Haver's estimate for older children is longer than the limit imposed by anti-bilingual education initiatives she supported (Proposition 227 in California, Proposition 203 in Arizona). Her estimates in general are also in conflict with the empirical research.

For English learners in New York City's all-English program who started school in the US at kindergarten, only 44% were ready for the mainstream after one year. This confirms that one year is simply not enough for many children, no matter what kind of program is used and when they begin school (New York City, 2000).

Even two years is not enough for many children. A University of California at Riverside study (Mitchell, Destino and Karem, 1997) reported that English learners in an English immersion program only improved from 2.18 to 3.4 on a 1-5 scale in after two years, where 4 = sufficient proficiency to survive in the mainstream. Ramirez (1992) reported that children in immersion were nowhere near ready for the mainstream after two and even three years, even though 70% knew some English before they started school. At the end of first grade (two years of exposure), only 21% reached the redesignation standard, and at the end of grade 2, only 38% achieved this level.

The Case of David Ho: Six Months with De Facto Bilingual Education

Haver presents the case of David Ho, who came to the US at age 12 without any knowledge of English. After six months, he was able to get along in English in school, and eventually went on to become Time Magazine's Person of the Year in 1996 for his work in molecular biology. Haver notes that this is a "rare" case, but it must be pointed out that Ho had the benefit of "de facto" bilingual education: He was a good student in Taiwan before he come to the US, and noted in an interview that there was "a great deal of emphasis on scholarly endeavors" in his family (Ho, 1998). Ho also came from a very academically inclined family: His father was an engineer who had done graduate studies in the United States. Ho had both of the advantages of bilingual education: He had a great deal of subject matter knowledge and was very literate in his primary language. These advantages helped make the English curriculum comprehensible and accelerated his English language development. It is possible that he also received extra help from family and friends using the first language. His case, and others like it (Krashen, 1996; Tse, 1997; Ramos and Krashen, 1997) provide strong support for bilingual education.

A BIASED BIBLIOGRAPHY

Haver ends her book with a biased bibliography of recommended readings, one that gives the reader the impression that there is a vast amount of uncontested evidence supporting her view.

She describes Baker's 1998 paper in the Phi Delta Kappan as a "convincing case for the superiority of structured English immersion over bilingual education" (p. 117). She does not mention my response to Baker, published in the same journal (Krashen, 1999b). She cites Carter's No Excuses, but does not cite Bracey and Biddle's response. She describes Eckman, Highland, Lee, Millham and Weber (1995) as a book containing studies that show that grammar teaching improves immersion education. Not so. Only one paper in this volume mentions this literature. The other 18 are on a wide variety of topics, including mine reviewing the evidence for free voluntary reading. She cites Rossell (2000), but does not cite any of the many critiques of Rossell's conclusions (e.g. Krashen, 1996, Greene, 1997). Readers should know that there is more than one point of view in the scientific literature.

7