. Land development, such as housing and infrastructure, is largely carried out by private individuals and developers, whose practices are not different from the simple form of rational economic behavior and allocated performance. Urban planners and developing authorities rationally support the expansion of urban areas in MLU form [3]. The vertical growth with the mixed form of land-use areas to cope with the housing backlog has also been considered [4].

MLU development was regarded as a desirable urban livability and public health pattern. The congress of new urbanism promoted the closer integration of residential, commercial, and recreational uses promoted by the congress of new urbanism in 2001 and the smart growth network in 2006. A greater MLU can promote active travel, reduce auto-dependence [5], and help to build a healthy community [6]. Advocates for MLU have argued that land-use segregation results in longer travel time, traffic congestion [7], air pollution, improper use of energy, open space, and loss of greeneries [8]. It relies on an unequal distribution of economic resources, housing imbalances, and losses that segregated societies. Thus, MLU development can be considered a solution to the problems caused by scattered development [9]. MLU development patterns have been followed by numerous countries, including Pakistan [10].

While the rapid changes in land use activities in Karachi and Hyderabad demand to accommodate many inhabitants, the MLU development has emerged as a solution for urban areas’ residents. To understand and promote viability based on determining factors deeply about MLU development, this study identifies the core factors responsible for MLU development. The factors can be used not only for selected areas but also where rapid changes in land use occur. It is evident from the available literature that other researchers worked on different factors separately such as physical [11,12,13], economic [4,11,14], social [1,9,15,16,17], and environmental [5,9,11,14,17,18,19,20,21,22,23]. It is pertinent to mention that only two studies were carried out in Pakistan on Mixed land use development [10]. Those studies had covered only limited factors like; reduction in commuting and congestion cost factors in different studies. Therefore, it is the novelty and contribution of this study that all physical, economic, social, and environmental are employed in this research. It is also evident from the available literature that the previous studies conducted their research on one city, whereas this study is based on comparing the two cities. The methodology is based on Delphi Method (DM) iterations to reach a consensus among experts on the most influencing factors causing MLU developments. The Delphi Method is a very famous tool that uses stakeholders’ opinions and requires consensus [24]. Therefore, this study has taken the opinion of the experts and practicing professionals through two iterative rounds of questions. In the first round, both open-ended and close-ended questions were asked of the respondents. Because of this collocational data strategy, we used the Delphi Method. The DM method is never used in MLU factor identification studies. The DM was initialized through a set of 42 factors resulting from an extensive literature search, and after two iterations of DM on experts with their own technical opinion also counted therein, the consensus was established for a smaller set of factors. The determination of the most responsible factor was based on the statistical parameter, the coefficient of variability, and reliability to rate and shortlist the factors.

The present study is organized as follows, Section 1, briefly describes the importance of MLU development, focusing on the study area, whereas the MLU features of the developing countries and study area, together with the study purpose and its significance, are been highlighted. Section 2, presents an extensive literature review to deeply illustrate the concepts and models used in MLU development over the years across different countries, mainly focusing on Pakistan. Additionally, the Delphi Method concepts and the rationale of the experts-based survey were adopted in this study. In Section 3, the details of the underlying method and techniques used and the list of identified factors, which encourage the inhabitants toward MLU developments. In Section 4, results and discussion of DM are illustrated in detail factor by factor as per their importance in the form of statistical parameters. In the end, detailed conclusions are given with a set of recommendations.

2. Literature Review

The idea of land use integration began to emerge in Roman cities from 700 BC to 1900 AD, considering the combination of residential and non-residential activities. However, with the dramatic increase in car occupations after World War II, issues of land use coherence began to emerge, and its principles were scrapped from all new projects [25]. MLU has become one of the key planning principles in contemporary planning strategies [9]. The combination of commercial facilities, single-family residences, and polygamous residences has given residents access to nearby locations over centuries. In the United States, this long tradition was broken with the advent of Euclidean zoning, which allowed locals to segregate land use to protect health by separating hazardous industrial areas from residential areas [11].

Advocates for MLU have argued that land-use segregation results in longer travel time, traffic congestion, air pollution, improper use of energy, open space, and loss of greeneries. It relies on an unequal distribution of economic resources, housing imbalances, and losses, which distinguish society. MLU is an antidote to the problems caused by the spread of citizens [9]. As mapped in

Table 1. List of Identified Mixed land use Factors.

S. No.	Factors Driving Mixed Land Uses	Empirical Evidence
Physical
1	Land Use Diversity/mix forms of housing	[1,4,11,13,15,16,19,20,26,27,28,29]
2	Land use intensity due to urbanization (vibrant land use features in mixed land use areas)	[1,4,9,13,25,26,27,29]
3	Low quality of dwellings	[4,19]
4	Inexpedient transit linkages	[1,4,5,11,16,25]
5	Synergy effects (combined)/Compact features of land use	[13,15,25]
6	The demand for livable developments (easy access to restaurants and other fast food franchises, etc.)	[9,15,18,25]
7	Low-skilled labor force (due to the absence of utility services in suburban areas)	[4,19]
8	Demand for walkable communities/Bicycling modes (reduction in commuting time, cost affects the demand for MLU)	[1,4,10,11,13,14,19,26,29,30,31]
Economic
9	Economic vitality	[4,11,18,20,25,32]
10	Increase in Infrastructure and Fuel Costs	[1,4,21,25,26,31]
11	Stimulation of Local Economy	[17,18,22]
12	Automobile Dependence	[1,13,14,21,23,25,26,30,31,33]
13	Detachment from jobs	[4,5,9,19,33]
14	Employment is one of the factors that people are getting more focused on MLU due to the generation of more employment opportunities	[4,5,9,19,25,33,34,35]
Social
15	Lack of Accessibility to nearby public services	[1,9,19,29,33]
16	Low Public Health	[4,11,26]
17	Lack of Spaces and activity-oriented destinations	[19,25]
18	Need for safer and more vibrant neighborhoods.	[17,18,22,27,28,30,31]
19	Demographic change and Social Poverty	[1,18,19,26]
20	Increase in travel need	[4,11,13,15,16,19,25,26,33]
21	MLU enhances the Sense of Place	[13,14,31,36]
Environmental
22	Degrading Environmental impacts	[1,14,18,25]
23	Energy Consumption	[9,15,25]
24	Pollution and Traffic Congestion	[5,9,10,11,14,15,23,26,31,33,34,35]
25	More Noise occurrence in segregated developments leads societies to focus on MLU	[17,22,37]
26	The revolutionary shift in Industrialization is a major driving factor for MLU	[11,25,26]
27	MLU may reduce property crime	[16,17,18,22,27,28,38,39,40,41,42,43,44]
28	Supporting Social equality/social differences/local community	[13,17,18,22]
29	Improving health and well being	[13,17,18,22,30,31]
30	MLU promotes Pedestrian Activity	[13,14,17,18,22,29,30,31,33]
31	MLU is influenced by Violent crime considering the nearby neighborhood	[27,39]
32	Property Values tend to fall with proximity to Mixed Land use	[13,31,45,46,47,48,49,50]
33	MLU improves Locational Convenience	[13,31,46,49]
34	Reduction in commuting time affects the demand for MLU	[13,33,51,52,53]

, several factors drive societies toward MLU, as extracted from previous studies and informal interviews with stakeholders. Factors were identified from previous literature and extracted for this study. As Table 1 describes, it provides strong support to the design of the research tool and finding research gaps for achieving objectives by bringing major factors at one point. These are according to previous data studied individually, considering the factors leading to mixed land use as today is a key feature in many urban development concepts, such as smart growth, transit-oriented development, walkable community, compact city, etc. Mixed-use of land is one of the modern examples of planning among P-Gosh’s witnesses: “Mixing pieces from different uses creates vitality and success in the neighborhood” [1]. Mixed-use walking convenience was considered an essential ingredient for a healthy environment. Mixed or diverse land use is one of the ‘3Ds’—density, pedestrian-friendly design, and diversity associated with walking [11]. Mixed use is part of a sustainable development strategy, as well as the goal of economic vitality, social equality, and environmental standards, as well as the idea of better urbanization [18]. Mixed-use increases housing stock facilitates transportation, and reduces costs in a tight housing market. More people in the cities are looking to adopt an urban lifestyle and take public transport, trying to make their short lives easier.

The mixed-use can be described as a fine combination of basic land use—a wide range of residences and workplaces, in terms of all other ancillary services, with outstanding housing, allows easy walking distance to the houses. The combination of land use can be described in three dimensions: (i) increasing the intensity of land use with different arrangements and duration; (ii) enhancing application diversity by promoting integration; (iii) combining separate applications [1]. In addition, a model considering socio-economic and physical dimensions was designed by Rowley, Hoppen Browuer, and Louw, which has been promoted by many other researchers. The mixed-use model can be classified on different parameters according to its typology and affects the surroundings as proposed by P. Ghosh’s [1]. Various issues are studied in this regard, and the models for considering the best implementation and addressing the policies for planning mixed land use in India. Different categories studied for the mixed land-use model covered are physical, socio-economic, income and lifestyle, commercial and industrial activities, facilities, temporal, land-use and open spaces, premises, and street markets. Rowley’s Model and Hoppen Brouwer and Louw’s Model [15,19] mixed-use models give the parameters based on which the mixed land uses can be categorized.

The empirical findings of one of the case studies, as shown in Table 1, are stated as follows.

The researcher seeks to emphasize the institutional policy reforms in mixed land use for urban planning for residential and commercial purposes. To measure the mixed use of land, this study aims to quantify the mixed use of urban land for commercial and residential purposes in major cities of Pakistan. The data gathered by the Urban Unit of Pakistan and some earlier studies examined the existing level of mix-use land utilizing data for commute and congestion costs. The results indicate that Rahim Yar Khan is at the bottom of the mix-land use scale, whereas Lahore is at the top. Additionally, in order to determine the hypothesis of research whether mixed land use leads to reduction and commuting cost and congestion costs, a survey was conducted in the two markets of the same area of Islamabad. One market had the construction and design of mixed land uses, and the other had commercial use only. The exploratory data analysis and non-parametric analysis of the survey technique are used. The results identified a significant reduction in commuting and congestion costs in mixed land-use development. The conclusion suggests that there is a great need for institutional reforms regarding mixed land use in big cities of Pakistan [10].

2.1. Rowley’s Model for Mixed Land Use Typology

The pattern of Rowley’s typology, as shown in Figure 1, does not consider the use of high-rise or building-level mixes but is referred to as an aspect of the internal structure of settlements with basic skills in the structure, scale, location, and layout of citizens. Density, grains, permeability building blocks, districts, roads, inner-city areas, suburbs, and greenfield locations that use engineered viability.

2.2. Hoppen Brouwer and Louw’s Mixed-Use Model

Mixed land-use attributes in Hoppen Brouwer (HB) model, as in Figure 2, based on function, dimension, scales, location, etc. Both models conflict in dimensional considerations.

The models were developed, and the study was conducted on the relational basis for MLU and urban decay [19] by conducting spatial analysis in ArcGIS overlying technique using ten variables for seven classes. Further, for extracting the binomial relationship and statistical investigation of the relationship applying the Pearson chi-square correlational test and the variance of this relationship applying the ANOVA test. It was found that the new approach (a combination of mixed land use and a view of the city’s inner decline) could develop a new analytical approach to the local analysis of urban degradation and performance. This study concluded that providing extensions to Rowley’s development models for mixed land use with the help of the urban renewal approach gets a new application angle.

2.3. Connotation of Mixed Land-Use

It is described as “activities carried out by the occupants on a particular piece of land employing natural resources via leveraging the finest human potentials.” It may alternatively be described as “Means used by a group of residents for the aim of obtaining their basic necessities” or “The human needs of land for living on it, utilizing it for various life functions, and increasing houses on it.”. Mixed land use is the term used to describe a plan that contains a variety of various building types with varying functions. For instance, a mix of residential buildings might be found next to office buildings, stores, movie theatres, schools, coffee shops, parks, and transportation hubs. According to some, it is a heterogeneous pattern of land use in geographically defined zones that often combines residential, commercial, institutional, industrial, recreational, and agricultural activities [25]. Three categories of mixed land use exist: horizontal, vertical, or both, as shown in Figure 3.

MLU is a sort of land use pattern that combines several land-use types, may be functionally integrated, and promotes the trend of sustainable development. This idea consists of the following four parts: (1) a range of land uses and activities, (2) a constrained physical area, (3) interactions and integration among these uses and activities, and (4) a specific development objective, such as meeting various human needs, boosting community vitality, and increasing spatial allocation [20].

3. Method and Materials

3.1. Study Area

This study is based on the mixed land use development in Sindh province, Pakistan. The study has focused only on the metropolitan cities of Sindh province, Karachi, and Hyderabad, whereas the other urban areas of Sindh province are secondary cities, small cities, and towns. As emphasized in Figure 4 and Figure 5, MLU development was increased in Karachi and Hyderabad, respectively, over the years, as shown in Figure 6 and Figure 7. For example, Karachi is an economic hub of Pakistan that rapidly sprawled over the years. The total area of Karachi is approximately 3600 sq. km. About 1300 sq. km was highlighted as a built-up area.

Karachi is one of the most densely populated cities in Pakistan. The population of Karachi is around 16,024,894 people [56,57]. It has 7 districts, 18 towns (6 cantonment areas), and 178 union councils. The city covers a total area of 3527 sq. km and is located at 24.8600 N, 67.0100 E Hyderabad is located in Sindh Province, 140 km east of Karachi; it is the second-largest city in Sindh has 2,199,928 people in Hyderabad, Pakistan [56,57,58,59,60].

Figure 6 and Figure 7 show the digitized maps of the study areas, i.e., Karachi and Hyderabad showing boundaries.

3.2. Methods

Firstly, qualitative data through empirical evidence from the literature review were extracted to obtain global identification of MLU’s factors, as shown in

Table 2. List of Assign Codes to Factors extracted from Literature and Proposed by Experts.

S No.	Extracted MLU Factors	Code Assigned	Extracted from Literature	Proposed by Experts in Local Context
Physical Factors
1	Land Use Diversity/mix forms of housing	P1	✓
2	Vibrant Land use features in mixed land use areas	P2	✓
3	Low-quality living conditions in scattered urban locations	P3	✓
4	Inexpedient transit linkages or Unavailability of transport facilities in low-density areas	P4	✓
5	Compact features of land use	P5	✓
6	Due to the absence of utility services in suburban areas	P6	✓
7	More travel opportunities in mixed land-use areas	P7	✓
8	Due to the housing backlog	P8	✓
9	People like to live near historical places around the city Center	P9	✓
10	Easy access to restaurants and other food fast franchises	P10	✓
11	Reduction in commuting time affects the demand for MLU	P11	✓
12	Availability of infrastructure in MLU areas (health and educational facilities)	P12		✓
13	The overall layout and size of a city with mixed land use suits the middle class that is in majority	P13		✓
14	Sustainable, environmentally friendly, and use of minimal designs in mixed land-use Neighborhood	P14		✓
15	A Range of Housing Opportunities and Choices	P15		✓
Economic Factors
16	Economic vitality due to commercial activities	EC1	✓
17	Increase in Infrastructure and Fuel Costs in scattered areas	EC2	✓
18	Stimulation of Local Economy	EC3	✓
19	More Automobile Dependence in areas other than MLU	EC4	✓
20	More employment opportunities in mixed land-use areas	EC5	✓
21	More household savings in mixed land use areas like; travel costs, low rent, availability of household items nearby, and shared utility services.	EC6	✓
22	Availability of a variety of options in purchasing of grocery and other shopping items	EC7	✓
23	Property Values tend to fall with proximity to Mixed Land use	EC8	✓
24	Things are available nearby, variety and part-time jobs and business opportunities	EC9		✓
25	The distance from nearby markets; Distance from markets affects land use because many perishable crops need to reach the markets in fresh condition.	EC10		✓
26	Availability of cheap labor in mixed land-use areas	EC11		✓
Social Factors
26	Accessibility to nearby public services	S1	✓
27	Low Public Health facilities and standards in sub-urban areas	S2	✓
28	Spaces and activity-oriented destinations	S3	✓
29	Need for safer and more vibrant neighborhoods.	S4	✓
30	Demographic change and Social Poverty	S5	✓
31	MLU enhances the Sense of Place/community life	S6	✓
32	The more civilized social environment in mixed land-use areas	S7	✓
33	social security and safety in mixed land use areas	S8	✓
34	Supporting Social equality/social differences/local community	S9	✓
35	Diversified cultural values, sects, languages, castes, religions	S10	✓
36	More religious and civic facilities	S11	✓
37	More security and safety for women	S12	✓
38	MLU may reduce property crime	S13	✓
39	Encourage Community and Stakeholder Collaboration	S14		✓
Environmental Factors
40	Degrading Environmental conditions in sub-urban areas	EN1	✓
41	The demand for livable environment	EN2	✓
42	The revolutionary shift in Industrialization is a major driving factor for MLU	EN3	✓
43	Good water supply and drainage facilities in mixed land use areas	EN4	✓
44	Improved solid waste system in mixed land use areas	EN5	✓
45	Healthy environment and well being	EN6	✓
46	MLU promotes Pedestrian friendly environment	EN7	✓
47	Easy access to green spaces	EN8	✓
48	MLU improves Locational Convenience	EN9	✓
49	Strong municipal environment	EN10	✓
50	Land Use Diversity/mix forms of housing	EN11	✓

.

Further, for quantitative data, with the help of two rounds of iterative DM from experts and stakeholders of Karachi and Hyderabad, we obtained consensus on factors extracted from the literature. The Delphi Method questionnaire survey was conducted by experts [61] and based on two sections in round one. In the first section, the ranking of the list of factors provided is depicted in Table 2. They were asked to select the most important factors by ranking using a Likert scale and give less value to the factor which they feel is irrelevant or less significant. In the second section of the questionnaire, they are open to suggesting factors, which they feel, are relevant in the local context, as shown in Table 2. After the first round of the DM survey, the factors remained for the second round, which obtained less consensus through the coefficient of variance [62]. The formula for the coefficient of variation, on the other hand, accounts for the reliability as of Equation (1):

$$CV=\left(\frac{s}{\overline{\mathrm{x}}}\right)\times 100$$

(1)

where s represents the sample standard deviation and $\overline{\mathrm{x}}$ is the sample mean response. The CV, thus, is expressed in a percentage scale in (1) to compute the percentage variation among the responses/perceptions in terms of ratings for different MLU factors. We have set the levels of consensus, as in

Table 3. Coefficient of variation cut-off points.

Co-Efficient of Variation	Level of Consensus Achieved
$0\le CV\le 20$	High
$20<CV\le 30$	Acceptable
$CV>30$	Not acceptable

, based on the values of CV. If the value of CV falls from 0 to 20, it is considered a high consensus level. If the value of CV is over 20 and up to 30, then still the acceptable consensus is expected so that the factors in this category may still be considered in the next iteration. However, if the value of the CV is over 30, it is not acceptable to conclude consensus and is thus immediately eliminated.

Those factors were eliminated from the second round, which obtained a high consensus level, and were thus considered to be sufficient evidence as the determining factors for MLU. In the second round, the least consensus factors by experts and their suggested factors added in the second round of the Delphi survey and questionnaire were sent to experts again for further elimination of the factors to develop consensus. Delphi survey, the experts were asked to rank the finalized factors as they think the most appropriate according to them.

3.2.1. Delphi Method (DM)

DM is used in our study because it is one of the most well-established means of collecting expert opinions and of gaining consensus among experts on various factors under consideration. In our study, the DM uses mail questionnaires to accomplish the aforementioned goals. Delphi was chosen to help determine what draws an individual’s choice to reside in mixed-use areas due to its potential benefits of low cost, quick response, and ease of administration. In comparison to the expense associated with using other techniques, it was projected that the DM could find factors within a short period of time at a relatively cheap cost. This is, however, the first time that Delphi has been used to identify MLU factors. The DM is used to determine the level of consensus among the experts involved in determining the factors influencing residents toward mixed land-use development. The DM process starts with an initial questionnaire that acts as a ranking of the factors that are taken from the literature. The relevant question for the research is to identify the factors considered and new factors that need to be considered in mixed land use areas. Responses from the initial questionnaire must provide as many relevant factors as possible, as these factors are the foundation for the continuation of the process, where the identified factors are used for further refinement that could lead to consensus on the important factors that need to be considered. The complete two rounds process, where the previous round forms the foundation for the next, is important in the process that assists in reaching a consensus between the responders or experts. These multiple rounds develop into a consensus on the different opinions concerning the topic in question. Every feedback process provides an opportunity for each of the participants to reassess his or her initial opinion of the responses of the other responders, who remain anonymous during the entire process. The feedback process consists of a summary of the previous iterations.

3.2.2. Demographic Information of Experts

The criteria for the selection of the experts were based on the persons involved in the development process of land uses in Hyderabad and Karachi cities of Pakistan. Therefore, the architects, urban planners, and civil engineers working in the governmental sectors such as building control authorities, metropolitan corporations, and master planning authorities of both case study areas, and the professionals practicing in consulting firms were also part of this data collection process. The list of experts is given in

Table 4. Demographic composition of experts for Delphi Method.

Demographic Information	Round 1	Round 2
Experts City
Karachi	18	12
Hyderabad	14	12
	32	24
Type of current organization
Private	18	13
Government	11	10
Semi government	03	01
Field of experts
Civil engineer	11	05
Urban planner	07	07
Transport planner	01	01
Environmental expert	01	01
Project manager	01	0
Private consultant	01	0
Architect	08	08
GIS Specialist	01	01
Other	0	0
National level expert	01	01
Total	32	24

, as this study requires a certain group of experts; therefore, the purposive and snowball methods of the non-probability sampling technique were adopted.

Demographic information in Table 4 highlighted the details of experts in rounds one and two. In the first round, there were 32; out of them, 18 experts were from Karachi and 14 from Hyderabad. Experts are working in the private sector, government sector, and semi-government sector. In round, two total of 24 experts, of which 12 experts were from Karachi and 12 were from Hyderabad.

4. Results

4.1. Round One Delphi Survey

This study was conducted in Hyderabad and Karachi, Pakistan, to find out the need for MLU, its causes, and its impact on the population of the study area. Two-round Delphi questionnaire survey was used to obtain the opinion of experts on factors identified from the literature review. The first round of the questionnaire is divided into two sections, Section 1, close-ended questions, and Section 2, open-ended questions. In close-ended questions already, identified factors were arranged on a Likert scale based on 1 to 5, where 5 are considered as Extremely Significant, 4 as Highly Significant, 3 as Significant, 2 as Less Significant, and 1 as Not Significant. Factors were further divided into four categories physical, economic, social, and environmental factors, and all the factors were assigned code as Physical as P, Social as S, Economic as EC, and Environmental as EN. Section 2 of the questionnaire is based on open-ended questions in which the experts are asked to suggest the related factors of mixed land use in Hyderabad and Karachi.

4.1.1. Physical Factors

In round one, eleven physical (P) factors were identified from the literature, as shown in Figure 8a. After obtaining the opinion from experts on the Likert-based questionnaire, the responses were analyzed using the mean value, standard deviation, and coefficient of variance, whereas the coefficient of variance (CV) value is set as a standard for the consideration of factors as a consensus developed or required second. If the value of the CV is <30, it is considered that factor is relevant, on which experts agreed. If the value of the factor CV > 30 is required, the second round of the Delphi questionnaire survey. As shown in Figure 8a, after analyzing the factors P2:29.02, P5:29.52, P6:26.91, P7: 27.092, P10: 26.23, and P11:26.01 were considered as a consensus developed as they obtained the CV value >30 and P1: 32.67, P3:36.11, P4:31.53, P8:30.11, and P9: 44.03 were considered as consensus not developed required another round as they obtained the CV value <30. P11 received the highest consensus having a CV value of 26.013, and P9 obtained the lowest consensus having a CV value of 44.03726.

4.1.2. Economic Factors

As shown in Figure 8b, after analyzing the economic factors, EC1: 18.69, EC2:27.81, EC3:22.53, EC4:22.52, EC5: 18.97, EC6:17.15, EC7: 18.09 were considered as a consensus developed as they obtained the CV value >30, and EC8: 30.61 was considered as consensus not developed and required another round as they obtained the CV value <30. EC6 obtained the highest consensus having a CV value of 17.15, and EC8 obtained the lowest consensus having a CV value: of 30.61.

4.1.3. Social Factors

As shown in Figure 8c, after analyzing the factors S1:24.42, S2:28.39, S4:2518, S5: 19.70, S7:25.18, S8:25.18, and S9: 26.17, they were considered as a consensus developed as they obtained the CV value <30, and S3:32.62, S10:33.03, S11: 31.49, S12:31.07, and S13:33.75 were considered as consensus not developed required another round as they obtained the CV value <30. S5 became the highest consensus having a CV value of 19.70, and S13 had the lowest consensus having a CV value of 33.75.

4.1.4. Environmental Factors

As shown in Figure 8d, after analyzing the factors EN2: 23.64, EN9:24.64, and EN10:29.33, they were considered a consensus developed as they obtained the CV value <30, and EN1:33.26, EN3:33.54, EN4:32.85, EN5:33.40, EN6:32.92, EN7:32.29, and EN8:31.91 were considered as consensus not developed required another round as they obtained the CV value <30. EN2 obtained the highest consensus having a CV value of 23.64, and EN3 found the lowest consensus having a CV value of 33.54.

Based on the results mentioned in Figure 8a–d,

Table 5. List of Factors and Consensus level in Round One.

S No.	Factor Code	Mean	SD	CV	Level of Consensus
1	P1	3.63	1.185	32.67802	Not acceptable
2	P2	3.59	1.043	29.02064	Acceptable
3	P3	3.34	1.208	36.11974	Not acceptable
4	P4	3.66	1.153	31.53775	Not acceptable
5	P5	3.63	1.070	29.52061	Acceptable
6	P6	3.63	0.976	26.91052	Acceptable
7	P7	3.91	1.058	27.09201	Acceptable
8	P8	3.81	1.148	30.11884	Not acceptable
9	P9	2.94	1.294	44.03726	Not acceptable
10	P10	3.75	0.984	26.23303	Acceptable
11	P11	3.88	1.008	26.01374	Acceptable
12	EC1	4.06	0.759	18.6917	High
13	EC2	4.03	1.121	27.81235	Acceptable
14	EC3	3.81	0.859	22.53149	Acceptable
15	EC4	3.72	1.023	27.52041	Acceptable
16	EC5	4.28	0.813	18.97991	High
17	EC6	4.34	0.745	17.15754	High
18	EC7	4.31	0.780	18.09396	High
19	EC8	3.41	1.043	30.61811	Not acceptable
20	S1	3.75	0.916	24.42163	Acceptable
21	S2	4.00	1.136	28.39809	Acceptable
22	S3	3.69	1.203	32.62792	Not acceptable
23	S4	3.97	0.999	25.18415	Acceptable
24	S5	4.06	0.801	19.70967	High
25	S6	3.97	0.822	20.72295	Acceptable
26	S7	3.97	0.999	25.18415	Acceptable
27	S8	3.97	0.999	25.18415	Acceptable
28	S9	3.81	0.998	26.17657	Acceptable
29	S10	3.66	1.208	33.03258	Not acceptable
30	S11	3.59	1.132	31.49699	Not acceptable
31	S12	3.84	1.194	31.07186	Not acceptable
32	S13	3.66	1.234	33.75519	Not acceptable
33	EN1	3.88	1.289	33.26221	Not acceptable
34	EN2	4.13	0.976	23.64864	Acceptable
35	EN3	3.47	1.164	33.54356	Not acceptable
36	EN4	3.94	1.294	32.8532	Not acceptable
37	EN5	3.78	1.263	33.4079	Not acceptable
38	EN6	3.72	1.224	32.92325	Not acceptable
39	EN7	3.66	1.181	32.29382	Not acceptable
40	EN8	3.63	1.157	31.91796	Not acceptable
41	EN9	3.91	0.963	24.64007	Acceptable
42	EN10	3.91	1.146	29.33	Acceptable

depicts the details of the factors which were accepted in the first round. The table also describes the factors which proceed in the second round.

4.2. Round Two Delphi Survey

4.2.1. Physical Factors

In round two, nine physical (P) factors were given in the questionnaire. Five factors among them were those that obtained a CV value above 30 in the first round, and four new factors were added, which were proposed by experts in the first round in the open-ended part of the questionnaire shown in Figure 9. After obtaining the opinion from experts on the Likert-based questionnaire, the responses were analyzed using the mean value, standard deviation, and coefficient of variance, whereas a coefficient of variance (CV) value was set as a standard for the consideration of factors as consensus was developed or eliminated. If the value of the CV was <30, it was considered that factor was relevant, on which experts agreed. If the value of the factor is CV > 30, it was eliminated. As shown in Figure 9, after analyzing the factors in two P1:21.30, P4:24.11, P8:25.48, P12:20.31, P13: 23.61, P14:29.07, and P15:27.52, they were considered as a consensus developed as they obtained the CV value >30, and P3:34.25 and P9:41.70 considered as eliminated as they obtained the CV value <30. P12 obtained the highest consensus having a CV value of 20.31, and P9 obtained the lowest consensus having a CV value of 41.70.

4.2.2. Economic Factors

As shown in Figure 10, after analyzing the economic factors in round two, EC9:26.31 and EC10:26.31 considered an unanimity developed as they obtained the CV value >30, and EC8: 36.11, EC11:32.34 considered as eliminated as they obtained the CV value <30. EC9 and EC10 obtained the highest consensus having a CV value of 26.31, and EC8 had the lowest consensus having a CV value of 36.11.

4.2.3. Social Factors

As shown in Figure 11, after analyzing the factors S3:21.74, S10:27.19, S11:21.30, and S14:19.86, they were considered a consensus developed as they obtained the CV value <30, and S12:31.14 and S13:31.83 were considered as eliminated as they obtained the CV value <30. S14 obtained the highest consensus having a CV value of 19.86, and S13 obtained the lowest consensus having a CV value of 31.83.

4.2.4. Environmental Factors

In Figure 12, after analyzing the factors in round two; EN3:29.19, EN6:26.58, EN8:27.11, they were considered a consensus developed as they obtained the CV value <30, and EN1:31.01, EN4: 30.24, EN5: 30.39, and EN7:31.69 were considered as eliminated as they obtained the CV value <30. EN6 obtained the highest consensus having a CV value of 26.58, and EN7 obtained the lowest consensus having a CV value of 31.69. The detailed results of factors and concensus level of round two including CV value and SD in

Table 6. List of Factors and Consensus level in Round Two.

S No.	Factor Code	Mean	SD	CV	Level of Consensus
1	P1	3.8333	0.81650	21.30019	Acceptable
2	P3	3.2500	1.11316	34.25108	Not acceptable
3	P4	3.9583	0.95458	24.11591	Acceptable
4	P8	3.6250	0.92372	25.48193	Acceptable
5	P9	3.0000	1.25109	41.703	Not acceptable
6	P12	4.0833	0.82970	20.31935	Acceptable
7	P13	4.0417	0.95458	23.61828	Acceptable
8	P14	3.7917	1.10253	29.07746	Acceptable
9	P15	3.7500	1.03209	27.5224	Acceptable
10	EC8	3.3333	1.20386	36.11616	Not acceptable
11	EC9	3.9583	1.04170	26.31685	Acceptable
12	EC10	3.9583	1.04170	26.31685	Acceptable
13	EC11	3.6250	1.17260	32.34759	Not acceptable
14	S3	3.7083	0.80645	21.74716	Acceptable
15	S10	3.5833	0.97431	27.1903	Acceptable
16	S11	3.8333	0.81650	21.30019	Acceptable
17	S12	3.9583	1.23285	31.14595	Not acceptable
18	S13	3.6667	1.16718	31.83189	Not acceptable
19	S14	3.8333	0.76139	19.86252	High
20	EN1	3.4167	1.05981	31.01853	Not acceptable
21	EN3	3.7083	1.08264	29.19505	Acceptable
22	EN4	3.9583	1.19707	30.24202	Not acceptable
23	EN5	4.0000	1.21584	30.396	Not acceptable
24	EN6	4.0000	1.06322	26.5805	Acceptable
25	EN7	3.7500	1.18872	31.6992	Not acceptable
26	EN8	4.2083	1.14129	27.11998	Acceptable

.

4.3. Discussion

The data gathered through the Delphi Method and the results of this study are depicted in the following main factors.

4.3.1. Physical Factors

After completion of two rounds of the Delphi expert-based survey, as shown in Figure 13 comparison to rounds one and two, physical factors are fifteen altogether. Eleven factors were in the first round, and four more factors were added in round two, as suggested by experts in the open-ended section of the first round. As in Figure 13, the red dashed line reveals the benchmark set for CV value 30. Those factors that obtained a CV value of less than 30 were considered as a consensus developed, and factors that obtained a value of more than 30 were considered as consensus, not developed. In round one, if factors obtained a CV value of more than 30, they were added in round two for reconsideration by the experts. However, even after round two, factors that still obtained a CV value of more than 30 were eliminated. As shown in Figure 13; P2, P5, P6, P7, P10, and P11 obtained consensus in the first round and P1, P4, P12, P13, P14, and P15 obtained consensus in the second round, whereas P3 and P9 were eliminated from the final list of the factors as they did not obtain the benchmark value of CV less than 30 even after the second round of the Delphi survey. The previous studies (see Table 1) had identified nine physical factors, whereas this study explores four additional factors; availability of infrastructure in MLU areas (health and educational facilities), The overall layout and size of a city with mixed land use suits the middle class that is in the majority, Sustainable, environmentally friendly, and use of minimal designs in mixed land-use Neighborhood, A Range of Housing Opportunities and Choices suggested by the experts.

4.3.2. Economic Factors

After completion of two rounds of the Delphi expert-based survey, as depicted in Figure 14, in comparison to rounds one and two, economic factors are eleven altogether; eight factors were in the first round, and three more factors were added in round two as suggested by experts in the open-ended section of the first round. As in Figure 14, the red dashed line indicates the benchmark set for CV value 30. Those factors that obtained a CV value of less than 30, were considered as a consensus developed, and factors that obtained a value of more than 30, were considered consensus, not developed. In round one, if factors obtained a CV value of more than 30, they were added in round two, for reconsideration by the experts. However, even after round two, factors still obtained a CV value of more than 30, and they were eliminated. As shown in Figure 14: EC1, EC2, EC3, EC4, EC5, EC6, and EC7 obtained consensus in the first round, and EC9 and EC10 obtained consensus in the second round, whereas EC8 and EC11 were eliminated from the final list of the factors as they did not obtain the benchmark value of CV less than 30 even after the second round of the Delphi survey. The previous studies (as in Table 1) had identified seven economic factors, whereas this study explores two additional economic factors, such as things available nearby, variety and part-time jobs, and business opportunities. The distance from nearby markets; distance from markets affect land use because many perishable crops need to reach the markets in fresh condition suggested by the experts.

4.3.3. Social Factors

In the completion of the second round of the Delphi expert-based survey, as shown in Figure 15 comparison of rounds one and two, social factors are fourteen altogether; thirteen factors were in the first round, and one more factor was added in round two as suggested by experts in open-ended section of the first round. As mentioned in Figure 15, the red dashed line shows the benchmark set for CV value of 30. Those factors that obtained a CV value of less than 30 were considered as a consensus developed, and factors that obtained a value of more than 30 are considered consensus, were not developed. In round one, if factors obtained a CV value of more than 30, they were added in round two for reconsideration by the experts. However, even after round two, factors still obtained a CV value of more than 30, and they were eliminated. As shown in Figure 15; S1, S2, S4, S5, S7, S8, and S9 obtained consensus in the first round and S3, S10, S11, and S14 obtained consensus in the second round, whereas S12 and S13 were eliminated from the final list of the factors as they did not obtain the benchmark value of CV less than 30 even after the second round of the Delphi survey. The previous studies (as in Table 1) had identified eleven social factors, whereas this study explores one additional social factor related to Pakistan, Encourage Community and Stakeholders Collaboration suggested by the experts.

4.3.4. Environmental Factors

At the completion of two rounds of the Delphi expert-based survey, as shown in Figure 16, in comparison to rounds one and two, environmental factors are ten in total. As shown in Figure 16, the red dashed line displays the benchmark set for CV value 30. Those factors that obtained a CV value of less than 30, were considered as a consensus developed, and the factors that obtained a value of more than 30 were considered consensus, not developed. In round one, if factors obtained a CV value of more than 30, they were added in round two for reconsideration by the experts. However, even after round two, factors that still obtained a CV value of more than 30 were eliminated. As shown in Figure 16, EN2, EN9, and EN10 obtained consensus in the first round, and EN3, EN6, and EN8 obtained consensus in the second round, whereas EN1, EN4, EN5, and EN7 were eliminated from the final list of the factors as they did not obtain the benchmark value of CV less than 30 even after the second round of the Delphi survey. The previous studies had identified six environmental factors, whereas this study explores no additional factors. Experts are satisfied with the identified factors from previous studies as consensus developed.

After completion of two rounds Delphi survey, as shown in Figure 13, Figure 14, Figure 15 and Figure 16, forty-two factors were in the round first, and out of them, twenty-four factors obtained consensus. In round two of the Delphi survey, out of twenty-six factors, sixteen factors obtained consensus. Results illustrate that forty factors are selected as the major factors that are influencing the residents to live in mixed land use areas. This methodology of using the Delphi survey technique in identifying the mixed land use factors has never been used. However, with the help of the Delphi technique, we identified the core factors with the help of a thorough literature review and stakeholders’ agreements between Hyderabad and Karachi. These factors are important to consider before designing mixed land use developments not only for Karachi and Hyderabad but also for other urban areas of the world. However, the Delphi Method is solely based on experts’ opinions. Public participation is ignored in this method during the data collection process. It is a very vague process of continued commitment. It is required from participants who are being asked a similar question multiple time. The results depicted the experts based or literature-based factors that they may think are responsible for MLU development, whereas some residents’ opinion-based factors maybe not be considered.

4.4. Policy Implications

Developing countries are experiencing rapid population increase over the years. Pakistan is also a developing country, where folks were found struggling in search of better jobs, business opportunities, basic healthcare, and decent education services, particularly in remote environments. Hence, people often look forward to urban centers in quest of better living conditions. It is obvious that policy makers have to accommodate the population influx in urban centers. To fulfill the need of residents and reduce the housing backlog, concerned authorities are considering the mixed land-use development approach as an optimum solution to eradicate land-use conflicts in urban areas of Pakistan. In this scenario, our study identified the main factors that dragged local residents to live in MLUs. This has emerged as a tool to tackle the issues of housing shortage and utilized the availability of scarce land.

The main purpose of conducting this research is to understand the reasons for MLU development and to explore residents’ points of attraction towards MLU development areas in Pakistan. This study found prominent factors with high consensus levels. This research clarified foremost MLU development factors, for example, travel costs and low rent, availability of household items and shared utility services, availability of a variety of options in purchasing grocery and other shopping items, and economic vitality due to commercial activities. The MLU factors such as demographic change and social poverty; accessibility to public services, household savings and reduction in commuting time; easy access to restaurants and other fast-food franchises, etc. In light of the aforementioned identified prominent MLU factors, this study put forward the following policy implications:

• The travel cost and low-rent facilities are highly endorsed by the experts as the key MLU development factors. Thus, urban planning policies should be made considering the low-cost multiple travel options available for the people living in MLUs. On the other hand, because of the high demand, the housing rents were found beyond the reach of the common man. It is recommended that public and private sector agencies should formulate policies focusing on low-cost residential facilities in the MLU centers.
• The second prominent factor is the availability of household items and shared utility services. Thus, the policies should be formulated accordingly while designing MLUs, considering the availability of household items access to residents. Provision of utility services in MLUs at ease for residents.
• The third important factor endorsed by experts is the economic vitality due to commercial activities in MLUs. The decision makers, while drawing the policies, should emphasize the provision of commercial activities together with suitable parking spaces.
• The fourth prominent factor is demographic change and social poverty. The policies should be designed in a way that should eradicate poverty and provide equal employment opportunities to all residents. Population density should also be maintained with the provision of policy implications throughout MLU zones. A livable environment for residents may be provided, so they perform their social activities freely. This research parameter is linked to SDG goal number 10, i.e., reduced inequalities.
• The fifth imperative factor is access to public services in MLUs. The planning of MLUs should be followed by the provision of public services within residents’ approach at walking distance.
• The sixth significant factor is easy to access to restaurants and fast-food franchises. Residents prefer ready-to-eat food available within walking distance. Policies should be designed to make it more convenient and cost-effective for residents.

The policy makers have to look for sustainable solutions to avoid the hindrances that residents are facing in MLU development areas. The identified MLU factors give a broader understanding of residents’ criteria for choosing MLU development factors. Authorities should review the core factors which are identified in this research before implementing policy for MLU development. The building bylaws should be revised prior to allowing MLU development in Karachi and Hyderabad. With the help of these factors, we can make our cities more sustainable.

5. Conclusions

The MLU technique allows for sustainable urban planning. in developing countries, especially in Pakistan. In this study, experts from Karachi and Hyderabad were asked to evaluate the factors which had motivated the inhabitants to live in MLU areas. To identify the core and relevant factors of MLU, it was necessary to obtain an endorsement from the experts in relevant areas. To obtain the genuine and core factors, the Delphi Method (DM) was used in up to two iterations. It was concluded from the study that several factors were responsible for inhabitants to live in MLU areas due to its vibrant surroundings and less dependency on automobiles for movement. There were four major categories of factors, which were further sub-classified as fifty factors overall. Forty-two factors were identified from the literature, and eight factors were proposed by experts. In round one of the Delphi questionnaire, forty-two literature-based factors were distributed in four categories as physical, economic, social, and environmental factors. Of these, eleven were physical, eight were economic, thirteen were social, and ten were environmental factors. Using the DM, out of forty-two factors in the first round, twenty-four factors could get endorsement as relevant factors, and the remaining eighteen factors, along with eight new factors suggested by experts in the first round, were placed in a questionnaire in the second round. Out of twenty-six factors in the second round, sixteen factors got an endorsement from experts, and ten factors were rejected. At the end of two rounds of DM, it is identified forty significant factors altogether. Therefore, the results show that residents were willing to live in MLU areas. Due to some lacking basic facilities such as inexpedient transport linkages, absence or lack of basic utility services, housing backlog, more travel cost and time, more distance to cover to reach workplace and market, fewer employment opportunities, insecure environment, and poor municipal services in separate land-use areas. Inhabitants showed their willingness to live in MLU areas because of its vibrant land-use features and design, socially diverse, multiple job opportunities, enclosed availability of services, near to marketplaces, more health facilities available, near to workplace, public transport availability, less dependency on automobiles for movement, low carbon exposure, livable environment, and strong municipal system. Consequently, all the significant factors were identified, and these final forty factors will be helpful for urban planners to make urban areas sustainable. The prominent factors with high consensus level were clarified as travel costs and low rent (17.15), availability of household items nearby and shared utility services (17.15), availability of a variety of options in purchasing grocery and other shopping items (18.09), and economic vitality due to commercial activities (18.69). The MLU factors such as demographic change and social poverty (19.70), accessibility to nearby public services (24.40), more household savings in mixed land use areas, the reduction in commuting time (26.01), and easy access to restaurants and other fast-food franchises (26.23) also attracted experts to be nominated as crucial MLU development factors. For the first time, this research identified the factors responsible for MLU development in highly urbanized settlements of Pakistan. The identified MLU factors can contribute to the development process of land-use planning in a country and the rest of the developing world. However, this research is limited to the identification of the factors only based on experts’ choices.