I plotted the center point of the aggregate time ranges, and the percentages in my graph neglect the don't know/not applicable responses. I cannot find an annual series for this survey question, though the survey is apparently performed annually.

Also, Methodological Report 56 cautions:

21. Two separate context effects have distorted the time trend on personal happiness (HAPPY). First, personal happiness is higher for married persons when preceded by an item on marital happiness (HAPMAR). Marital happiness has preceded personal happiness on all surveys except 1972 and Form 3 in 1980 and Form 3 in 1987. To make a consistent series we have a) used the experimental comparisons in 1980 and 1987 to adjust the 1972 personal happiness figures and have eliminated the variant experimental forms in 1980 and 1987 (Smith, 1979). Second, personal happiness is lower when not preceded by the five-item, satisfaction scale (SATCITY, SATHOBBY, SATFAM, SATFRND, SATHEALT). This scale has routinely preceded personal happiness except in 1972, in 1985, on Form 2 in 1986, and on Forms 2 and 3 in 1987. Using the experimental comparisons in 1986 and 1987, we have adjusted the figures for 1972 and 1985 and have eliminated the variant forms for 1986 and 1987 (Smith, 1986).