1 Introduction

Since 1988, the provision of Specialized Educational Service (SES) as a mechanism for guaranteeing access to the common school for the Target Population of Special Education (TPSE) students (^{Constituição da República Federativa do Brasil, 1988}) has been advocated in Brazilian legislation, reinforced, later, in 1996, in the National Education Guidelines and Framework Law (known as LDB) (^{Lei nº 9.394, 1996}). However, the adoption of the term preferentially in the legislations in question gave rise to the creation and maintenance of environments that served these students in segregated spaces of education (^{Mazzotta, 2005}; ^{Bueno & Meletti, 2012}). Still, the Federal legislations of this period were about signaling the possibility of existence of the SES, without, however, systematizing in which spaces and in what form it should occur.

In 2003, there was a reorientation of education policies, both in the Basic and Higher Education (^{Dourado, 2007}). In line with these changes, in 2007, the Fund for the Maintenance and Development of Basic Education and Valorization of Education Professionals (known as FUNDEB) was established (^{Lei nº 11.494, 2007}), which directly influenced subsequent education policies, since education funding policies concern the financing of educational management processes, including the role of the Federal Government and the articulation between the public and private spheres (^{Dourado, 2007}). In this sense, the education policies for Special Education also underwent major modifications, also referring to international documents that provided the inclusion of the TPSE students in the regular system of education (^{Kassar, 2016}). Thus, there was the creation of programs and actions that aimed to promote the school inclusion of TPSE students in the common classroom (^{Kassar, 2012}).

The current policy of school inclusion is composed of 28 legal texts that specifically deal with the characteristics of this policy and the service provided therein. It also includes other more specific legislation, such as the evaluation of students with intellectual disabilities, the duties of the school support professional, among others (^{Santos, 2017}). In this sense, the enrollment of the TPSE student in the common classroom and extra hours classes in the SES is currently required; thus there is with double accounting of FUNDEB for those students who are properly enrolled in the School Census in both spaces (^{Decreto nº 7.611, 2011}). According to this policy, the SES consists of a "set of activities, accessibility and pedagogical resources organized institutionally and continuously" (^{Decreto nº 7.611, 2011}, p. 2), and it must occur in Multifunction Resource Rooms (MRR) or in Centers of Specialized Educational Service (CSES), in a complementary way for students with disabilities or with global developmental delay (GDD) and supplementary for students with high abilities/giftedness (HA/GD).

This is the only legally established national service to promote support for the inclusion of TPSE school students, whose limitations have been pointed out in the literature (^{Mendes, Cia, & Tannús-Valadão, 2015}) and signify a simplification of the concept of SES, which is not based on the literature of the area regarding the assistance of the distinct population that makes up the TPSE (^{Santos, 2017}). It is observed that there is the absence of governmental proposals that focus on activities that promote the partnership between teachers of the common classroom with those of the SES, through strategies of pedagogical support or co-teaching, for example (^{Mendes, Vilaronga, & Zerbato, 2014}). Also, there is a clinical view about SES that is based on the assumption that full knowledge of Braille or the Brazilian Sign Language (named LIBRAS) could be sufficient to respond to the specificities of the blind and deaf students respectively (^{Bueno, 2011}).

From the analysis of 28 legal documents that make up the current school inclusion policy, it has been observed that there are texts that only include students with disabilities, leaving aside those with GDD or HA/GD (^{Santos, 2017}), thus showing the possibility that the voices of students with disabilities are more heard than the voices of the other two populations that compose the TPSE. As a possible reflection of this observation, there is the inclusion, in 2012, for legal purposes, of students with Autism Spectrum Disorder (ASD) in the group of students with disabilities (^{Lei nº 12.764, 2012}).

According to the fifth edition of the Diagnostic and Statistical Manual of Mental Illness, the characteristics of Autism Spectrum Disorder (ASD) consist of: a) persistent impairment in reciprocal social communication and social interaction; and b) restricted and repetitive patterns of behavior, interests or activities. Previous conditions of early childhood autism, childhood autism, Kanner's autism, high functioning autism, atypical autism, global developmental delay with no other specification (GDD-WNOS), Childhood Disintegrative Disorder and Asperger's Disorder were included in the ASD terminology, in which a dimensional (spectrum) perspective is shown to represent the great variability within the disorder and within a symptomatological set of signs and symptoms, and there are classifications of mild to severe that are distinguished into three main levels (1, 2 and 3), depending on the amount of help needed to develop daily activities (^{American Psychological Association [APA], 2013}). In addition, the symptoms must be present in the initial period of development, but may not manifest fully until certain social demands are presented, such as when the child starts attending school. The symptoms cause clinically significant impairment in social, occupational, or other important areas of functioning, and such disorders are no better explained by intellectual disability or global developmental delay (^{APA, 2013}).

Regarding the analysis of education policies, ^{Jannuzzi (2005)} defends the position of educational indicators as a way to support the diagnosis, formulation, implementation and evaluation of public policies. In this same sense, ^{Meletti and Bueno (2010)} put the analysis of the indicators provided by the School Census as a possible way to analyze this policy, since these data are of great relevance, as, according to information officially provided, they consist of "the main tool for collecting basic education information and the most important statistical survey of Brazilian education in this area" (^{Instituto Nacional de Estudos e Pesquisas Educacionais [INEP], 2018}), being an "indispensable tool so that the educational actors can understand the educational situation of the country, (...) as well as of the schools and, with this, to follow the effectiveness of the public policies" (^{INEP, 2018})

This methodology of policy analysis has been used by Brazilian researchers throughout the country. From a literature review, which considered both scientific papers and dissertations, it was observed that, at the national level, these data were analyzed considering both TPSE students as a whole (^{Caiado & Meletti, 2011}; ^{Castro, 2015}; ^{Côrrea, 2012}; ^{Cruz, 2011}; ^{Meletti, 2014}; ^{Meletti & Ribeiro, 2014}; ^{Nascimento, 2014}), and the students with disabilities only (^{Carvalho, 2012}; ^{Kasper, Loch, & Pereira, 2008}; ^{Laplane, 2014}; ^{Hass & Gonçalves, 2015}; ^{Martins, 2012}; Meletti, & ^{Bueno, 2011}; ^{Rebelo, 2012}; ^{Sá, & Cia, 2015}; ^{Souza, 2012}). ^{Castro (2015)} and ^{Côrrea (2012)}, besides conducting the national analysis, also compared the data by each country region, allowing to know the specificities of each region. Both authors considered TPSE students in general and had a historical cutoff from 2007 to 2010 (^{Côrrea, 2012}) and from 2009 to 2013 (^{Castro, 2015}). This specification by region of the country is relevant, since there are disparate realities when considering both population and demographic aspects.

As for the public with ASD, only two studies were identified (^{Lima & Laplane, 2016}; ^{Talarico & Laplane, 2016}). ^{Talarico e Laplane (2016)} had as an objective to map the school trajectory of the students with autism, Asperger's Syndrome, Rett Syndrome and Childhood Disintegrative Disorder, in the city of Campinas, with a historical cutoff between 2009 and 2012. Although the authors considered the students identified as having Rett Syndrome from DSM-5 (^{APA, 2013}), this syndrome does not fit the population with ASD, going against the population indicated by the authors in the introduction of their paper. ^{Lima and Laplane (2016)} had as their objective to analyze the access and the permanence of the students with autism in the school and to verify what therapeutic and educational supports they received. For this, part of the data of the authors consisted of the microdata of enrollment of the students with autism of the city of Atibaia (São Paulo - Brazil) registered in Basic Education between 2009 and 2012. Again, the conceptualization of ASD as a whole was made in the introduction; however, in the method, the authors considered only the students with autism, as predicted by the research objective. The authors also pointed out that students with autism constituted the majority of the sample of students with ASD.

Regarding the analysis of education policies, ^{Power (2011)} stated that there are two possibilities for analysis, depending on the direction of the researcher's gaze. The first, within a macro scope, has a larger sample range (for example, states, regions of the country or the country as a whole), aiming to observe broader issues, while the second , within a micro scope, is focused in detailing and accumulating descriptive information about a particular location. In this sense, the studies of ^{Lima and Laplane (2016)} and ^{Talarico and Laplane (2016)} focused on the students with ASD in a single city, which did not allow an analysis of the school inclusion policy at the national level. Thus, there was a gap in the literature, including with regard to the specific analysis of ^{Law No. 12.764/2012}, whose impacts were not discussed by the authors and, according to ^{Ball (1995)}, this Law can be considered as an attempt to achieve greater political representation for this public and directly implicate in the practical scope of the policy of school inclusion and enrollment of these students.

While legal documents and legislation address the right to education, preferably in the common classroom, for TPSE students there are no data in the literature that allows to identify its effectiveness or whether these laws have produced significant changes in the inclusive national educational setting of students with ASD. Thus, the aim of this study was to characterize the enrollments contained in the microdata of the School Census of students with ASD in the common classroom in the five Brazilian regions, considering the possible impacts of ^{Law No. 12.764/2012}

2 Method

This study consisted of a comparative research, in which it was sought to know similarities and to explain divergences (^{Marconi & Lakatos, 2003}), of documental character, that aimed to analyze documents with the objective of collecting facts (^{Caulley, 1983}), which may have as its source: a) public archives; b) private archives; and c) statistical sources (^{Marconi & Lakatos, 2003}).

3 Results and discussions

The results and the discussions were divided by studied variable. The first variable consisted of the gross total of students with ASD and other disabilities (indicated as DDEF in Table 1) enrolled in regular education and the percentage of students with ASD in relation to the total enrollments of these two populations added. When considering the population rate of each region as a whole, according to the demographic census of 2010, between the ages of 0 and 19, the Southeast region led (38.2%), followed by the Northeast region (30.6%), South (13.2%), North (10.5%) and Central-West (7.5%) (^{Instituto Brasileiro de Geografia e Estatística [IBGE], 2010}). In relation to these percentages, it can be seen in Table 1 that the Southeast region had the highest enrollment concentration of students with ASD and other disabilities, followed by the Northeast and South regions The North region, with regard to the concentration of enrollments of students with disabilities, remained, with the exception of 2009, in fourth place, followed by the Center-West region. For the students with ASD, there was a greater oscillation between these two regions, with the Center-West region, although with a small gross difference in relation to the North region, with the fourth largest concentration of enrollments in 2009, 2010, 2015 and 2016 . However, when considering the data from the 2010 IBGE statistical summaries about the population between 0 and 19 years old, it was observed that the South region had 6,595,166 people, while the Central-West region had 4,706,098 (^{IBGE, 2010}). Thus, the difference in enrollments of students with ASD in this second region in relation to the North region was tangible.

Regarding the percentage of enrollments of students with ASD, considering the total number of students with disabilities, the Southeast region began the period studied leading in 2009 (18.7%) and in 2010 (13.5%). This percentage fell in the following years, occupying the third place, together with the Northeast region in 2011 (10.5%), fourth place in 2012 and 2013 (10% in both years), returning to the highest percentage in 2014, remaining in this way until the end of the studied period; however, with a smaller percentage difference in relation to the other regions. These data were similar to the ones found by ^{Castro (2015)}, who observed that the region with the highest percentage of enrollments of TPSE students in relation to total enrollment in Basic Education by region was the Central-West region. In addition, the author also found that this region suffered the highest percentage increase in the period studied (216%), followed by the North (116%), the South (109%) and the Northeast (98%) regions, while the Southeast region, despite having the largest gross enrollment of TPSE students, increased by only 31.8%. Thus, it can be inferred that there was a greater growth in the percentage of enrollments of students with ASD in the period studied in all regions, except for the Southeast region, which declined and grew again in 2014. This growth may signal the impact of ^{Law No. 12.764/2012}.

In order to obtain more accurate data regarding the possible impacts of this law on students' enrollments, Table 2 was developed to bring the percentage variation of enrollments of students with ASD and those with other disabilities between 2009 and 2012 and between 2012 and 2016.

As it can be seen in Table 2, there was a large difference between the percentage change in enrollment of students identified with ASD during the first period studied in relation to the other students with disabilities. In the first period analyzed (2009-2012), it was observed that, with the exception of the Southeast region, the growth of enrollments of students with ASD was higher than the growth of students with the other disabilities, being more than double in the North and Central-West regions, considerably higher in the Northeast and similar in the South. The Southeast region consisted of the one with the greatest difference in these data, since the enrollment of students with ASD fell between 2009 and 2012, and enrollments of other students with disabilities showed a considerably lower growth in relation to the other regions.

These data were similar to those presented by ^{Castro (2015)}, although the variation of the studied period and the grouping of the data modified the order of greater percentage variation in relation to this study. The second period studied showed that both enrollment groups continued to grow, however in a lower form from 2009 to 2012, except for the Southeast region, which demonstrated a great change in the enrollment of students with ASD between 2012 and 2016. It was still possible to observe that, with the exception of the Central-West region, growth was quite similar. In the other regions, the enrollment growth of students with ASD was considerably higher than that of the other disabilities (the group whose percentage enrollment variation was close to 50% in all places studied) - in the South region, for example, between 2009 and 2012, growth was similar for both populations (105% for ASD and 104% for DDEF), while between 2012 and 2016 the growth was 111% for students with ASD and only 49% for other disabilities. This difference in percentage growth may show that there was an impact of ^{Law No. 12.764 / 2012} on the enrollment of students with ASD throughout the country, except for the Central-West region, which, from the beginning of the period studied, as demonstrated in ^{Castro's (2015)} study already presented a percentage growth superior to the others, which may have reflected in a lower growth between 2012 and 2016.

Table 3 shows the percentage of enrollment in the regular education of students with ASD separated by gender and also the average of this percentage by region. It can be observed that there was little variation in the percentage by region, initially maintaining a ratio of about 30% of female gender enrollments and 70% of male. This proportion increased over the years, with the most distinguishing regions being the Southeast and South, so that the increase in male enrollment was greater in relation to the others, reaching close to 80% of enrollments in these places in 2016, while the others maintained a percentage of about 75% of male students. It should be noted that, as can be seen in Table 3, in all regions studied, growth occurred as of 2012, year in which ^{Law No 12.764/2012} was promulgated. This fact, when confronted with the literature, corroborated in the sense of identifying impacts of this Law.

When the proportion of students with disabilities by gender was discussed according to national census studies, it was observed that there was a prevalence of approximately 60% of males and 40% of females. ^{Rebelo (2012)} identified this proportion in all Brazilian states, thus emphasizing the importance of the issue of gender with regard to students with disabilities. This was discussed by ^{Mendes and Lourenço (2009)}, who pointed out that there was a prevalence of boys in relation to girls (from 60% to 40%), as shown in the census studies on the subject. In contrast, according to a report by the US Centers for Disease Control and Prevention (CDC) about the prevalence and characteristics of eight-year-old children diagnosed with ASD in different states of the United States in 2010 and 2012, the prevalence was 4.5 and 4.4 of boys for one girl, respectively. In percentage terms, approximately 82% and 81% (^{Christensen, Baio, & Braun, 2016}). This data backs up the study conducted by ^{Lima and Laplane (2016)}, which showed that, between 2009 and 2012, in the city of Atibaia (São Paulo - Brazil), enrollment data of students with ASD presented a prevalence four times higher in male students than in female students.

Therefore, considering the national and international literature and the data contained in Table 3, it can be inferred that the prevalence by gender of students with ASD has changed over the years in the country, moving towards the proportion expected by the literature in relation to the distribution of this population according to the gender, starting to grow from 2012, strengthening the hypothesis that ^{Law No. 12.764/2012} would be positively impacting the number of enrollments of students with ASD in the regular network of education.

Table 4 shows the distribution of enrollments of students with ASD by aggregated educational teaching. As can be seen in Table 4, when comparing the enrollments of students with ASD in relation to the educational stages, it was observed that the data of the initial grades of Elementary School were disproportionately higher in relation to the other stages, showing the concentration of enrollments in this educational stage during the period studied. This information was consistent with the literature of the area, which demonstrated the same concentration both in students with disabilities in general in different historical cutoffs after 2008 and also in different locations (^{Carvalho, 2012}; ^{Laplane, 2014}; ^{Martins, 2012}; ^{Meletti & Bueno, 2011}). The same occurred with the studies of TPSE students in general (^{Meletti & Ribeiro, 2014}; ^{Nascimento, 2014}), with a concentration of 74% of TPSE students enrolled in 2012 throughout the country in the initial grades of Elementary School (^{Meletti & Ribeiro, 2014}). Similar data was found by ^{Françozo (2014)}, whose target population consisted of students with visual impairment. Regarding ^{Law No. 12.764/2012}, it did not seem to impact the distribution of student enrollments with ASD in the period studied.

In addition, the data presented in Table 4 appeared to be inconsistent, that is, if it is compared to the percentage variation of 2016 in relation to 2009 by region of the country, it was observed that, in the Central-West, Southeast and South regions, there was a growth in Early Childhood Education (56%, 122% and 62%, respectively), and decrease in the Northeast and North regions (-13% in both). In the first years of Elementary School, there was a decrease in all regions, which were: -16%, -14%, -11%, -10% and -15% for the Central-West, Northeast, North, Southeast and South regions, respectively. In the final grades of Elementary School, there was an inverse movement, that is, an increase in the Central-West, Northeast and North regions (10%, 95% and 113%, respectively) and a decrease in the Southeast and South regions (-99% in both regions). Finally, in the High School, there was growth in all regions above 100% for the Northeast (173%), Southeast (136%) and South (355) regions, while the North Region had an increase of 33% and the Central-West a drop of -98%.

These inconsistent data, that is, abrupt increases followed by fall and vice versa, corroborated those of ^{Lima and Laplane (2016)} and ^{Talarico and Laplane (2016)}. The authors, who studied the school trajectory of student enrollment with ASD in municipal contexts, observed that only 12.92% of the students, in Talarico and Laplane's (2016) study, and 6.38%, in ^{Lima and Laplane's (2016)}, between 2009 and 2012, had full schooling. Both studies - ^{Talarico and Laplane (2016)} and ^{Lima and Laplane (2016)} - verified a high number of enrollments characterized as partial trajectories without school retention (40% and 25.53%, respectively), that is, those student enrollments appeared, disappeared and reappeared in the period studied without any student retention in any of the years. This data, associated with the low index of full trajectories, signaled the high rate of school dropout.

Reports of interviews with parents of students with ASD in ^{Talarico and Lima (2016)} showed concern about the lack of progress in their children's learning, with speeches indicating that students were not learning in the classroom, that many complete lessons were made by other people and that the parents themselves had to teach at home. The authors suggested that the lack of support was a barrier to the schooling of this population, given the low percentage of students with ASD enrolled in the SES in the municipality investigated, as well as with the reports of the relatives of these students, who presented to the author that despite of the SES in the municipality, their children did not receive assistance (^{Talarico & Lima, 2016}).

In order to verify the incidence of students with ASD enrolled in the SES in the country, Table 5 was elaborated. As it can be seen in Table 5, with the exception of the Southeast region, in 2009, the percentage of enrollments of students with ASD in the SES was lower by region. And in all regions, there was a gradual increase over the period studied, so that, in 2016, this percentage was approximately 30%, again with the exception of the Southeast region, whose percentage was 25.5%. The percentage variation in the period studied did not allow to infer that enrollments in the SES suffered impacts related to ^{Law No. 12.764 / 2012}, thus requiring other study methodologies or waiting for the census data that will emerge in the coming years to verify a possible existence of impact in this regard.

According to ^{Meletti (2014)}, in 2014, most of the TPSE students did not receive SES in the country. ^{Rebelo (2012)}, when studying the data from 2009, observed that the Southeast and South regions made up about 68% of the total enrollment of TPSE students in the SES, and only about 40% of the total enrollment of this public received the attendance and the majority of those students had intellectual disabilities. This low rate and the main target population, when considering TPSE as a whole that year, corroborated with the data presented in this paper about students with ASD in 2009. In addition, the author also verified that there was a direct relationship between the Human Development Index (HDI) and the percentage of enrollments in the SES, that is, the four states that had the lowest HDI also had the lowest number of students enrolled in the SES (Roraima, Sergipe, Amapá and Amazonas).

In relation to the students with ASD in the municipal contexts whose census data have already been studied, it was observed that there was a low enrollment rate of this population in the SES. In the municipality of Campinas, between 2009 and 2012, the percentage of students with ASD enrolled in the SES was 13.9% in 2009, 7.48% in 2010, 5.31% in 2011 and 7.12% in 2012. Thus it was observed that there was a reduction of enrollments of this public in the SES (^{Talarico & Laplane, 2016}). ^{Lima and Laplane (2016)}, similarly to the Atibaia municipality, concluded in their study that a small number of students with ASD attended the SES and that the state network did not provide any extra support for this population, which may even be a factor that may lead to school evasion, previously discussed, since the students carry out their schooling in the municipal network until Elementary School and, when they go to the next stage of education, they do not find support, which leads to school abandonment due to the lack of pedagogical support. Of the 96 students with ASD whose trajectories were studied in Atibaia, only six completed the educational stages of Basic Education, indicating the great difficulty in school permanence. Both municipalities mentioned above are located in the state of São Paulo, confirming the data in relation to the Southeast region, presented in this paper, as the one in which, despite having a higher enrollment rate in 2009 in the SES, suffered the lowest percentage growth both in the total number of students with ASD in the common classroom and in the SES.

It is also worth noting that, as punctuated by ^{Meletti (2014, pp. 806-807)}:

Another aspect is the SES being officially the realization of special education in the regular school through work that is not characterized as parallel to regular education. It is worth questioning what kind of support students with special educational needs, who do not receive SES, are having in regular school.

This problem has a direct impact on the learning of students with ASD, since they require intervention and individualized, multidimensional and multidisciplinary support, aiming to maximize functional independence and quality of life throughout development and learning, improving social skills and communication, as well as reducing difficulties and promoting independence and providing support to families (^{Lai, Lombardo, & Baron-Cohen, 2014}). In this way, the SES, even if it had a high enrollment concentration, would not fit the appropriate intervention and support methodology for this public.

It is worth emphasizing that the current policy of school inclusion consists of a policy whose only way of support focuses on a service that occurs in the school extra hours and without providing support in the common classroom, associated with an education that has one of the worst performances in the world, does not show enough to meet the very distinct needs of the target population for which it is intended (^{Mendes et al., 2015}), pointing to the need to reformulate this policy.

4 Final considerations

Based on the data presented in this study, we believe that it was possible to achieve the objective of characterizing the enrollments in the micro-data of the School Census of students with ASD in the common classroom in the five regions of the country, considering possible impacts of ^{Law No. 12.764/2012}. This characterization was essential to understand more about the students with ASD enrolled in the common school and to look for studies that had the focus to intervene so that this public remains in the school and, in fact, learns.

It was observed that these enrollments suffered a growth superior to those of the other students with disability during the period studied, with the Southeast region having the highest gross enrollment concentration of students with ASD and with other disabilities, followed by the Northeast and South regions. However, the Southeast region was the one that had the lowest percentage growth, specifically between 2009 and 2012, while the other regions showed a growth of more than 100% in the enrollments of students with ASD. This situation was reversed between 2012 and 2016, a period in which the Southeast region showed a growth of 120% in enrollment, leading the percentage of this period. In relation to the enrollment of other students with disabilities, in both periods, the Southeast region maintained its lowest percentage growth when compared to the others (except in the Northeast region between 2012 and 2015, whose percentage growth was only 1% lower than the Southeast).

Regarding the gender of the students, it was observed that over the years the proportion of boys and girls with ASD has been approaching that expected by the literature, that is, about one girl for approximately 4.5 boys.

In both of the aforementioned data, it can be inferred that there was a possible impact of ^{Law No. 12.764/2012}, since they had a sharper percentage variation after 2012. However, with regard to school attendance of this population, it was not possible to verify impacts of this law, considering the percentages studied. In relation, for example, to the distribution of enrollments by stage of education, it was noted their concentration in the initial grades of Elementary School, as happened with the other TPSE students. Another factor raised was the inconsistency in relation to the percentage variation and the distribution of enrollments by stage of education. More specifically, while enrollments grew, in all regions of the country, in Early Childhood Education and in the initial grades of Elementary School, there was a fall in enrollments referring to the final grades of Elementary School and High School. Such information indicated a high rate of school dropout. It should also be pointed out that, once again, the Southeast region was the one whose percentages differed from the others, pointing to a lower number of access to the common classroom for students with ASD, corroborating with data previously discussed.

Once the school dropout was verified, the SES data were analyzed, from which it was possible to verify, in agreement with the literature, that only a small percentage of this population was attended, so that the lack of support demonstrated be the reason why there is high school dropout rate. In addition, the literature showed that the support promoted in the SES was not sufficient to provide the special needs of students with ASD. Therefore, it was inferred that ^{Law No. 12.764/2012} had an impact on access to the common school for students with ASD, but this access was not enough to guarantee their permanence. This perspective reflected not only the data of the students with ASD, but of the TPSE in general, since there is a policy whose only way of support is parallel to the common classroom, which has been shown not to be sufficient to meet the demands of the student of Special Education.

The limitation of this study refers to the inconsistencies in the census data, as well as the fact that ^{Law Nº 12.764/2012} was implemented in the middle of the historical period of the eight years studied, which may not be enough to verify its real impacts. Thus, studies are required that have other complementary forms of data collection or even a longer historical period after the implementation of the law.