Will the Real Hyderabad Please Stand Up?
Each passing day, there are new announcements affirming the Telangana State Government’s resolve to make Hyderabad a world class city and a smart city. These announcements are coming against a backdrop of enormous amounts of digital data being generated and shared to various degrees of openness by government agencies, citizens and corporations. There has hardly been a month – since the new government has assumed power – when some spectacular high tech event has not been organized in Hyderabad. In all this excitement however, it appears that some basic issues of data hygiene have not received the attention they deserve.
Let us be clear. This is not a problem unique to Hyderabad. Data scientists and modeling experts struggling to generate useful insights about our cities will vouch for it that data-work in India is in large measure simply grunt work. It is simply about ensuring that data is clean, portable and that one dataset is usable in relation to another. Some times, these problems are intractable because we do not even know where to begin. But Hyderabad has the opportunity and leadership now to attend to these issues with alacrity and diligence. We appeal for stronger commitment and a reorientation of collaborations between government agencies, corporate entities and civic groups towards problems of data. These problems need to be attended with diligence if we are to mitigate urban chaos to any degree. Specifically, we call for greater clarity on precisely what we mean by Hyderabad city in spatial data terms. This means clarity and transparency on the spatial boundaries of the various jurisdictions that govern and organize the urban agglomeration that houses twelve million people (according to the Integrated Household Survey) .
The problem that we describe here may seem too narrow and technical and of interest only to experts in the first instance. But bear with us for it will soon become clear why this problem has serious consequences for everyday life in the city for very ordinary people.
Here is how it all began. The Census of India recently uploaded literally hundreds of excel sheets of information pertaining to the House-listing and Housing Census Data of 2011 at a Village and Ward level. This data is of enormous value for research and for policy-making at the national level and the local level. So, we were very excited at the new opportunities as the data reveals percentages of households to total households by amenities and assets. There are 290 columns of data (from percentage of houses with thatched roofs to how many of them use kerosene for cooking) right down to the lowest administrative unit of a sub-district. So we began to assemble the data for spatial analysis we found that it was impossible to match the data with the boundaries. In lay terms, we just found too many shapes and datasets all of them claiming to be Hyderabad. Hence the title of this essay: Will the real Hyderabad please stand up? Unless we can pin that down, one would always have to take any claim about Hyderabad with a strong measure of skepticism.
Let us begin with a brief explanation of the source of this problem and its consequences, for those less familiar with urban data.
Multiple Sources of Data and Confusion
Most of us assume that we know exactly what we mean when say Hyderabad city. But in reality it is not that simple. The city of Hyderabad can be many different things. Overall, the contiguous urban agglomeration of Hyderabad is known as Hyderabad Urban Agglomeration (HUA). HUA is only a physical description of continuous urban spread. It does not refer to any administrative boundaries. It is spread over three districts – Hyderabad, Ranga Reddy and Medak.
When we say Hyderabad, we could be referring to the central or core part of the HUA that is the Hyderabad district administered by the district collector. The district collector, an IAS officer, heads the administration of several line departments such as health, education, industries etc., including the revenue department which is responsible for all land related information and transactions. For ease of administration, the collectorate oversees sub-districts known in Andhra Pradesh as Mandals. Each Mandal is administered by a Mandal Revenue Officer. Hyderabad district has 16 mandals. The Hyderabad district has a boundary which separates it from the surrounding Ranga Reddy district which is again administered by a collector overseeing 37 sub districts or mandals. 12 of these are part of the HUA.
But, then when we say Hyderabad city, we could also actually be referring the Greater Hyderabad Municipal Corporation (GHMC). Until 2007, Hyderabad city was administered by the Municipal Corporation of Hyderabad (MCH) and parts of Secunderabad was under the Secunderabad Cantonment Board. Boundaries of both these collectively formed the boundary of Hyderabad District. The GHMC is an urban local body with an elected house headed by a mayor and an executive headed by the commissioner (an IAS officer reporting to the Municipal Administration and Urban Development department secretary).
The Multiple Administrative Divisions within Hyderabad UA – Before 2007 on the left and After 2007 on the Right.
In 2007, more parts of the HUA were incorporated into the civic corporation limits to form Greater Hyderabad. The Greater Hyderabad Municipal Corporation that administers the city is divided into 150 wards (proposed to go up to 172 soon) and 18 circles. These 18 circles are clubbed into 5 zones very appropriately named North, South, Central, East and West Zones. These wards cover all of the subdistricts (mandals of Hyderabad district) and in addition include 12 mandals of Ranga Reddy district.
The GHMC is not alone in administering Hyderabad UA. The Secunderabad Cantonment Board (SCB) takes care of most of the defense lands and the residential colonies in Secunderabad. It is divided into seven wards. Osmania University is independently handled by the administration of the college.
A rough sketch of the civic boundaries of Hyderabad. Hyderabad District is surrounded by Rangareddy District on all sides.
HUA is spread over four parliamentary constituencies – Hyderabad, Secunderabad, Malkajgiri and Chevella. In.In terms of the state assembly constituencies, the HUA is spread over 26 of them across three districts.
IALAs and other special jurisdictions
In addition to these there are also 11 Industrial Area Local Authorities which serve limited service provisioning role but without an elected citizens body in HUA. Then there is the Secunderabad Cantonment Board under the Defence Ministry where the state’s revenue related laws do not apply automatically.The HUA itself is ensconced in the Hyderabad Urban Development Authority development plan administered by the even bigger Hyderabad Metropolitan Development Authority’s planning jurisdiction.
It can be easily surmised from the above that multiple overlapping jurisdictions with unclear function and resource distribution among all these agencies can lead to a very high degree of informality which cannot be easily resolved by any single agency. The problem, however, is not restricted to a lack of clarity with regards to shapes on maps. Each of these agencies not only have their own jurisdictions, but they produce their own data. The challenge that arises is to make sense of this data in the absence of clarity about the boundaries of jurisdictions. Also, it becomes extremely difficult to compare data generated by different agencies.
It is against this backdrop that we began trying to make sense of the census data.
Census data tables and maps do not match
In the “Percentage of Households to Total Households by Amenities and Assets (India & States/UTs – Village and Ward Level)” datasheet for Andhra Pradesh, the data is given according to the Revenue divisions of the administrative unit (ie district) .
Case 1 : Golconda
In the case of Hyderabad district – according to the data sheet, the Golconda Sub District is composed of four sub parts – GHMC (MCorp) (Part) – Ward 68,69,70 and 71. A quick coloring of the GHMC election wards would reveal the shape in Image 1.
Image 1 is what Golconda Subdistrict should look like according to the data tables. Image 2 shows the Census’s map that is provided in the administrative atlas provided by the Census of India – Golconda Subdistrict in yellow.
The Golconda Subdistrict, according to the Census Data’s Excel Spreadsheet shows a rhino horn like protrusion in Image 1 . This is with the assumption that when the sheets says ward no 68,69,70, 71 – they mean all of it. Or does the part in brackets mean that only a part of the GHMC wards are part of the mandal? This ambiguity is also found in other wards that fall in the sub-districts of Ranga Reddy and Medak district.
Case 2 : Secunderabad Cantonment Board
In the Census Database sheet that we have been referring to , all of the cantonment board is under the Tirumalgiri Sub District. All the seven wards of it. But if the below image is the Cantonment Board area, how is it missing in the Hyderabad district shape? Refer to images 4 and 5 below.
Image 3A is the ward map of the Secunderabad Cantonment Board. Image 3B is a rough sketch of the GHMC area being spread over the Hyderabad, Rangareddy and Medak District. Image 3C shows the Census’s portrayal of Hyderabad District. Note the missing shape of the Cantonment board in 3C.
Census map and Telangana state government’s map do not match
This discrepancy is not limited to the subdistricts. In fact, the shape of the Hyderabad District as per the Census’ Administrative Atlas is at odds with the shape of the District as per the Telangana state.
Image 4 is Telangana State Government’s official representation of Hyderabad. Image 5 is the Census’s version of the Hyderabad District
The shape of the district in both the images is not the same. The Southern tip of Barkas and the northern protrusion of Marredpally and the Cantonment is missing in the left.
What is the actual boundary of Hyderabad district?
The shapefile of Hyderabad district that is floating around on the Internet and among planning students is this. Surprisingly, this shape is seen even on most Government related GIS portals such as this one.
Hyderabad District in Yellow
This is again in contrast with the above two versions of Hyderabad District. It is not clearly defined and areas such as Masab Tank and Punjagutta are depicted as being outside the district, which they are not.
Consequences of spatial data chaos
This lack of clarity on what actually is the boundary of Hyderabad District poses the following issues:
- Which areas actually fall under Hyderabad District and which ones fall under Rangareddy district? This is unclear in the case of the outer boundaries.
- Which areas fall under which sub district? This is unclear too, because the boundaries of the revenue department do not follow the civic boundaries.
- How does one make sense of the rich sub district level data when the boundaries of these sub districts are not available?
- The description of the boundaries in the Census Data is explained using Municipal Wards. How can it be better explained?
To get some clarity on these issues, we visited the Survey Bhavan, Narayanguda to get the official district maps. The official maps in their possession were the old revenue maps of the Nizam state and the maps from when Rangareddy District was formed (somewhere in the 70s) They did not have a more recent map because the boundaries of the district haven’t changed since the formation. Their maps resembled the Census’s ones.
According to knowledgeable sources, the revenue district boundaries (mandal boundaries) in the case of Hyderabad are arbitrarily drawn and knowledge of the actual boundaries is available only either with the Revenue Department, Census Department or the NIC. These mandal boundaries do not have much in correlation with the Civic boundaries of wards.
Understandably, policymakers and executives primarily working in routine maintenance tasks or in project mode have imaginations which are limited to a single project and do not have time to rectify these problems. However, even for long term policy making and meaningful project execution – not to speak of researchers, informed citizens and consultants – these inconsistencies in data can be crippling. Without straightening these glitches, generating more data as we did in the Integrated House Hold Survey can only complicate the problem even further.
As Hyderabad aspires to become a smart city, the only thing we can say with any certainty is that the first step would be cleaning up and aligning crucial datasets and making them public and transparent.