Dr. Rakesh Mohan Deputy
Governor Reserve Bank of
India Central Office
Mumbai-400001

Dear Sir,

Sub: Report of the Technical Advisory Group on Development of Housing Start-Up Index in India

We are pleased to submit the Report of the Technical Advisory Group on Development of Housing Start-Up Index in India appointed vide the R.B.I, memorandum dated 31-07-2007

Housing start is considered to be a lead indicator in many developed and developing economies because of the strong forward and backward linkages it has with various sectors. The number of housing starts during a given period reflects the institutional response in a country to the current demand and supply situation in the market, as reflected though operationalisation of the existing building permits into actual starts. This has an impact not only on the construction activities but also on several consumer durables and investment goods sector of the economy.

The decision to set up a Technical Advisory Group (TAG) by the Reserve Bank of India to consider bringing out Housing Start up Index (HSUI) on a regular basis is extremely timely. This is particularly so because the current meltdown of the economy at global  level as also in many less developed countries have been linked partially to the ‘developments’ in the housing sector. Indeed, these have powerful multiplier effects on the economy, operating through the intersectoral linkages in the production system. It makes a significant impact on financial sectors as well, as has been realized with some amount of concern, in recent months.

The members of the TAG consider construction of HSUI and its regular release to be an extremely important and challenging responsibility which has been long overdue. They plead for taking urgent measures to put into operation an institutional structure entrusted with the responsibility of bringing out HSUI, keeping in view the global practices as also the ground reality of India into consideration.  This indeed, can become a powerful tool for monitoring the movements in several segments of the economy.

The start up coefficients, computed from the data in recent past, reflect institutional and social response to housing permits, in terms of their conversion into actual housing starts. The time required for administrative and procedural clearances after the issuance of permits, to complete the formalities of obtaining loans, organizing materials, construction process etc. can be considered to be somewhat rigid or fixed in the short run. As housing is a long term decision, predictions based on these coefficients, that reflect procedural and social rigidities governing the house construction process, are likely to be fairly reliable.

Understandably, the actual housing starts at any point of time are likely to be influenced by a host of other factors like price of building material, interest rates in general and that for housing loans, policy pronouncements, legislations, administrative orders affecting construction sector etc. To an extent, these would affect the demand and supply parameters in the housing market that, in turn, would determine the number of application for permits. The institutional response - in terms of the number of permits actually issued - would have some time lag. It is nonetheless clear that market based factors would get incorporated in the computation of the index through the key variable – the number of permits issued in recent past, say the last two years.

The TAG believes that the HSUI can be used by housing related agencies as the basic or core predictor. These agencies can combine the values of this index with other short term indicators and policy variables to arrive at more detailed projections of housing activity, if they so desire.

Let me put on record my gratitude for the keen interest and enthusiasm with which the TAG members have participated in the deliberations and finalisation of the Report. The analytical insights and meticulous care with which they have commented on the conceptual and methodological issues and examined technical aspects of data availability and reliability in the meetings as also through internet communication have been commendable. It is only because of the full cooperation of the members that the complex methodological issues as also those related to institutional grounding could be resolved and the Report submitted within a short time.

INTRODUCTION

(i) To review base paper on concepts, methodology, approach to generate the database for construction of the indices and suggest a feasible methodological framework for construction of HSUI for the Indian economy, with a view to assist monetary policy formulation, and to guide and oversee its implementation.

(ii) To recommend modalities of entrusting the work for construction of HSUI by appropriate external agency or institution, including scope of work and deliverables.

(iii) To evaluate the work of the external agency/institution and recommend its acceptance by the Bank.

(iv) Any other issue as deemed necessary for development of the HSUI.

1.1.5. The Department of Statistics and Information Management (DSIM) (Statistical Analysis Division) provided the secretarial support to the Technical Advisory Group.

1.2   Report Outline

1.2.1 The Group deliberated on issues regarding the development of sound and reliable Housing Start-Up Index during its four meetings held in Mumbai.   During the first two meetings, the issues concerning the scope, coverage, relevance and operationalisation of the index were discussed in some detail. The issues relating to the concept of housing start up, sources of data, data collection mechanism, periodicity of compilation, need for pilot survey etc. were also deliberated. The Group had the benefit of the presence and participation of the Deputy Governor Dr. Rakesh Mohan in its third meeting which finalized the methodology of the pilot survey and detailed out the procedures and institutioinalisation of the index building exercise. Decisions were taken also regarding the selection of the urban centres for the survey, sampling technique, survey questionnaires and the methodology of data analysis. The fourth and final meeting, where again Dr. Rakesh Mohan was present, discussed the empirical results of the surveys conducted in Coimbatore, Mumbai, Delhi, Delhi, Villupuram and Saswad. The Group felt there was a need for conducting a survey on the processes including the formal requirements for issuance of building permits by the various municipal bodies across the country. As suggested by the Group a meeting of the municipal commissioners and officers from town planning departments/urban development authorities from some selected cities with members of the Group was held at National Building Organisation, New Delhi to deliberate on the issues relating to the existing system of data collection on building permits and explore the possibility of setting up a mechanism for compilation and collation of the exiting data and collection of some additional data on the building permits. The study note on permit issuing processes across the country based on the information given by the municipal commissioners and officers from town planning departments/urban development authorities from some selected cities were circulated among the members. Based on the discussions and decisions in earlier meetings as also the analysis of the data gathered by NBO, the Group finalised the methodology for construction of the index on a regular basis and proposed an institutional structure that would be responsible for its operationalisation, as presented in this Report.

1.2.2 The Report is divided into five sections. Section 2 discusses international practices in constructing house construction related indices and the experiences of building related to data/information base for housing sector in India.  Section 3 gives insight into the methodological issues for construction of HSUI, taking the empirical context of Indian urban scenario into consideration. Section 4 presents the pilot survey results and enumerates the limitations of the data used in this and similar empirical studies in the country.  The recommendations of the Group are presented in Section 5.

1.3 Acknowledgements

1.3.1 Preparation of HSUI being the maiden venture of its kind, the challenges were many. The contribution by each of the members and their institutions were crucial in completing the projects in a meaningful manner and all of them must be sincerely thanked for their efforts. The Group expresses sincere thanks to Dr. Rakesh Mohan, Deputy Governor, Reserve Bank of India for his valuable insights and suggestions provided during the entire period of the Group’s functioning. The Group also thanks Dr P. K. Mohanty of the Ministry of Housing and Urban Poverty Alleviation, the Joint Secretary in charge of NBO, for providing information regarding the present system of data generation in the Ministry and assisting in formalizing a system of data compilation for HSUI on a regular basis. The Group is thankful to  Shri. Radhey Shyam, former Adviser, DSIM, RBI, Shri. Sangeet Shukla, CGM, State Bank of India,  Shri. T. Prabhakaran, former Director Finance, HUDCO and Shri. P.K Ray, former Director General and CEO (In charge), NSSO who were part of this Group in its initial phases. Special thanks are for the Directorates of Economics and Statistics (DES), Tamil Nadu, Maharashtra and Delhi that undertook the responsibility of conducting the pilot survey. The Group is thankful to Dr. M. Murughan and Shri S. Sudalaimuthu of DES – TN,  Shri. K.S.P Rao, Ex-Deputy Director General, NSSO, Shri. Raj Pal, Principal Adviser, NHB, Smt. Uttara Dasgupta, GM, SBI Shri. K. L Paulson, DGM, SBI and Shri. Avanish Mishra, NBO for representing their organizations in various meetings and putting forward invaluable suggestions as also to Mr. Deepak Gahlowt was a special invitee in the third meeting to present his work on municipal housing permit system. The Group is also thankful to Shri. A.B Chakraborty, Adviser, Monetary Policy Department (MPD), RBI and Dr. O.P Mall, Director, MPD, RBI for their contributions.

1.3.2 The Group is thankful to Dr. Amal Kanti Ray, Officer-in-Charge, Department of Statistics and Information Management, RBI for creating an excellent environment and facilities in overseeing the pilot survey and coordinating the Group meetings. His continuous persuasion for completion of the Report is highly appreciated. The Group also recognizes to Shri. Sanjoy Bose, Director, Dr. A.K. Tripathi, Director, Smt. Sushila Augustine, Director and Shri Joice John, Research Officer of Department of Statistics and Information Management, RBI for shouldering the total responsibility of analyzing the survey results and preparing the  Report.

1.3.3 The Group places on record the valuable contributions, encouraging thoughts and support for design of the survey schedules, conduct of pilot survey and firming up the methodology for compilation of Housing Start Up Index (HSUI) by  ex-officio members,  Dr. R. B. Barman, former Executive Director, RBI, Dr. S. K. Nath, former Director General, Central Statistical Organization (CSO) and Shri. S. K. Sinha, former Director General and CEO, National Sample Survey Organisation (NSSO). 

Section 2

INTERNATIONAL PRACTICES AND DOMESTIC EXPERIENCES

2.1 Background

2.1.1 Housing Starts as a Lead indictor: Housing Start indices are considered to be lead economic indicators because these give an idea regarding the level of activities in a number of sectors of the economy in immediate future and in this sense it is forward-looking. A high level of housing activity can trigger economic growth, cause interest rates to rise and may have inflationary impact. Similarly, decline in housing activity could slow down the economy, cause yields and interest rates to fall, dampen investments in linked sectors and push the real economy into recession. The current meltdown of the economy is being linked to the crisis emanating from the housing sector. The developments in the latter have direct causal effect on the real economic activities as also the financial sector, which amplifies macroeconomic shocks. Furthermore, these can become autonomous sources of macroeconomic and financial fluctuations. Because of the high outlays needed to start construction projects, an increase in housing starts is often taken as an indication of commitment of related investment in other sectors. It reflects an enhancement of business and consumer optimism. The housing starts figures provide insight into the upcoming demand for consumer durables in near future, since new house constructions/purchases are typically followed by large expenditures on a wide range of consumer products. Conversely, an economy that is growing rapidly is noted as having a high demand for housing and large number of housing starts.

2.1.2 Ripple effect of housing demand: The housing sector has powerful multiplier effects across the commodity and service markets that impact on the overall growth performance of the economy. Changes in the rate of housing starts reflect demand for new dwelling units, impacting on the outlook for construction industry. As new house/building constructions get started, the demand for construction materials goes up. Further, employment in the construction activities rises immediately, causing a higher demand for a large number of consumption goods including durables, which eventually may cause the general price rise in the country. Once the houses are sold, these generate revenues in the hands of the house-builders and open a myriad of consumption opportunities for the buyer. Refrigerators, washing and drying machines, furniture, etc. are a few things that new house buyers would often spend their money on. The economic 'ripple effect' in the Indian context has been noted to be substantial, especially when new houses are coming up at a rate higher than in the past. In a more specific sense, the housing starts data carry valuable clues for house-builders, producers and suppliers of construction materials, banks, lenders, and house furnishings companies, for their future decisions.

2.2  International Practices

2.2.1 Internationally, countries like Canada, United States, Japan, France, Australia, and New Zealand are compiling data related to building permits/housing starts on a regular basis. Most of these countries compile housing starts using housing permits data, collected either through census or sampling method.  The practices followed in these countries are summarised below. 

2.2.2 Statistics Canada publishes data on house permits on a monthly basis. The monthly Building Permits Survey of the Canadian municipalities collects data on the value of construction intentions in the non-residential sector; and the number of dwelling units authorized in the residential sector and their value. The Survey collects information also on the number of dwelling units demolished. It covers all the municipalities that issue permits. At present more than 2,350 Canadian municipalities, representing all provinces and territories are covered by the survey. Data sources and methodology are detailed in Annex-2.

2.2.3 Building permits data are widely used as a lead indicator for the construction industry in Canada; the issuance of a building permit is one of the first steps in the construction process. Statistics on building permits are essential for the computation of capital expenditures in building construction, depreciation by components and estimation of net capital stock on quarterly and annual basis. The results of this Survey are used by Canada Mortgage and Housing Corporation (CMHC) as a reference base for conducting a monthly survey of housing starts and completions. There are, thus, a wide range of users – from economists in public and private production sectors and development planners to construction industry analysts and housing market analysts in Canada.

2.2.4 The United States Census Bureau compiles and publishes data on 'New Residential Construction' on a monthly basis, based on sample survey. The purpose of the survey is to provide statistics on the construction of new privately owned residential structures in the country. The data relate to new housing units intended for occupancy and maintenance by the occupants. These include single-family unit as well as multiple-family unit buildings but exclude hotels, motels, and group residential structures such as nursing houses and college dormitories.  Also excluded are the publicly owned housing and manufactured mobile housing units. Units in structures built by private developers with partial public subsidies are all classified as private housing and hence included in the database.

2.2.5 Statistics on housing units authorized by building permits include those that are issued under local permit-issuing jurisdictions by a building or zoning permit agency. Statistics are based upon reports submitted by local building permit officials in response to a mail survey. Approximately 9,000 of the 20,000 permit issuing places in the United States are surveyed monthly, the remaining being surveyed annually. Estimates of Housing Units Authorized, but Not Started; Housing Starts; Housing Units under Construction; and Housing Completions are obtained from the Survey of Construction (SOC). SOC comprises two parts: (i) Survey of Use of Permits (SUP) which estimates the number of new construction in areas that require a building permit and (ii) Non-permit Survey (NP) estimating the amount of new constructions in areas that do not require a building permit. Data from both parts of the SOC are collected by Census field representatives. For SUP, they visit a sample of permit offices and select a sample of permits issued for new housing. These permits are then followed through to record the date of their starting and completion. From these sample surveys, related housing statistics are estimated. The detailed methodology is presented in Annex-3.

2.2.6 The Japanese Ministry of Land, Infrastructure and Transport announces Japan's total housing starts every month under official statistics of Japan. The Housing Starts figure gives insight into consumer activity in Japan, since new home purchases typically require a large investment for consumers.

2.2.7 In France, the National Institute of Statistics and Economic Studies, Directorate-General of the Ministry of the Economy, Finance, and Industry, publishes the information on house starts on monthly basis. The rate of growth in housing construction is released as percentage change over the preceding year.

2.2.8 Australian Bureau of Statistics publishes dwelling starts on quarterly basis.  The number, which is officially called Construction of Dwellings, measures growth in the construction sector and reflects the overall health of the housing market.  The headline number is the percentage change in Dwelling Starts from the previous month's figure.

2.2.9 Statistics New Zealand, a government department and New Zealand's national statistical office publishes data on building permits on monthly data.  Building Permits or Building Consents, are issued when a building project is authorized for construction. Since Building Consents are the earliest signals of expanded housing supply, this is taken as a lead indicator by most actors in the housing market. The headline figure is the percentage change in new consents for house construction in the month.

2.3 Explorations into the Existing Data on House Construction  in India

2.3.1 Considerable information on house construction is available from permit issuing authorities in India, as is the case of several other countries. The permit issuing authorities in India vary depending on the nature of settlement, as discussed below:

- Municipal Corporation

- Town Planning Authority

- Tahsildaar (Nagar Parishad / Palika)

- Gram Panchayat

- Building Plan Register

- The individual files for the building permits that contain various documents like Project Proposal, Building Plan Approval, Commencement Certificate, Occupancy Certificate and Completion Certificate.

• The survey did not include non-residential constructions. Also, addition and alteration of constructions to the existing buildings were not covered. New constructions, taken up in an unauthorized manner, were also not covered in this survey.

• A fairly representative year 2004-05 is selected as the reference year for the study. As the construction permissions issued by local bodies are valid for three years, those not reporting any construction at the time of the survey can be taken to have lapsed. The names and addresses of the permission holders, numbering about 5792, who were sanctioned the Construction permissions during the reference year, were collected from the Municipal Corporation. Individual permission holders were the respondents of this survey.

• Chennai Corporation has got 10 administrative zones. Variability was noticed in zone-wise distribution in the number of permissions issued. Consequently, a stratified random sampling method with proportionate allocation to the zones was adopted in the pilot survey. Two per cent of permission holders, numbering about 114 were randomly selected for detailed data collection. The questionnaires have been canvassed to the permission holders or their household members.

• Out of 114 building permissions pursued, in 111 cases, house constructions have started while in 3 cases, this has not happened. Out of 111 cases where house construction had started, only 10 were observed to have not been completed till the date of the survey.

• Responses from eleven municipal bodies were received in NBO. These include Ahmedabad, Delhi, Kolkata, Mumbai, Coimbatore, Puri, Bilaspur, Korba, Bhilai, Agarthala and Bhopal.

•The data on building permissions issued during the quarter April -June 2008 through Schedule A, Annex-5 is received from all the eleven cities. However information on house construction linked processes and practices through Schedule-B in Annex-5 is received only from two centers namely, Puri and Mumbai .

• As regards construction linked processes and practices, the information (Schedule B, Annex-5 ) is received only from two centres. It is, therefore, difficult to come to any conclusion regarding the nature of practices followed in various cities across the country. However, the procedures adopted in the cities of Puri and Mumbai are similar except a few exceptions.

• One can derive only a few conclusions from the limited information (Schedule B, Annex-5) that have been collected and analysed. The attempt nonetheless revealed that the information required for constructing HSUI are largely available at the city level and are already being complied, although not very systematically. One can also argue that the system can be strengthened to generate whatever additional data requirements may come up for undertaking the exercise. For this, the inter-institutional linkages are to be strengthened and there must be political will at the highest level, backing up the effort.

Section 3 

METHODOLOGY

3.1 Background

3.1.1 The Group observed that the present data collection system as organised and updated by National Building Organization can be strengthened and fine tuned to have the requisite base data for constructing a housing start-up index (HSUI) on a quarterly basis. 

3.1.2 The objective of the HSUI is to track the changes in the level of construction activities in housing sector, which can identify and signal growth or reversionary tendencies in the housing sector. The housing starts in a particular quarter can be estimated from the permits issued in that quarter and the various past quarters by using the rates at which the permits have got converted into starts in the recent past. It would therefore be important to construct a series of start rates (coefficients) for the permits given during the preceding quarters based on the information on actual starts after the issuance of the permits.  These coefficients are expected to be different for different quarters due to the seasonality involved in the housing starts. The data on housing starts for a two year period or eight quarters has been considered appropriate for building up the series of coefficients. For the housing starts, out of the permits given before two years, an 'aggregate coefficient’ may be calculated based on the actual empirical data. Thus, there will be nine coefficients for each quarter of the year, eight for the preceding quarters, and one more for the residual permits that are two-year old. Since these coefficients are likely to be different for each of the four quarters, one would end up building up a matrix with four rows and nine columns.

3.1.3  Once the matrix of start-up rates is constructed based on survey data, the number of house construction started in a particular quarter, say A out of the permits issued in a preceding quarter, say B, can be obtained by multiplying the number of housing units authorised through issuance of permits in quarter B with the corresponding start rate (coefficient) in the matrix. Aggregation of the nine values thus obtained would give the total number of housing starts in the quarter A.

3.1.4 The Group decided that the scope of the index should be limited to new built residential buildings in urban areas of India, whose construction is authorised through issuance of building permits. Consequently, the pilot surveys undertaken to estimate the coefficients do not include permits for non-residential buildings including commercial, institutional and industrial buildings. Furthermore, the surveys do not include the publicly owned/built residential housing units. These cover only urban areas because the residential construction activities here are likely to affect macroeconomic parameters much more than in the rural areas. Un-authorised constructions can be excluded from the scope of the analysis. It is possible to assume that the excluded components are multicollinear with the formal residential units and hence the index can signal the direction of movement for both.

3.1.5 After overviewing the results of the survey conducted by DES-TN, the Group felt about the need to conduct a comprehensive pilot surveys in few cites/town in order to generate housing start coefficients and test their sturdiness. It is only then that these numbers can be recommended for application to the information on the number of building permits for constructing HSUI. The objective of the pilot surveys should be to construct the Start up Coefficient Matrix with adequate empirical strength so that HSUI can be constructed on a quarterly basis. It should help in identifying the difficulties and challenges in this exercise as also sorting out the issues related to sampling design, selection of variables, weighting pattern, choice of base year etc.

3.1.6 The Group suggested that the pilot survey, using a common methodology and schedules, should be conducted in 3 class I cities and 3 small towns. A subgroup was constituted for designing of the schedules for data collection. The comments/suggestions on the schedules prepared by the subgroup were obtained from the members based on which the methodology for data collection for the pilot study was also finalized. The three class-I cities identified for the survey were Mumbai, Delhi and Coimbatore. The respective Directorate of Economic and Statistics (DES) were entrusted with the data collection job. It was considered important to collect information from one small town in the states of Maharashtra and Tamil Nadu and one near the National Capital Territory of Delhi for the calendar years 2003 and 2004. The choice of two calendar years was made with a view to identify temporal differences in the start rates as it may help to fix the periodicity for conducting such surveys for generating the coefficients, to be used for constructing HSUI. The concerned DESs were requested to conduct the surveys by selecting the town as per the framework of the project and report the results to the Group.

3.1.7 The Terms and Definitions used in the study are presented in Annex 6.

3.2  Data Collection

3.2.1 The data related to housing starts were collected in two stages as per the survey schedule given in Annex-7. The survey schedule has two parts.

3.2.2 Collection of Information on Building Permits:  Information on the  permits issued for new residential construction was collected from the permit issuing authorities in all cities and towns through Schedule –Part I given in Annex-7. The permits for alterations of the existing building were not included in their survey. However, the permits given for additional housing units in the existing building; construction of new building by demolishing the old existing building were included. The survey excluded all non-residential buildings, as noted above. However, mixed-use houses like residential cum commercial, residential cum industrial units etc. were included.

3.2.3  Survey conducted for determining the coefficients of Housing Starts:  The data on housing starts were collected by drawing a sample from the permits issued for new residential buildings in city during the four quarters in certain reference year. The reference year was assumed to be two or three years old, generally coinciding with the period of validity of the license. The survey tracked these sample permits in order to ascertain in which quarter and year during the subsequent period, the owner or the builder who obtained the permit actually started the construction. The information was obtained using the Schedule –Part II, given in Annex 7.

3.2.4 Sampling method: The sample selection for the survey was based on a stratified sampling method in which the units in each stratum were randomly selected. In each administrative/tax zone/ward, the data on permits were further stratified based on the type of the building (Single Housing Unit (SHU) or Multiple Housing Unit (MHU)). For example, if a particular centre had 5 zones, each zone was further stratified into 2 strata. i.e. in total 10 strata. In each such stratum, 5 per cent sample of the total building permits for new residential construction was selected based on systematic sampling procedure. If the 5 per cent of the total happens to be fraction, the next integer was taken as the sample size. If total number of permits in a stratum was less than 10, then all permits were taken to constitute the sample. If 5 per cent of the total number of permits in a stratum turned out to be less than 10, then the sample size was taken as 10.

3.3 Housing Start Rates (coefficients)

3.3.4 This matrix HSRM_transformed is to be used for estimating the housing start figure for a particular quarter. The first row of HSRM_transformed  gives the coefficients of housing start for the first quarter of the calendar year, based on permissions given in previous quarters. Similarly the second, third and fourth rows correspond to the second, third and fourth quarters of the calendar year respectively. The four series of start rates (coefficients) corresponding to the four quarters are estimated to incorporate the factor of seasonality.  This matrix of start rates can be obtained separately for SHU and MHU and can be used for estimating the housing start in each quarter separately for SHU and MHU.

3.4  Compilation of Housing Starts

3.4.3 The number of housing starts can be estimated by two methods: i) by adding the number of housing starts corresponding to SHU and MHU giving equal weights; ii) by adding the number of housing starts corresponding to SHU and MHU giving weights proportional to the average Floor Space Area (FSA) corresponding to each category.

3.5 Compilation of Housing Start-Up Index

Where n is the number of centres, Ai0 is the average FSA in the i^th centre in the base period; S_it is the number of housing starts in the t^th quarter in i^th centre; S_i0 is the number of housing starts in the base period in i^th centre.

4.1.8 Table 4.1.3 provides the average Floor Space Area (FSA) in Sq.ft in the 4 zones for SHU and MHU. The average FSA did not vary much across the zones. However, the SHU were found to have higher average FSA than MHU.

x: Not computable  --: Nil/Negligible

x: Not computable  --: Nil/Negligible

4.2 Case Study of Villupuram

MHU. The numbers of permits disaggregated by different quarters of the year are given in Table 4.2.1.

      x: Not computable  --: Nil/Negligible

          x: Not computable  --: Nil/Negligible

Table 4.4.1 - Total Number of Permits Issued in Saswad

        x: Not computable  --: Nil/Negligible

         x: Not computable  --: Nil/Negligible

                                                      --: Nil/Negligible

4.7 Difficulties and Limitations of the Exercise

4.7.1 All requisite information sought to be collected through Schedule Part I are available with the permit issuing authorities. Unfortunately, these are not available in the computerised form in many cases. It therefore became necessary to cull out the information from the registers and files maintained mostly in hand written form.

4.7.2 In the permit data, collected through Schedule Part-I (Survey of Building Permits), the data corresponding to some variables, for example number of housing units in the building, plinth area of the building and height of the building are not available even in the registers maintained by the concerned authorities. Hence the field staff had to go through the individual files for collecting the requisite information. In some centers (for example, Delhi), even the files did not contain the required information in several cases.

4.7.3 The information on the number of units in a Multiple Housing Unit (MHU), for which permission has been given, have been collected through Schedule Part-I but not very satisfactorily. This is a very important piece of information for working out housing starts. Importantly, this information was captured in the survey through Schedule Part-II (Sample Survey of Housing Starts) which was used for estimating the total number of housing units in the permits.

4.7.4 In case of MHU, it is difficult to ascertain the exact date by which construction can be taken to have started for each individual housing units. Hence all the housing units in an MHU are assumed to be started when the excavation or footing was laid for the building.

4.7.5 Some of the sites for which permits to construct houses were issued were found to be not occupied at the time of the survey and consequently the Part II schedule could not be canvassed. In some cases, the builder after handing over the building to the owner had shifted to a new address. In these cases, too, certain vital information like time of the building start could not be collected.

4.7.6 Some divergence was noted in the type of the permit issued as per the official records and the actual permit document, during the field survey. Some of the SHU permits were found to MHU and vice versa. However, for the compilation of the start rate matrices, the data from the field survey (Schedule Part II) were used.

Notes:

ξ Test statistics for testing the equality of proportions is given by [(s₁-s₂)/SE] which follows N(0,1) asymptotically, where s1 and s₂ corresponds to the start rates obtained in 2003 and 2004.
$ The SE in the difference of start rates is obtained by SQRT (s₁(1-s₁)/n₁+s₂(1-s₂)/n₂)

Section 5

RECOMMENDATIONS

5.1 Objective and Scope

5.1.1 The objective of constructing a Housing Start-Up Index (HSUI) is to measure the change in the level of activities in housing sector and to identify the growth/recessionary tendencies in this and related sectors of the economy.

5.1.2 The scope of HSUI would be limited to new built residential units in urban India, whose construction is authorised through issuance of building permits.

5.1.3 The HSUI may be constructed based on two sets of data

(a) The start up coefficients reflecting the recent experience of conversion of housing permits into housing starts and (b) the number of permits issued during the last two years or so. The Group recognized the fact that besides these, actual housing starts in any quarter are likely to be influenced by a host of other factors like price of building material, interest rate for housing loans, policy pronouncements, legislations, administrative orders affecting construction sector etc. To a certain extent, these factors would affect the demand and supply parameters of the housing market and consequently the number of permits issued, with possibly a time lag of a few months. One could, therefore, argue that these factors are included in the model through the number of permits.

5.1.4 The start up coefficients, computed from the data in recent past, reflect institutional and social response to housing permits in terms of their conversion into actual housing starts. The time required for administrative and procedural clearance after the issuance of permits, time taken to complete the formalities of obtaining loans, organizing material, community rituals etc. are considered to be rigid or fixed, at least in the short run. As housing is a long term decision, predictions based on these coefficients, that reflect to some extent procedural and social rigidities governing the house construction process, are likely to be fairly reliable.

5.1.5 The objective is to release the HSUI that can be used by housing related agencies as the basic or core predictor. The agencies can combine the index with other short term indicators and policy variables to come to more definitive projections of housing activity.            

5.2 Periodicity

5.2.1 The Group observes that the present system of data collection, as reported by National Building Organization along with its formats may be fine tuned to obtain the requisite data on building permits on a quarterly basis. This would constitute the basic information for constructing the Housing Start-Up Index. 

5.2.2 Conducting a field survey with adequate coverage of the urban centres that can be representative of the country as a whole would be the first step in institutionalising a system for regular release of HSUI. The Group recommends that this survey may be conducted once in three years for estimating/updating the start rate matrices for each of the selected centres. These start rates are to be used for computing the housing start figures in each of these centres using the data on building permits. These can then be aggregated to construct and release the HSUI on a quarterly basis.

5.2.3 The field survey for estimating the start rate matrices can be done in two phases. In the first phase, the data on building permits can be collected in the selected centres using Survey on Building Permits (SBP). The reference year for this survey would have to be three to four years before the date of conducting the Survey on the Housing Starts (SHS).  The SHS is to be conducted in the second phase to determine the percentage distribution of the housing starts over the eight quarters (after the issuance of the permit, including the quarter of issuing) and thereafter and build start rate matrices in each of the centres.


5.3 Geographical Coverage

5.3.1 The Group recommends that as an initial effort, the HSUI may be launched based on such coefficient matrices constructed for 6 Metros and a select sample of Class I towns based on  SBP and SHS. Small towns can be included at a later stage as it is found from the pilot survey that the number of housing starts in small centres is relatively less.
           
5.4 Sampling Method and Tools
           
5.4.1 The urban centres can be selected based on the number of permits issued as per the latest available data on building permits, by using probability proportional to size (PPS) method, so as to adequately represent of the total building permits issued at All India level. At the initial stage, when the exercise is limited to large cities only, appropriate adjustment in the method of city election may be made to make the sample representative of all class I cities.

5.4.2 Survey of Building Permits: The SBP can be conducted by collecting the details of all permits issued for new residential construction from permit issuing authorities of the selected centres, using Schedule –Part 1 given in Annex 7.

5.4.3 Survey on the Housing Starts: In this survey, a sample of permits issued for new residential buildings can be identified. The follow-up of this sample permits can be done by canvassing the Schedule–Part II of the questionnaire, as given in Annex 7, to the person responsible for the permit or the concerned builder. Information relating to starting of construction and other necessary aspects are to be obtained through this survey. A stratified sampling procedure as described in section 3.2.4 can be adopted for this survey in each center.

5.5 Estimation

5.5.1 Housing Starts Rate Matrix: From the data collected from Schedule –Part II,the number of sample houses started in all the succeeding quarters, staring from the quarter in which the sample permits are issued till the latest period can be obtained. In all the centres where pilot study has been conducted, it was noted that within 8 quarters of permit issue, more than 95 per cent of the construction gets started. Hence, all the housing units where construction started, after two year of issue of permits (8 quarters including the quarter in which permits are issued) are added to obtain the final residual aggregative coefficient. Based on this data corresponding to each quarter of year, 9 start rates (1 for the quarter in which permits are issued + 7 for the 7 succeeding quarters + 1 for all the starts after 1 year) corresponding to 4 different quarters of a year are worked out. Thus we estimate a 4x9 matrix of start rates (coefficients). This matrix can be transformed as described in section 3.3.3 to get the housing start rate matrix. The methodology for estimating housing start figures for release on a regular basis has been discussed in detail in section 3.3 of this Report.

5.5.2 It is found in the pilot study that the information on the number of housing units in a  MHU is not available in most of the urban centres. The Group suggested that SHS can also be used to identify the average number of housing units in a MHU for different cities. This would be useful for estimating the number of housing units authorised to build through permits, in cases when the figure is not available in official records relating to issuance of permits.

5.5.3 Housing Starts: These start rate (coefficient) matrices can be used to obtain the housing starts for the selected Metros and Class I towns for which the regular data on building permits can be obtained without any difficulty on quarterly basis. The number of  houses started in a particular centre for a particular quarter can be obtained by multiplying the start rates coefficients with the corresponding number of permits issued in that quarter and preceding quarters. The choice of the set of start rates to be used depends on the quarter for which the start rates are to be estimated. The methodology for compiling the housing starts is described in detail in section 3.4 of this Report.

5.5.4 Housing Start-Up Index: Separate HSUI can be compiled for different Classes of the centre (for example, Metros, Class 1 cities etc.) as well as for All India level using the year of Survey on Housing Starts as the base year.  The formulae given below can be used to estimate the HSUI for the quarter t.

5.6 Institutional Arrangement 

5.6.1 A Standing Committee may be set up by the Reserve Bank of India to launch this initiative, monitor its progress, commission and overview the surveys for constructing start up matrices and consider increasing the scope and coverage of HSUI over time. It would have official members from the Central Statistical Organisation, office of the Registrar General, Ministry of Housing and Urban Poverty Alleviation and other concerned government departments, besides a few experts in the field.

5.6.2 National Buildings Organisation, Ministry of Housing & Urban Poverty Alleviation, Government of India is the nodal agency for collection and dissemination of housing and building construction statistics in the country. The Group recommends that NBO may collect the data on building permits issued for the new residential buildings in various centres (metros and class I cities at the first stage) across the country on a quarterly basis under the overall guidance of the Standing Committee.

5.6.3 Surveys to determine housing starts coefficients may be conducted every three years to examine the validity the matrix in use and identifying the areas where further research needs to be done to increase reliability of the estimates. RBI may coordinate with NSSO and NBO for the survey, based on which start rate matrices can be constructed for compilation of the HSUI.

5.6.4 An Advisory Committee on HSUI may be formed at NBO to guide and oversee the entire process of compilation of housing permit data from concerned local bodies and the Department of Economics and Statistics of the state governments, as specified by the Standing Committee, on a regular basis. The Advisory Committee may have members from Reserve Bank of India, National Statistical Commission, Central Statistical Organisation, National Sample Survey Organisation, Directorate of Economics and Statistics of select states that have a large number of Class I cities.

Technical Advisory Group on Development of Housing Start-up Index
In view-of the need, for development of a Housing Start-up Index to monitor the movements of the Indian economy on a regular basis based on appropriate methodology and the international best practices, it has been decided to set up a Technical Advisory Group on ' Development of Housing Start-up Index '.

2. The Terms of Reference of the Technical Advisory Group are as given below:
(i)           To review base paper on concepts, methodology, approach to generate the data base for construction of the indices and suggest a feasible methodological framework for construction of HSUI for the Indian economy, with a view to assist monetary policy formulation, and to guide and oversee its
implementation.
(ii)         To recommend motjalities of entrusting the work for construction of HSUI by appropriate external agency or institution, including scope of work and
. deliverables.
(iii)        To evaluate the work of the external agency/institution and recommend its acceptance by the Bank.
(iv)        Any other issue as deemed necessary for development of the HSUI.