Consider the following benchmark:

package escaping

import (
        "html"
        "strings"
        "testing"
)

const noHTMLString = `Can you hear the people sing, singing the song of
angry men. It is the music of a people who will not be slaves again.`
const withHTMLString = `We're going to demand <strong>one million
dollars</strong> or we're going to threaten Earth with a "Laser".`

var escaper = strings.NewReplacer(
        `&`, "&amp;",
        `'`, "&#39;",
        `<`, "&lt;",
        `>`, "&gt;",
        `"`, "&#34;",
)

func BenchmarkHtml_EscapeString_nochanges(b *testing.B) {
        for i := 0; i < b.N; i++ {
                html.EscapeString(noHTMLString)
        }
}

func BenchmarkStrings_Replacer_nochanges(b *testing.B) {
        for i := 0; i < b.N; i++ {
                escaper.Replace(noHTMLString)
        }
}

func BenchmarkHtml_EscapeString_changes(b *testing.B) {
        for i := 0; i < b.N; i++ {
                html.EscapeString(withHTMLString)
        }
}

func BenchmarkStrings_Replacer_changes(b *testing.B) {
        for i := 0; i < b.N; i++ {
                escaper.Replace(withHTMLString)
        }
}

The performance is as follows:

BenchmarkHtml_EscapeString_nochanges      500000              3912 ns/op
BenchmarkStrings_Replacer_nochanges      5000000               368 ns/op
BenchmarkHtml_EscapeString_changes        500000              4597 ns/op
BenchmarkStrings_Replacer_changes        1000000              1360 ns/op


I've written https://golang.org/cl/141930043/ to change html.EscapeString so
that it uses a replacer internally, but it needs to wait until Go 1.5 development opens.