Alterations to the municipalities in the Population Censuses since 1842



The main activity is to provide a catalogue of all the municipalities that have been considered in the Censuses in Spain from 1842 to our times. Limitations have appeared given that the institutionalised Population Censuses have been used as sources of information, as well as the so-called Cadastral Register Census dated 1842. This was done because it was the first to present the total ensemble of municipalities covering the Spanish territory, meaning that those municipalities created and those that disappeared during the same intercensal period are excluded.

Secondly, the evolution of these municipalities is presented, noting the alterations to the main characteristics included in the censuses considered. Three quantitative and three circumstantial characteristics have been chosen as the main municipal characteristics. The first three are the de facto population, the de jure population and the number of households (registration certificate). One line is reserved for each of these characteristics in the space allocated to each municipality. In terms of the circumstantial characteristics, change of name, variations in territory size and the existence of the municipality have been included. These are indicated by symbols located in the census year column and are explained in the last row of the table corresponding to each municipality.

Definition of the characteristics considered

De facto population:Number of persons that spent the night in the municipality on the census reference date.

De jure population:Number of persons that were officially resident in the municipality on the reference date.

Resident population:Term used in the 2001 Census that is the complete equivalent of de jure population.

Linked population:Ensemble of persons that can be considered in the Census (in other words, with regular residence in Spain) who have some kind of regular link with the municipality because they live there, study there or because, even though it is not their regular residence, they spend certain periods of time there (summers, bank holidays, weekends). The linked population concept was used for the first time in the 2001 Census with the aim of better estimating the real population load that each municipality had to bear and replacing the de facto population concept, which disappeared in accordance with the withdrawal of the non-resident concept in the Register of inhabitants.

Similarly, the three first Censuses do not qualify the legal situation, de facto or de jure, and deal with the regular population. In the first census, the de jure population is assumed, whereas in the 1857 and 1860 census, the population can be assumed to be the de facto population.

Households:Similar but not identical concepts are included in each Census. This concept more or less defines the group made up of one resider and those persons who live with him/her. In the Censuses, this is used as the information unit when obtaining data and all members should be registered on the same certificate, form, sheet or family questionnaire.

More specifically, these are the concepts that appear in each census: 1842, residers. 1857, 1860, 1877, registration certificates. 1887, 1897, 1900, 1910, 1920 and 1930, certificates included. 1940, census certificates. 1950 and 1960, number of sheets included. 1970 and 1981 , number of families. 1991 and 2001, number of households (the difference between family and household is that the household can be an individual and where there is more than one person, the members don't necessarily have to be related). In turn, the concept of household, which in the 2001 Census is defined as a group of persons resident in the same family dwelling, differs from the household concept used in the 1991 Census. The condition of the residents sharing certain common expenses has been removed.

Municipality denomination:The denomination that appears at the side of the table corresponds to the current one (2005). If the municipality has disappeared, the name that is included is the one that appears in the 1970 Census, and if it had already disappeared by this date, the denomination registered at the time of its disappearance is included. Similarly, if the municipality has now disappeared, but was created after 1970, the denomination that appears at the side of the table is the one it had when it was created.

A correlative digit is allocated to successive name changes that are different from those that appear at the side of the table. This is done by chronological order of use and in the last row, this number is repeated stating the name taken.

Territory variations:The only variations to be considered are where the municipal area has grown or decreased in the period between the Census in question and the previous one.

Municipality existence:It is not necessary to indicate whether the municipality exists or not at the moment when the Census is taken, as seeing if there is data or not is enough. In this characteristic therefore, only the intercensal period in which it was created and when it disappeared is indicated.

Coding of the municipalities

The best solution for distinguishing between municipalities is to allocate them a code that they will keep, regardless of the changes that could be made and even if the municipality is dissolved or disappears.

All municipalities in existence on the Census reference date, together with their de facto and de jure populations, are presented in alphabetical order by province in Book I of the appropriate publication since the 1877 Census (in previous censuses, they were ordered by areas) and, as of the 1887 Census, they are numbered correlatively. This numbering should not be considered a code, as it is not usually used for anything else and changes from one Census to the next. It was in the 1970 Census that this numbering acquired the level of coding, because as well as this purpose, it also methodises, includes collective and singular population entities and does not change in subsequent Censuses. The official coding of municipalities in the 1970 Census is composed of five numbers: The first two numbers correspond to the province code and the other three are a code used to differentiate the municipality in the province, in other words, an exclusive code for the province that grants each municipality a correlative number from 001 onwards by alphabetical order of the names. The highest numbers reached were 485 in Burgos, 382 in Salamanca and 335 in Guadalajara.

In this database of Intercensal Alterations to the Municipalities, the official coding composed of the code allocated in the 1970 Census and the regulation that states that municipalities created in the future will take the province code followed by the number allocated correlatively according to the order in which it is created within the province itself starting at 900, will be taken.

Consequently, all municipalities that disappeared before 31-12-1970 are excluded from the official coding. This publication has created use codes for these municipalities, which are explained hereunder: As with the previous codes, the first number corresponds to the province, the following number corresponds to the correlative number, in alphabetical order from 500 onwards. If the municipality disappeared between the 1842 Census and the 1857 Census, the same applies, although numbers start from 5000 onwards. In summary, the municipality code is composed of the province code + exclusive province code; the latter is established as follows:

From 001 to 449 .............municipalities existing in the 1970 Census

From 500 to 899 .............municipalities that disappeared between the 1857 Census and 1970 Census

From 901 to 999 .............municipalities created after the 1970 Census

From 5000 to 5999 .............municipalities that disappeared between the 1842 Census and the 1857 Census

Symbols and terminology used

   Municipality created.

   Municipality disappeared.

   Municipal area grew.

   Municipal area decreased.

(*)  Alterations with singular characteristics.

[1], [2], [3],etc.: Used for municipality names, by chronological order of use, that differ from the primordial denomination that appears on the side of the table.

..  To indicate the non-existence or lack of information.

Terminology:Following the data, there is an explanation of the alterations, indicating the municipalities involved in the changes and the reasons. In order to be able to do this in the available space, the language used is extremely concise. Each case is identified by the symbol and Census year in which it appears, followed by a description of the circumstances:

   SEPARATES FROM x:  This means that municipality M is created from a part that separated from municipality x.

  • Counterpart: In municipality x the following indication features:      RELEASES M.

   Municipalities x, x', x'' MERGE:  This means that M is created from the union of the municipalities or entities x, x', x''.

  • Counterpart: In the municipalities x, x' etc. the following indication features:      JOINS M and if there was any entity in municipality x that it contained, it would say:      TRANSFERS x to M.

   CHANGE OF NAME of x:  This means that M is created because it takes the area, population and assets of municipality x, which in turn disappears.

  • Counterpart: In municipality x the following indication features:      CHANGE OF NAME for M.

   INTEGRATED INTO x:  This means that municipality M has disappeared and its area, population and assets pass to municipality x.

  • Counterpart: In municipality x the following indication features:       INCORPORATES M.

   JOINS x:  This means that municipality M has joined others to form a new municipality x.

  • Counterpart: In municipality x the following indication features:      x, x', x'' MERGE.

   INTEGRATED IN PARTS into x, x’, x’’:  This means that municipality M has disappeared and its area, population and assets have passed to municipalities x, x’, etc.

  • Counterpart: In municipalities x, x’, etc. the following indication features:      INCORPORATES PART of MorSEPARATES from M.

   CHANGE OF NAME by x:  This means that M disappears and reappears with the new name of x.

  • Counterpart: In municipality x the following indication features:      CHANGE OF NAME of M.

   NOT LOCATED:  This means that municipality M appears in the 1842 Census, but is not located in the following Census of 1857.

   INCORPORATES x:  This means that municipality x becomes part of municipality M.

  • Counterpart: In municipality x the following indication features:      INTEGRATED INTO M.

   INCORPORATES PART of x:  This means that M has incorporated one of the parts H that belonged to municipality x, which disappears.

  • Counterpart: In municipality x it states:      INTEGRATED IN PARTS into M, M’, etc.

   H RECEIVED from x:  This means that M has incorporated part H into its area from municipality x.

  • Counterpart: In municipality x it states:      H TRANSFERRED to M.

   H TRANSFERRED to x:  This means that municipality M has given part H to municipality x.

  • Counterpart: In municipality x it states :      RECEIVES H from M.

   RELEASES x:  This means that part x of municipality M separates to form a new municipality.

  • Counterpart: In municipality x the following note is found:      SEPARATES from M.

Different municipality denominations: The number features in chronological order of use [1], [2], etc. followed by the name.

Institutional Censuses

Chapter taken from the piece of work entitled Censos de población españoles (Spanish Population Censuses), published in the INE magazine Estadística Española (Spanish Statistics) in volume 33, number 128 dated December 1991, written by D. Eduardo García España.


Soon after the start of the second half of the 19th century (1851), a Universal Industrial Exhibition was held in London, which demonstrated the difficulties and, at times, the impossibility of comparing figures from different exhibitors due to the variety of concepts, classifications and methods used. This led to the intervention of Visschers and particularly Quetelet, who founded an organisation named the International Statistics Congress, which aimed to reach a consensus on the unification of all terms, definitions and stages in statistics processes. Among the wide range of topics dealt with at subsequent meetings of this Congress, the most important were those relating to Population Censuses.

At the first two meetings and the fourth (London 1860), it was continually stated that they be nominative, referred to the de facto population, were ten-yearly, included information from family certificates and contained the personal data of all members, including those that were il l and disabled. For the first time, those at the meeting in London also requested that the data refer to one day only and that this day be when the population is most settled, possibly on New Years Eve. At the fifth meeting held in Berlin in 1863, it was announced that for a Census to be useful for all administrative issues, the de jure population must be known. Nevertheless, at the next meeting (Florence, 1867), it was argued that Censuses should be based on the de facto population, even though special treatment was advised for those residents who are absent. At the St. Petersburg meeting in 1872, this topic was dealt with and everything set out up until this point was ratified with some small, but significant changes: The reference date must show the day and the moment and should be carried out during years ending in zero; de facto, de jure and regular populations must be distinguished between, but only the de facto population will be used until uniform definitions are achieved that are valid in all countries.

The Population Censuses carried out in Spain during the second half of the 19th century, feature in the Censuses universally qualified as Modern Censuses. In the INE Statistical Yearbook they are called Official, but this characteristic does not differentiate them from the previous ones carried out at the beginning of this century, as these were also ordered by the authority in charge. For this reason, we prefer to call them institutional, thus avoiding them being considered as the only ones to be officially acknowledged. The two last Censuses of the 18th century could also be considered as modern censuses, but they could not be institutionalised, even though Godoy wanted this and they remained as two isolated Censuses. From 1857 onwards, there is continuity and they followed the same guidelines, thus constituting an institution in itself, which is is how the authorities who wrote the corresponding official publications presented them.

1842 Census

This was undertaken without strictly respecting the imputation procedure and therefore lacks a reference date. For this reason it is known as the "Cadastral Register Census", but for convenience and questions of uniformity, here it is called by the year in which it was taken. As a Census it shouldn't feature in this piece of work, as it was carried out in a very different way from the modern censuses and its poor quality doesn't contribute any reliable numerical data. For this reason, the population provided and presented in the column dedicated to it in this publication should be taken on board with reservations. It has been included, as it is the oldest, most exhaustive list of municipalities known. It is the first Census to take the municipality as the information unit and covers the whole Spanish territory, which means it is the ideal starting point for the topic under study in this piece of work.

1857 census

Soon after it was created, the Kingdom's General Statistics Commission had to plan a new Census that adapted to the recommendations of the International Statistics Congress and four months later, Royal Decree 14-3-1857 was announced, which contained a long explanation of how to carry out the Census composed of 80 articles grouped into eight chapters. A month and a half afterwards (3-5-1857), a new Decree set the 21st of this same month as the date for its simultaneous undertaking across the whole national territory. The results had to comply with this deadline and at the end of the year, the classification by professions was renounced, as there were difficulties in interpreting the different professions, which made it hard to classify them correctly. It took another ten months to finish the work and at the end of this time, the new President of the Government and the Statistics Commission, Leopoldo O'Donnell, was able to present the Decree dated 30-9-1858 to the Queen, thus approving the Census.

Its main characteristics are as follo ws: Its reference date is before the one advised by the International Statistics Congress, it includes the de facto population, dividing it into those that are "settled" and "non-residents" and the age groups used allow comparisons with previous Censuses to be carried out (Aranda, Floridablanca and Godoy).

The results did not meet the creators' expectations and they doubted that this piece of work would reliably reflect reality and so warned in the introduction that "the census has an inherent right to be believed in legal terms, even where this is not the case in strictly material terms" and afterwards, in this same section, the Census was allocated an experimental role and they recommend that another one be done straight away in 1860 (art. 3), to use the experience acquired "without losing recent impressions with regards the difficulties encountered".

It was published in a book dated 1858.

1860 Census

The new Census was carried out with a reference date of 25th December 1860. The most significant differential characteristics are: The trouble taken during the preliminary work to make sure that the Census framework was as complete as possible and to record the absent legal residents, thus obtaining the de jure population. When considering the first characteristic, it is important to note the following regulations:

In order to improve and complete the population entity nomenclature, leaflets were sent out giving instructions to the Governors, dated 14-8 and 31-12 1859 and 30-1 and 2-4 of 1860. The Government Ministry demanded that the municipal boundaries were marked by Royal Order 5-11-1859; regulations were brought in to signpost the streets and number the buildings by another Royal Order dated 24-2-59.

In order to understand the "de jure population", Spaniards resident abroad, sailors at sea without residence on land and troops stationed in extraterritorial positions needed to be added to Spaniards resident in Spanish territory. Royal Orders 27-12-1859 and 13-1 and 5-11 of 1860 are dedicated to this purpose and were aimed at consulates. The Orders dated 22nd and 29th November 1860 were sent to the military authorities of the Spanish populations in Africa and the occupied troops in Tetuán. A dispatch dated 23-12-1860 was sent to the Governors of the coastal provinces on how to take a Census of personnel at sea and finally, a circular dated 30-11-1860 was issued on persons resident in border towns, working temporarily in the neighbouring country.

In order to achieve truthful information, the Circulars dated 15-8 and 19-9 1860 urged the Governors, who were the Presidents of the Census' provincial committees, to maximise their conscientiousness. The same was done with t he Primary Education Inspectors (7-12-1860) and the provincial representatives (8-12-1860).

These regulations and others that we have not included demonstrate the great care with which the Census was prepared and, in turn, we are obliged to point out the most important legal regulations: The order to carry out the Census is found in Royal Decree dated 30-9-1859, together with the approval of the previous figures. Another order dated 31-10-1860 indicates the reference date (25-12-1860), the Instruction dated 10-11-1860 sets out the regulation to be followed and the Royal Decree dated 12-6-1863 approves the figures.

Once these data were compiled, another series of circulars appeared during the first few months of 1861 finishing off or clarifying the instructions given for classifying professions, giving other instructions on non-residents (for example military personnel stationed within national territory) and stipulating inspections in towns where the truthfulness of the data raised doubts. The age classifications ignored the intervals of previous censuses in order to take on others that varied between 4, 5 and 10 years and 1 year for certain ages considered to be strategic (<1 year and 21, 22, 25 years old). The classification of professions was undertaken with interest, but did not really satisfy the authors, especially because of the difficulties presented by those people with a number of professions, which was common in a rural environment. Elementary education and the traditional classifications of marital status and sex completed the classifications and, in general, correspond to the those presented in the 1856 Belgian Census and the French and English Censuses of 1861.

It was published in a book dated 1863.

Withdrawal of the 1865 Census and 1870 Census

The 1858 Royal Decree, which at the same time ordered the undertaking of the 1860 Census, established that this operation would be repeated every five years. On the 12-6-1863, at the suggestion of the Marques of Miraflores, Isabel II endorsed the Royal Decree that approved the results of this second institutional Census and insisted on repeating it in 1865. Faced with such illusion, one could think that Miraflores had well founded suspicions that it was not his place to do this. And in effect, following the end of his role during the first few days of the following year, he was succeeded by Arrazola, Mon and Ramón Mª. Narváez, Duke of Valencia on 16-9-1864, who presented the decree dated 30-9-1864 fifteen days later, stipulating that the Census would be taken at ten-year intervals as advised by the International Statistics Congress at its meeting in Brussels (1853).

However, from the end of September 1868 onwards, when preparation work on the 1870 Census should have been taking place, the political scene in Spain was unsettled. On the 30th, Queen Isabel II went into exile. On the 8th October, General Serrano took over the Regency and eight months later General Prim was ruling (18-6-1869). A new king was searched for and not found until finally, Amadeo I accepted the crown on the 31-10-1870. His reign lasted less than two and a half years and there were seven different heads of government. The first republic was set up for eleven short months, during which time there were four presidents. The Serrano Regency returned (3-1-1874) with two governments until the 29-12-1874 when Alfonso XII was declared king and a period of relative calm began with the Cánovas, Jovellar, Martínez Campos and Sagasta governments.

During Espartero's reign and just days before Prim was named head of the government, the formation of the Census corres ponding to 1870 (Decree dated 7-6-1870) was ordered, despite the undecided circumstances surrounding the country. To have undertaken this would have meant that its execution and the most awkward work would have fallen to Amadeo I's ephemeral governments. It was not the time for activities that demanded political calm and it was continuously postponed, in the hope of quieter times, until the King's last government decided to renounce its undertaking in Decree 12-8-1872, extending the validity of the 1860 Census.

1877 Census

Work on the Population Censuses began again with the restoration of the monarchy and it fell to Alfonso XII and the Martínez Campos government to announce Royal Decree 1-11-1877 ordering the formation and whose reference date is the very end of 1877. The next day, the regulations to be followed in order to carry out this work were stated and were taken from those prepared for the 1870 Census that was not taken. The data were entered during the Cánovas del Castillo government and this was done by the Geography and Statistics Institute, led by D. Carlos Ibáñez and Ibáñez Ibero. The de jure population concept was introduced for the first time, which presented difficulties it seems, as the city councils divided their population into residers and residents, taking it to the extreme of dedicating an entire page to this in the prologue to the Census publication. Data on religion and disabilities were also included and all the concepts advised by the International Statistics Congress were demanded. The Marques of Mulhacén assured the Minster of Development of these details in a letter that presented the 1887 Census.

It was published in three books dated successively in 1879 and 1883.

1887 Census

The reference date for this Census is the 31-12-1887, just within the ten year period lapsed since the previous Census and in accordance with what had consistently been ordered in the time of Godoy, but will certain problems that interfered with its taking. Finally, for the first time in this century, they managed to comply with the intercensal path they so desired. This was only achieved one other time in history and was between the Floridablanca Census (1787) and the Godoy Census (1797).

It was carried out in accordance with the Law dated 18-6-1887, signed by the Regent María Cristina de Hansburgo and approved by Práxedes Mateo Sagasta, President of the Government. In agreement with the instructions set out in the Royal Decree dated 20th September of this same year, it was undertaken by the Geography and Statistics Institute, then led by the Marques of Mulhacén during the five-year period of the Sagasta government. The Ex-General Director of Demographic Statistics at the INE, D. José Aranda, highlights the following special characteristics: the omission of the classification of inhabitants into residers and residents, the classification of persons with disabilities and the declaration of religion, as well as the inclusion of nationality, which was only considered as place of birth in previous Censuses and the population of smaller entities. This last characteristic was done in the Nomenclature of the 1857 Census.

The INE has publi shed a wonderful piece of work written by the professor David-Sven Reher based on this Census (Spain in the light of the 1887 Census).

It was published in three books dated successively in 1889, 1891 and 1892.

1897 Census

This is the last of the Censuses carried out on reference dates ending in seven, the 31-12-1897. Its execution was stipulated in Royal Decree dated 9-11-1897 and signed by the same heads of state as the previous one, but this time Sagasta had only spent 36 days enjoying his sixth government. The data were obtained as foreseen with difficulties and the first provisional results were published in the first book after being approved on 16th June 1899.

The definitive results were available when the Law of 3rd April 1900 was announced, which agreed with the international plans expressed in the last meetings of the International Statistical Institute on the undertaking of a Census where the reference date falls at the change of a century. The law goes further than that and changes the years ending in seven for those ending in zero, which is maintained until the 80s when it is delayed a few months. With the proximity of the end of the century, it was advisable to suspend work on this Census and start preparing the next one. The definitive data that had been obtained, would be published together at the beginning of 1900.

It was published in two books: One in 1899 and another in 1902.

1900 Census

There are very few new features in this Census, not only because of its closeness to the previous one, but also because since the beginning of these institutional Censuses, each one has provided relevant experience for future Censuses and the necessary improvements have been made, which means that acceptable results are obtained that are difficult to better with available means. This is the case with the previous Census and even more so with the one we are talking about now, in other words its regulations would hardly change from now on, but this doesn't mean that the methodology employed remained the same: The classifications were increased, the definitions were assessed, the procedures for obtaining data were improved, the terminology became more rigorous, the data was processes with new methods, the data omitted were evaluated and processed by increasingly sophisticated computer programmes. The tabulation process and the calculation methods were improved, from the manual tra nscription of personal data on papers of different colours and calculations carried out by hand, to powerful computers, punched cards and the multi-adding machines of the 30s and 60s.

Nevertheless, we will identify the main new features of this Census, even though they are not overly important.

As the Public Treasury used the Census population to fix certain tax bases, some city councils fraudulently changed the population to improve their quota, which meant that the checks on fraudulent persons were intensified. The tabulation tasks were reinforced so that the provisional results did not have to be published and the population of lesser entities was provided within the municipalities, something that wasn't completely new as it had been included before and was also taken in the 1887 Census. In the introduction to the first book, it is stressed that since the Geography and Statistics Institute took over the responsibility for the Censuses, it has aimed to adapt them to the recommendations agreed by the International Statistics Congress.

The data were approved by Royal Decree 24-4-1902.

Censuses from the 20th century

The only new feature of the first two entire censuses taken in this century, in 1910 and 1920, is that the concepts are better defined. The processing of dubious cases, such as the residence of civil servants and military personnel and their families or prisoners, etc, was clarified.

The next Census in 1930 was half done by the Monarchy of Alfonso XIII, who did the design and the data collection and the Republic, which did the tabulations. In doing this, electromechanical statistical machinery was used for the first time. This was conventional punching equipment and Remington classifiers. The budget allocated to the Census by both regimes amounted to 1,750,000 pesetas.

The municipal register of inhabitants relating to years ending in zero was done at the same time as the Census on separate questionnaires, but from the 1940 Census onwards, the register questionnaire was rejected and the personal data of inhabitants was passed to the city councils from the Census without taking statistical secrecy into consideration. In the next Census (1950) the identity of both figures was demanded.

Sampling techniques were used for the first time in the 1950 Census, where a 10% sample of the questionnaires was taken. Tabulation was done by punched files obtained through the "mark sensing" technique. The results were not convincing. In the next Census (1960), only a 1% sample was used to obtain the general national results, but in the two following Censuses (1970 and 1981), the sample was increased to 1% or 2% of family dwellings and all the group dwellings in order to obtain a results preview. Another 25% was used to obtain more detailed data with cost sa vings, which meant that some of these savings could be used to improve the quality of information. Sampling techniques were also used for an analysis of coverage and content errors. In the next Census, the last of the 20th century, the amazing progress in IT and the spread of private companies specialising in data recording allowed the data to be exhaustively processed and stored electronically. This meant that new tabulations could be created according to user needs and sampling in censuses was therefore abandoned.

A directive from the Council of European communities dated 22-11-1973, states that the reference date of population censuses should fall between the 1st March and the 31st May of years ending in one. Spanish law 70/1980 dated 16th December imposes this directive in Spain when the date of 1st March 1981 at zero hours was already established by Royal Decree 2810/1980 dated 14th November before the law was announced. The date is maintained for the next Census, the last of this century.

As said before, it was during the 1930 Census when mechanical tabulation procedures were used and in 1950 the mark sensing technique was used for punching cards. The conventional electric classification and tabulation equipment was by then being used and acted as "punched cards". This meant that a large number of punching machines and personnel was needed and also cutting machines to make the individual cards from rolls of card. Electronics played a role in the 1960 Census when a 1401 16K IBM computer was used, which was increased to a 360 262K IBM computer in the next Census. The card punchers were replaced with recorders and GTS is used, which is a French table generating programme. From now onwards IT equipment takes over and their power increased spectacularly. Software follows a similar pattern: In the 1981 Census, a North American table generation programme was used, as well as TPL and the error processing programme AERO from Hungary, which was replaced in the last Censu s of the century with the DIA.V2 by the INE's higher professional statisticians. The DIA programme was improved for the 1991 Census and it is used extensively.

The 2001 Census

The 2001 Census constitutes the fifth Buildings and Premises Census, the sixth Housing census and the sixteenth Population Census carried out in Spain within the group we have called institutional Censuses.

The main new feature of this Census with regards previous ones is the use of administrative registers, in particular the municipal register and the urban land register, which meant huge savings in money and less inconvenience for citizens.

As of law 4-1996 dated 10th January, which modifies law 7-1985 dated 2nd April, regulating the Basis of Local Regimes in relation to the municipal register, the National Statistics Institute is legally obliged to coordinate municipal registers in order to detect duplications and errors contained therein. In order to do this, the INE can control their accuracy, as the Population Census is one of these operations, according to what is set out in article 79 of the Regulation on the Population and Territorial Demarcation of Local Institutions set out in the aforementioned law 4-1996.

Using this regulation as a base, questionnaires were designed that separated the administrative, register data from the other Census questions and subjected to statistical secrecy. Moreover, the register information was then personalised, meaning that if it was correct, it just needed to be confirmed by signing the corresponding questionnaire. If there were any errors or it didn't reflect the latest changes to the inhabitant's administrative situation, these data were corrected on the same questionnaire and the INE was responsible for conveying the changes (registrations, cancellations and changes to the municipal register data) to each of the city councils.

The corrected register data was incorporated individually to the peoples' other Census characteristics, this avoiding duplication of questions and respecting the separation of the two information sources.

Another new feature of this Census was the way in which other organisations and even individuals were encouraged to take part via the dissemination of working documents and the compilation of suggestions. The draft, in which the pros and cons of the inclusion of each of the characteristics under investigation were evaluated, gave rise to the census project, once the proposals of potential users had been analysed.

Reducing the dissemination deadlines was also a priority. The majority of planning decisions were taken on the basis of this focus, from the questionnaire design and the choice of questions to the use of the most advanced technology for processing the information: data entry using high performance scanners, processing with character recognition and automatic codification with the help of dictionaries and dissemination via the census datawarehouse.

It is interesting to highlight the new methodological features that affect, above all, comparisons with previous censuses and which consist of two variations with regards 1991:

  • - The non-resident concept is removed and as a consequence, the de facto population concept. This is replaced with the linked population, which better estimates the population load received by the municipality.

    - The household concept becomes that of household-dwelling and is defined as a group of persons resident in the same family dwelling without requesting the requirement of sharing certain common expenses that was asked for in the 1991 Census.

The Census reference date was the 1st November 2001, which should be taken flexibly, as on the one had the fieldwork was carried out over a period of three months and on the other hand, some questions referred explicitly to activities that are carried out on a regular basis or which were carried out in the week before the interview.

The data was made available to the public free of charge on the INE website and the population figures (July 2002), the preview data (between December 2002 and March 2003) and the definitive data (from February 2004 onwards) were added step by step. Other types of media were also used for dissemination purposes: High circulation of informative leaflets, the creation of tables on the Internet, microdata files, CD-ROMs with files and graphic facilities and paper publications.