ECONOMIC ANALYSIS GROUP
EAG Discussion Papers are the primary vehicle used to disseminate research from economists in the Economic Analysis Group (EAG) of the Antitrust Division. These papers are intended to inform interested individuals and institutions of EAG's research program and to stimulate comment and criticism on economic issues related to antitrust policy and regulation. The analysis and conclusions expressed herein are solely those of the authors and do not represent the views of the United States Department of Justice.
Information on the EAG research program and discussion paper series may be obtained from Russell Pittman, Director of Economic Research, Economic Analysis Group, Antitrust Division, U.S. Department of Justice, BICN 10-000, Washington, DC 20530, or by e-mail at email@example.com. Comments on specific papers may be addressed directly to the authors at the same mailing address or at their e-mail address.
Recent EAG Discussion Paper titles are listed at the end of this paper. To obtain a complete list of titles or to request single copies of individual papers, please write to Janet Ficco at the above mailing address or at firstname.lastname@example.org or call (202) 307-3779. Beginning with papers issued in 1999, copies of individual papers are also available from the Social Science Research Network at www.ssrn.com.
40 states and the District of Columbia have laws that authorize public charter schools to provide public education with public funds, in competition with regular public schools. However, many of these laws contain provisions that have hardly any rationale except to restrain competition, especially: (a) explicit caps on the number of charter schools, (b) exclusive reliance on local school districts as authorizers of charter schools, and (c) impediments to multi-school charter systems. Removal of these anticompetitive constraints should have both short- and long-term benefits. However, the evidence to date suggests that the principal benefits are likely to be long-term and evolutionary, as the result of competitive pressures that encourage good schools to grow, bad schools to exit, and new schools to imitate good schools rather than bad.
Public charter schools are schools authorized to provide public education with public funds, in competition with regular public schools. They are required to comply with some but not all regulations for public schools in their state. (For example, they do not teach religion or charge tuition, and they must comply with civil rights and health/safety regulations. In many states, charter schools may not select their students except through a neutral lottery.) They operate under a limited-term charter (usually 3-5 years) that may be renewed, non-renewed, or even revoked before the term ends.
Vouchers are another form of school choice, more controversial than charters but growing in public acceptance. The principal differences are that vouchers may be used at private and religious schools, voucher schools do not face oversight by a chartering authority, and vouchers generally provide lower funding than charters (which are themselves generally underfunded compared to regular public schools).
Public charter schools first appeared 15 years ago, and currently serve about 2% of all students nationwide, with the highest share among poor, minority students in the urban inner cities. Enrollment is growing 10-15% per year (compared to 0.3% per year for regular public schools). In 2002-03, 39% of charter schools reported having a waiting list, averaging 135 students each (nearly half their total enrollment).
Some states require applicants to obtain their charter from the school district with which they will compete, while others offer alternative authorizers such as universities, the state school board, or a specially created state agency. Some state laws impose explicit caps of various types e.g., caps on the number of schools that may be chartered statewide or in specific cities, the number that may be chartered by particular authorizers, the number of new start-ups per year, etc. Many state laws either prohibit charter operators from operating more than a single school, or impose constraints that impede the development of multi-school charter systems.
Those three provisions explicit caps, exclusive reliance on local school districts as authorizers, and impediments to multi-school charter systems have hardly any rationale except to restrain competition. They were initially included as a necessary compromise with the opponents of public charter schools. However, support for charter schools has grown steadily along with their enrollment, so the trend in charter laws is toward relaxation of these anticompetitive constraints.
Removal of these anticompetitive constraints should have both short- and long-term benefits. However, the evidence to date suggests that the principal benefits are likely to be long-term and evolutionary. To be sure, the current customers of charter schools perceive immediate benefits safer schools, a different instructional emphasis but these "product differentiation" benefits are not well documented, and the most careful comparisons of "quality" i.e., academic gains in reading and math show that on average, public charter schools currently perform about as well as the regular public schools.
So the principal benefits are likely to be incremental, evolutionary, and long-term. They will be the result of competitive pressures that encourage good schools to grow, bad schools to exit, and new schools to imitate good schools rather than bad. Some regular school systems may adapt while others disappear, and the current wave of "mom and pop" charter schools may itself be replaced by nationwide chains.(1) This process is just beginning, so there is very little evidence to indicate how this will play out.
Some state legislators want reassurance that charter school competition will not impose short-term harms. To provide that reassurance, they look to charter school authorizers to weed out the bad schools by denying or revoking their charters. By and large, the authorizers have not done this (for good reasons discussed below), except for financial misconduct.(2) Instead, their principal focus has been to assist struggling entrants, to require charter schools to adopt clear goals and measures of progress, and occasionally to force management or other changes in failing schools. Those measures facilitate competition rather than supplanting it.
Section II of this paper describes the most prominent anticompetitive restraints in current state laws. Section III presents theory and evidence of static, short-term competitive effects from public charter schools, and Section IV discusses evolutionary, long-term effects. Section V explains what charter school authorizers do. Section VI presents recommendations.
II. Anticompetitive restraints on supply
Explicit caps. Four heavily-populated states New York, Illinois, Massachusetts, and Ohio are unable to open new charter schools because of explicit caps imposed by their state charter laws. This poses major obstacles to planned structural reforms in New York City and Chicago. In Michigan, no more schools may be chartered by state universities, which have authorized the bulk of the charter schools in that state. Explicit caps are also constraining charter school activity in Hawaii, Iowa, and North Carolina.(3)
New York City is an especially interesting case, because the Chancellor of the New York City Public Schools is an expert in antitrust as well as education, former AAG Joel Klein. "From the day I arrived as Chancellor I made clear that charters are a critical leveraging force in public school reform,"(4) he said. "I want parents to say, 'Look, we could lose people to charter schools if our school doesn't improve.' Competition in this thing works."(5) Klein's ambitions are stymied by the caps in the New York State charter law:
The New York State Board of Regents cannot approve any additional schools recommended by the Chancellor and the State Education Department due to a legislative cap, which Chancellor Klein today yet again urged the State to remove and which limits the number of charter schools that may be created in New York State. As a result, two high-quality schools recommended by the Chancellor and considered by the Board of Regents were not approved today, denying the children of New York City much-needed educational options.(6)
Exclusive reliance on local school districts as authorizers. Ten states have no charter law, and therefore no charter schools. Of the 41 states with charter laws, roughly 10 designate local districts as the sole authorizer, another 10 allow appeals to the state school board, and 20 offer alternative authorizers without having to go through an appeals process. Typically, an appeals process does little to ameliorate exclusive reliance on local school districts, because the result of a successful appeal is to place the charter school under the oversight of a negligent or hostile authorizer.
For those states that offer alternative authorizers, the most popular choice is the state school board (about 15 states). Other choices are universities or community colleges (5 states), a state charter board (5 states), a county or regional district (3 states), city council or mayor's office (2 states), or a nonprofit organization (1 state).(7)
Two econometric studies found that offering alternative authorizers has a large impact on the number of charter schools.(8) It appears from their results that adding one class of alternative authorizers (roughly a one-standard deviation change in their authorizing variable) increases the number of charter schools by about 20-25%.
Impediments to multi-school charter systems.(9) State laws commonly include provisions that discourage successful charter schools from replicating their success. Typically, the charter confers authority to operate only a single school. Some states allow charter holders to operate more than a single school, but require each school to have its own governing board.
In addition, scaling up from one school to many requires investment funds, which are most readily obtained from investors seeking a profit but only a few states will grant charters to for-profit companies. Some states even prohibit charter holders from contracting with for-profit companies to obtain comprehensive management services.
The prejudice against for-profits greatly limits the ability of charter schools to achieve scale economies even indirectly, through management organizations that have sprung up to support charter schools. These management organizations include both nonprofits and for-profits, but only the for-profits have achieved national scope. Indeed, only one nonprofit (the national chain of KIPP schools) operates in more than one state.
The prejudice against for-profits is unwarranted by their performance. Charter schools managed by for-profit organizations serve a far higher percentage of poor and minority students than the national average of these students (as do nonprofits).(10) Moreover, when one compares their performance on state assessments with the average statewide performance, "the average gains for the for-profit managers relative to state gains round to 5, 6, and 8 percentage points for one-, two-, and three-year intervals."(11) (emphasis added)
III. Static, short-term competitive effects of public charter schools
Static competitive effects are the effects that show up in a typical antitrust investigation. Firm capabilities are assumed fixed, and the investigation looks for consumer benefits in the form of price, quality, and variety as consumers shift suppliers and suppliers change their behavior.
Analytically, one usually examines these effects in the context of an equilibrium model, in contrast to the evolutionary analysis in Section IV.(12) In general, static equilibrium analysis may be useful for long-term effects as well as short-term, and evolutionary change may occur short-term as well as long-term; however, I will be using the static equilibrium analysis mainly for short-term effects, and the evolutionary analysis for long-term effects.
Consumer choice in K-12 education is similar to other markets that the antitrust authorities commonly deal with. The primary idiosyncracy is the same as in healthcare i.e., that most of the expenses are covered by a third-party payer. That is a lesser difficulty in education than in healthcare, because education consumers typically have little ability to demand additional costly services for themselves (with the possible exception of special education a complication I will not deal with in this paper). In all other respects, consumer choice in K-12 education is unexceptional.(13)
Public charter schools are for-profit or nonprofit firms whose profit opportunities are determined by marketplace competition for students, public funding on a per-student basis, and (for nonprofits) tax-deductible contributions. Regular public schools are nonprofit firms whose profit opportunities are determined mainly by voters acting through elected officials.(14)
"Nonprofit" doesn't mean a firm doesn't make profits. Nonprofit firms do have revenues and costs, and therefore earn profits but they are required to consume those profits internally rather than distributing them to investors. Government-owned school systems are no different than nonprofit charter schools in that regard. Their profit calculus differs only in the source and amount of funding. Government-owned public schools obtain their funding almost entirely from government funds, while nonprofit charter public schools obtain revenues from donors as well. Also, public charter schools are typically underfunded relative to regular public schools. They often receive public funds for operating expenses but not facilities, or they receive the state contribution to school funding but not the local contribution.(15)
For-profit and nonprofit charter schools have different competitive strengths and incentives. In states where they are permitted, for-profit charter schools enjoy a competitive edge in operating efficiency and access to investor funding. This creates strong incentives for growth, and it improves their survivability in states where charter funding is especially meager. For nonprofits, the obligation to consume profits internally is an inducement for entrepreneurs to follow their vision instead of maximizing profits. This effect is inconsequential if meager funding and intense competition severely restrict profits; however, in the near term, with long waiting lists and generous donors, this constraint may foster more product variety among nonprofit charter schools than for-profits.
Regular public schools have fatter budgets than public charter schools, and they experience a severe principal-agent problem i.e., it is difficult for voters, relying on newspapers for information and elected officials for action, to exercise control over public school decision-makers. This invites rent-seeking by teacher unions (the most vigorous opponents of public charter schools) and by superintendents. Rent-seeking by superintendents takes the form of policy choices that make them look good but don't help children learn. The two most costly examples are (a) frequent disruptive changes that create the appearance of aggressive reform(16) and (b) deceptive testing (teaching narrowly to the test, using tests that don't require any knowledge of the subject tested except the ability to read the test questions, and changing tests to eliminate long-term trends and to create spurious gains as teachers adapt to a new test).(17)
These incentives suggest at least two reasons that the regular public schools might want to respond to competition from public charter schools: they may look bad if the charter school appears to be more effective, and they may lose rents if the marginal revenue from lost students exceeds the marginal cost.(18) If they do choose to respond, they have a range of options to consider e.g., self-promotion, assigning better principals to the schools that are losing students (this is a popular and effective quick-fix for troubled schools), etc. Some of these strategies improve student learning, some don't, and some benefit some schools at the expense of others.
In summary, the different incentives of regular and charter public schools suggest that competition from charter schools will yield improved quality and variety in two ways: consumers will switch to preferred schools, and regular public schools will respond to the new competition.
The most-studied static competitive effect is the impact on students who choose charter schools: do these students learn more or less than they would have learned in a regular public school?
That's a difficult question to answer, because the students who choose public charter schools may be quite different than students who remain in the regular public schools. (Demographically, public charter schools closely resemble the regular public schools nearby, but their students could also differ in motivation, prior school success, etc.) The most highly regarded studies control for this by looking at the annual gains of children who switch from regular to charter public schools, or vice versa.
Unfortunately, that significantly limits the scope of the conclusions that can be drawn, for several reasons: (1) students who switch may be quite different than students who do not; (2) very few states have a longitudinal database of individual student test data for both charter and regular public schools, and those states may be quite different from each other and from other states; and (3) this data is only available for reading and mathematics as measured by state tests, so the studies are only able to evaluate the effectiveness of public charter schools in what has become the core learning agenda of the regular public schools.
The general conclusion of this handful of highly regarded studies is that on average, after a few years of experience, public charter schools perform about as well as regular schools as measured by annual gains in state test scores in reading and mathematics:
Frustratingly, regardless of the methods used, the results are mixed, some positive about charters and some negative, with null or mixed findings the most common. One additional fact is that whether studies draw positive or negative conclusions about charter school effectiveness, the differences are not strong....The five most sophisticated studies focus on the four states where especially good data on student achievement are becoming available (i.e., Arizona, Texas, Florida, and North Carolina)....Of these studies, two are positive about charter school effects, two report mixed results, and one finds in the negative.(19)
Another important question is the impact on students who remain in the regular public schools: does competition from public charter schools benefit these students, or not?
Again, the evidence is mixed. In Michigan, Hoxby (2003) found positive effects, while Benninger (1999) did not. In North Carolina, Holmes et al (2003, 2006) found positive effects, while Bifulco and Ladd (2006) did not.(20)
Booker et al (2005) found positive effects in Texas, with the interesting wrinkle that in districts facing competition from public charter schools, student performance improved at low-achieving schools and fell at high-achieving schools. "These results raise an interesting question," they say. "Does charter competition lead districts to reallocate resources to poorly performing campuses and away from highly performing campuses?"(21) If so, I suspect the mechanism is probably by reassigning their best principals (who know how to get rid of bad teachers and attract good ones).
IV. Evolutionary long-term competitive effects of public charter schools
Evolutionary competitive effects arise from three sources, which act in concert to create positive change:
Increased variety comes mainly from new entrants, rather than from incumbents testing varied approaches to education. Selection pressures are weak at the present time, because excess demand ensures that very few charter schools fail. Replication and development of the preferred varieties is most likely to emerge through coaching by authorizers and the proliferation of growth-oriented chains like KIPP; however, there are also indications of an emerging industry infrastructure to provide public charter schools with technical assistance, back office and management support, etc.
Removing the restraints on supply should markedly increase evolutionary competitive effects, because often the closest competitor to a public charter school is another charter school.(22)
It's important to note that the evolutionary benefits of competitive entry do not rely on entrants being generally more efficient or more innovative than incumbents and, indeed, generally they are not. Metcalfe (2005), summarizing the results of a Canadian study of manufacturing plants over the period 1970 to 1989, says: "Greenfield entrants start with a disadvantage in respect of...productivity and profitability and it takes 8/9 years for them to improve their performance to approximate to the industry average."(23) As in nature, all that is required is that some of the challengers have a competitive edge.
Moreover, Metcalfe (2005) adds, "it is behaviour that is far from the norm that causes the most significant changes in the population structure."(24) That is, for evolutionary effects, the most important comparison between charter and regular public schools is not how they compare "on average" (as discussed above under static effects), but rather how they compare at the upper levels of performance. For these "upper tail" schools, annual gains in reading and math scores are significantly higher in public charter schools than in regular public schools. Hanushek et al (2005) provides a chart for Texas schools, depicting the school average achievement gain after controlling for a number of student demographic characteristics:
Distribution of Estimated Quality for Charter and Regular Public Schools in Texas(25)
Another requirement for positive evolutionary pressures for change is that parents should prefer the more effective schools, so that students switch out of schools in the lower tail and into schools in the upper tail. They do:
Even though parents undoubtedly have a variety of motivations for choosing individual charter schools, most are likely to be sensitive to the narrow question of quality in basic skills. The results show that the probability of exiting a charter school declines with school quality, although the relationship is weaker for lower income students. The quality responsiveness of families satisfies a necessary condition for the education market to favor higher quality charter schools over time....(26)
However, the vast majority of public charter schools are survivors for now. Gau's (2006) survey found that only 10 percent had closed voluntarily (presumably due to financial or organizational problems), 3 percent had their charters revoked before the end of the term (1 percent for low student achievement, 2 percent for other reasons), and 1 percent were non-renewed (mostly for low student achievement).(27) That survey also reported that authorizers are becoming choosier about approving new charter schools after January 2003, the approval rate dropped from 68 percent to just over half.(28)
Eventually, this should shift the quality distribution of charter schools, with fewer schools in the lower tail and more in the upper tail. Currently, there is good evidence that start-up charter schools improve their performance in their early years, but I have not seen evidence that more recent cohorts of entrants are superior to previous cohorts.
V. Charter school authorizers
As a policy matter, one might regard consumer choice as the objective measure of social welfare so if charter schools are expanding to meet consumer demand, public policy should not constrain their growth. Moreover, one can be confident that consumer benefits will grow steadily and perhaps sharply, as competition thins out the weak performers and KIPP-style chains build national reputations.
However, the political consensus on public charter schools is that in addition to satisfied customers, they must also demonstrate successful learning outcomes. So public charter schools occupy a middle ground in the regulatory spectrum, between voucher programs (largely unregulated beyond a market test) and the regular public schools (which rely mainly upon regulatory oversight rather than a market test).
The charter laws assign this oversight responsibility to authorizers who decide whether a charter will be issued, renewed, or revoked. But the laws offer little guidance on how to carry out this responsibility. They typically say that charter schools must participate in the statewide assessment program and submit an annual report on their academic progress, and leave everything else to the authorizers.
Consequently, public oversight over charter schools' learning outcomes has been evolving along with the schools. The original expectation was that charter school authorizers would exert their influence through selective approval, renewal, and revocation of charters. In practice, they have mostly chosen to use this power indirectly i.e., as a threat to induce improvements in the charter schools they oversee rather than by directly exercising this power.(29) There are several good reasons:
The last point is the most interesting one. Looking at this through the prism of the evolutionary triad variety, selection, and replication/development authorizers have decided to focus on assisting struggling entrants (which fosters variety), requiring the schools to articulate clear goals and measures of progress (which fosters selection through the competitive marketplace), forcing management or other changes in failing schools (which fosters selection through efficient exit, while preserving the salvageable capabilities of the school), and encouraging new-school developers to emulate the types of charter schools that appear to be succeeding (which fosters replication and development of the successful varieties).
Some of the current "best practices" of authorizers could go into charter laws as they are revised.(30) That would facilitate efficient entry and exit a procompetitive benefit and it could help persuade legislators to provide alternative authorizers and remove caps on charter schools.
The most anticompetitive constraints on public charter schools are explicit caps, exclusive reliance on local school districts as authorizers, and impediments to multi-school charter systems. Some state legislators might be willing to relax those constraints if they felt more confidence in the public oversight provided by authorizers. I think that confidence is warranted, and I think we can also feel confident that authorizers are acting mainly to enhance competition rather than to supplant it.
* This paper benefited from generous comments by John Bender, Ron Drennan, Nicholas Hill, Diane Owen, John Sawyer, and Peter Woodward. The views expressed are not those of the U.S. Department of Justice.
1. In the District of Columbia, charter schools now serve nearly one student in four, and it is still uncertain whether the regular school system will adapt or disappear. A recent front-page article in the Washington Post asked, "Will traditional public schools improve with competition? Or will charters take over?...With public confidence in the [District] schools at an all-time low, more than 17,000 public school students nearly one in four have rejected the traditional system...That share is one of the largest in the nation and is expected to rise when six more charter schools open their doors this fall." (Lori Montgomery and Jay Mathews, "The Future of D.C. Public Schools: Traditional or Charter Education?" The Washington Post (8/22/2006) p. 1.)
2. Ironically, it is the proponents of public charter schools who are urging authorizers to become more assertive in weeding out the bad schools.
3. Todd Ziebarth, "Stunting Growth: The Impact of State-Imposed Caps on Charter Schools," The National Alliance for Public Charter Schools (January 2006) http://www.publiccharters.org/content/publication/detail/533/
4. NYC Public Schools Chancellor Joel Klein's Remarks to the NY Charter School Association's Conference (March 27, 2004) http://www.ppionline.org/ndol/print.cfm?contentid=252665
5. Susan Saulny, "Harlem, a Test Lab, Splits Over Charter Schools," The New York Times (June 2, 2006)
6. New York City Public Schools Press Release, "Schools Chancellor Joel I. Klein Welcomes Four New Charter Schools Approved By State Board Of Regents And Renews Call For Lifting Of Charter School Cap As Final Slots Are Filled" (1/10/2006) http://schools.nyc.gov/Administration/mediarelations/PressReleases/2005-2006/01102006a.htm
7. Source: Summaries of state charter laws prepared by the Center for Education Reform (http://www.edreform.com/index.cfm?fuseAction=cLaw).
8. Simona Kúscová and Jack Buckley, "The effect of charter school legislation on market share," 12 Education Policy Analysis Archives (no.66) (November 30, 2004) http://epaa.asu.edu/epaa/v12n66/v12n66.pdf
9. This section draws mainly from John E. Chubb, "Should Charter Schools Be a Cottage Industry?," chap. 5 in Paul Hill. Charter Schools Against the Odds: An Assessment of the Koret Task Force on K12 Education (2006) http://media.hoover.org/documents/0817947620_127.pdf
10. "African Americans make up 35 percent of the for-profit enrollment and 53 percent of the not-for-profit enrollment versus a national average enrollment for African Americans of 13 percent. Poverty has a similar tendency. Students eligible for free or reduced-price lunch represent 54 percent of the for-profit enrollment and 71 percent of the not-for-profit enrollment; the national public school average is only 38 percent." Chubb, op. cit., p. 148.
11. Chubb, op. cit., p. 155.
12. For a current review of evolutionary theory and evidence, see Stanley Metcalfe, "The Evolution of Industrial Dynamics" (June 2005) http://www.druid.dk/conferences/summer2005/papers/ds2005-605.pdf
13. Based on survey evidence, "parents rate schools the same as experts (showing they have sufficient information to choose correctly)" and "most parents choose schools on the basis of their perceived academic quality (showing they are choosing in the child's best long-term interests)." Joseph L. Basta and Herbert J. Walberg, "Can Parents Choose the Best Schools for their Children?," 23 Economics of Education Review 431 (2004) http://www.heartland.org/pdf/18622.pdf
14. This model of firm behavior focuses on a few influential characteristics, ignoring a host of complications. In particular, I am assuming here that both charter and regular public schools have the authority to manage their own operations. In some states, that does not hold for charter schools, because the charter law gives so much authority to the school district granting the charter. On occasion, a publicly-owned school system may also have its decisions overruled by a higher authority.
15. The funding gap averages 20-30 percent, but varies greatly by state. Gregg Vanourek, "State of the Charter Movement 2005: Trends, Issues, & Indicators," Charter School Leadership Council (May 2005), p. 30 (http://www.publiccharters.org/files/543_file_sotm2005pdf.pdf); and Thomas B. Fordham Institute et al, "Charter School Funding: Inequity''s Next Frontier" (Aug 2005) http://www.edexcellence.net/doc/Charter%20School%20Funding%202005%20FINAL.pdf
16. Large urban school systems are in a constant state of reform, but they don't identify and adopt successful innovations in preference to unsuccessful ones. See Frederick M. Hess, Spinning Wheels: The Politics of Urban School Reform (1998) and "The Spinning Wheels of Urban School Reform," National Charter School Clearinghouse Review (July 2003) http://www.nationalcharterschoolclearinghouse.net/NCSCReview/Journal_three.pdf
17. John Jacob Cannell, " 'Lake Wobegon' Twenty Years Later," 2 Third Education Group Review (March 2006) http://www.tegr.org/Review/Articles/vol2/v2n1.pdf
18. In most states, when a regular public school loses a student to a public charter school, the lost marginal revenue is less than the average per-pupil revenue, because only a portion of the per-pupil revenue follows the child to the charter school. On the other hand, the marginal cost of educating these children may also be less than the average cost in some school districts. So the net effect on rents is uncertain, and may vary from one school district to the next.
19. Paul T. Hill, Lawrence Angel, and Jon Christensen, "Charter School Achievement Studies," 1 Education Finance and Policy 139 (Winter 2006) http://www.mitpressjournals.org/doi/pdf/10.1162/edfp.2006.1.1.139
20. Eric Bettinger,"The Effect of Charter Schools on Charter Students and Public Schools," 24 Economics of Education Review 133 (2005) Working paper: http://www.ncspe.org/publications_files/182_OP04.pdf (November 1999)
Robert Bifulco and Helen F. Ladd, "The Impacts of Charter Schools on Student Achievement: Evidence from North Carolina," 1 Education Finance and Policy 50 (Winter 2006) http://www.mitpressjournals.org/doi/pdf/10.1162/edfp.2006.1.1.50
George M. Holmes, Jeff DeSimone, and Nicholas G. Rupp, "Friendly Competition: Does the Presence of Charter Spur Public Schools to Improve?" Education Next (Winter 2006) http://www.educationnext.org/20061/67.html
George M. Holmes, Jeff DeSimone, and Nicholas G. Rupp, "Does School Choice Increase School Quality?" NBER Working Paper No. W9683 (May 2003) http://www.nber.org/papers/w9683
Caroline M. Hoxby, "School choice and school competition: Evidence from the United States," 10 Swedish Economic Policy Review 11 (2003) http://www.economics.harvard.edu/faculty/hoxby/papers/hoxby_2.pdf
21. Kevin Booker, Scott Gilpatric, Timothy Gronberg, and Dennis Jansen, "The Effect of Charter Schools on Traditional Public School Students in Texas: Are Children Who Stay Behind Left Behind?" (September 2005), p. 20. http://ncspe.org/publications_files/OP104.pdf
22. When two popular D.C. charter schools closed this summer, parents scrambled to find another charter school for their children. Adding to their difficulties is the uniqueness of each charter school and the added competition for slots from unhappy D.C. public school parents. Theola Labbé, "Charter School Closures Strand D.C. Students," The Washington Post (8/17/06)
23. Metcalfe (June 2005), op. cit., p. 9
24. Ibid., p. 20.
25. Figure 1 in Eric A. Hanushek, John F. Kain, Steven G. Rivkin, Gregory F. Branch, "Charter School Quality and Parental Decision Making with School Choice," SIEPR Discussion Paper No. 04-24, Stanford Institute for Economic Policy Research, Stanford University (March 2005). http://siepr.stanford.edu/Papers/pdf/04-24.pdf
"Quality" is the residual of a regression of school average achievement gain in a given year on a number of student characteristics. The authors note that "the much larger variation in the charter sector results at least in part from the much smaller enrollment levels in most charter schools. Nevertheless, the systematic quality differences by years of operation capture real performance differences...." (p. 22)
26. Ibid., p. 2
27. Rebecca Gau, Chester E. Finn, Jr., Michael J. Petrilli, "Trends in Charter School Authorizing," Thomas B. Fordham Institute (May 2006), p. 10.
29. Katrina Bulkley, "Educational Performance and Charter School Authorizers: The Accountability Bind," Education Policy Analysis Archives 9(37) (October 1, 2001) http://epaa.asu.edu/epaa/v9n37.html
30. See, for example: Bryan Hassel and Paul Herdman, "Charter School Accountability: A Guide to Issues and Options for Charter Authorizers" (2000) (http://www.uscharterschools.org/gb/account_auth/index.htm); National Alliance for Public Charter Schools (NAPCS), "Renewing the Compact: A Statement by the Task Force on Charter School Quality and Accountability" (9 Aug 2005) (http://www.publiccharters.org/content/publication/detail/544/); Progressive Policy Institute (PPI), "From Margins to Mainstream: Building a Stronger Charter School Movement," Conference Essays and Transcript (July 17-18, 2003) (http://www.ppionline.org/documents/Charter_Conf_Essays_0703.pdf and http://www.ppionline.org/documents/Charter_Conf_Transcripts_0703.pdf)