FORMAS DE ENSUCIAMIENTO

“School choice” is defined as policies which allow parental preferences to influence which school their child will attend. The central problem in designing a school choice system is in allocating pupils to schools, accounting for both parental preferences, school capacities and wider policy objectives (e.g. diversity in schools). One motivation for school choice focuses on the importance of respecting parental preferences, and allowing them to choose from a diversity of school types. A second motivation, not necessarily implied by the first, is to induce competition among schools with the aim of improving school standards. Both of these motivations have been referenced by different UK Governments in relation to school choice policy.

The rationale behind using school choice to enhance competition, is that by allowing families to choose which school they attend, popular schools will flourish, and poor per- formers will be forced out of this “quasi-market”. A number of conditions are required for this system to work, including: (i) parents engage with this process, they can observe the

Human capital formation 2.3. SCHOOL CHOICE

quality of different schools, and they choose schools based on effectiveness; (ii) there is a choice of schools available to parents; (iii) good schools can expand to meet increased demand; (iv) and poor performing schools face consequences (they are shut down or pro- vided with remedial assistance).

In England, features of parental “choice” and a “quasi-market” in compulsory school- ing were introduced largely from 1988, formalising the role of parental preferences, award- ing extra funding for popular schools, and allowing school some discretion in the pupils they admitted. The specific operation and political motivation of the school choice setting in England has evolved substantially over the last few decades. Very generally, the pro- cess to allocate places works as follows: to apply for a place at a state school in England, parents submit a ranking of their preferred schools to the local government. They can list between 3 and 6 places depending on their local area.

For oversubscribed schools, allocation of places is prioritised based on a set of ob- servable criteria (“school priorities”), typically including: whether the child has exception “medical or social needs”; whether the child has an older sibling at the school; and, fi- nally, the distance from the school as the tie-breaker. In some cases, faith may also be considered. An algorithm is used to allocate school places, given parental preferences, school priorities and school capacities (for example, a Deferred Acceptance algorithm is used). If a pupil cannot be allocated to any of their listed preferred schools (often because the parents do not list enough schools), they are assigned to a school with spare capacity (which is typically a less desirable school).

The LSYPE cohort were applying to schools in 2000, to start in September 2001, when the admissions arrangements were less regulated, and subject to less stringent legislation, compared with today. The following section describes in detail the policy settings appli- cable for the LSYPE cohort.

Admissions and oversubscription criteria used in September 2000

The admission process varied both within and between local authorities, and in Septem- ber 2001 was more heterogeneous and under-developed compared with the present day ar-

2.3. SCHOOL CHOICE CHAPTER 2. SECONDARY SCHOOL CHOICE

rangements. At its most general level, the admissions process had three stages (i) parents submit their preference(s) to the admission authority (or authorities), (ii) the admission authority tries to assign families to their preferred schools, and (iii) in the case of over- subscription, applications are ranked based on a set of observable criteria to ration scarce places. The School Code in force at the time provided some suggestions for the specific cri- teria to be used: “Commonly used and acceptable criteria include sibling links, distance

from the school, ease of access by public transport, medical or social grounds, catchment

areas and transfer from named feeder primary schools, as well as parents’ ranking of pref-

erence”(DfE(2001), para. 3.14). In practice, the actual procedures followed vary and are difficult to categorise (seeWest et al. (2004);Williams et al. (2001) for surveys of the admission protocol used, and specific examples from LEA brochures).

Once the families’ preference rankings have been submitted to an admissions authority, a procedure is required to allocate school places, given school capacities and priorities. There are a number of approaches for doing this. The two used in 2000 were: the FPF

first-preference algorithm, also called the Boston or immediate-acceptance mechanism;

and theequal-preferencealgorithm, also called the Gale-Shapley, or deferred-acceptance

(DA), mechanism.

The FPF approach proceeds in rounds. First, each school starts by assigning places to pupilswho put that school as their first choice, based on the ranking determined by the

school’s admission criteria—until either there are no places left, or all students who put that school as first choice are assigned. The next round conducts the same procedure for pupils who put that school as their second-choice, and so on. This approach maximises the share of families getting their first-choice school. The drawback of this approach is that families who are not offered a place at their first preference school can also be rejected by their second preference school in favour of a child of parents who placed that school first, despite living further away. If families put their true preferred school as their first choice, but where they have a very low chance of acceptance, they may not only fail to secure a place at their preferred school, but also fail to be placed at any school they deem acceptable. This means that families have an incentive to trade-off their true preferences

Human capital formation 2.3. SCHOOL CHOICE

against their probability of acceptance, in order to increase the chance of gaining a place at a satisfactory, but not most preferred, school. 41% of schools used a ‘first-preference’ approach to offer places for the September 2001 intake (Westet al.,2004).

In contrast, the DA approach does not use the preference ranking as a method for schools to rank applicants. In this method, parents submit their preference ranking over schools. The families’ application is considered at each school that they listed as a prefer- ence without reference to its ranking. Places are offered to applicants who fulfil the over- subscription criteria of a school to the greatest extent. This may lead to a parent whose child does not meet the criteria of his first preference school being offered a place at a nearby school for which he has named a second preference. Where proximity is a key cri- terion, one effect of this system is to confer advantage to parents who live near to several popular schools.

Westet al. (2004) used survey evidence to explore the use of various admission criteria and identified older siblings at the school (96%), distance from home to the school (86%), medical/social need (73%), catchment area (61%) and ‘first preference’ (41%) as the most common admission criteria practices (see Appendix 2.E, Table2.20for a longer list). In 2000, it was not required to give “looked-after” children first priority. From 2006, it was a statutory requirement that children in care should be given top priority in the event of a school being oversubscribed, and in 2008, almost all schools (99%) had an admissions criterion relating to children in care compared with 2% in 2001 (Westet al.,2011).

As described further in Appendix2.E, there are a range of sources of uncertainty about whether one will be admitted to a particular school. Catchment area policies, while be- coming less popular coming up to 2001, were used in both general admissions (14.9%)

and featured somewhere as an oversubscription criteria in (63.1%) of LEAs Williams

et al. (2001). Catchment area policies create uncertainty because the catchment areas can change over time in unpredictable ways. For example, in the survey byWilliamset al.

(2001), one LEA deployed a flexible catchment area system based on the numbers of

children applying to particular secondary schools from particular primary schools over a period of time. Additionally, sibling and catchment criteria can interact; for example, sib-