If a bunch are NOT vaccinate, then herd immunity has been eroded. It has been eroded enough that about dozen babies each year die from pertussis, and outbreaks of diseases to occur. Plus there is no such thing as herd immunity with tetanus.

Then compare that kind of study with research like this:
http://archpedi.ama-assn.org/cgi/content/full/160/3/302

With measles being so easily transmitted (one child coming back from Switzerland caused an outbreak of at least 11 more cases!, three of them babies too young to be vaccinated), it would not be ethical to withhold medical protection.

Daedalus2u; I do not think that the study would be unethical. The benefit to the control group may be the possibility of being spared a complication. The placebo can be a subQ saline injection. If there was no immune response, only a antibody titer can determine who got the vaccine and who did not. We would not expect the naive vaccinated subjects to develop infectious disease as they would be protected by herd immunity. As for the statistics, I would assume that a statistician would have to stipulate an apriori null hypothesis. So, we would assume that taking the vaccine makes it, say 30% more likely to develop Autism. If so, the statistician would have to determine the number of patients who would need to be enrolled in order to detect a 30% difference as being statistically significant. These kinds of studies go on all the time in clinical medicine. What’s the issue?

http://www.wma.net/e/policy/b3.htm

Those in the placebo side of the study are receiving no benefit by being in the study, only risk of harm.

What would you use as the placebo vaccine? Vaccine without antigen but with aluminum? If there was no immune response, it would be easy to tell who was vaccine and who was placebo.

How many cases of autism in each group would be enough to have good statistics? How big does each group have to be in that case? At a cost of a few thousand per participant, the cost of the trial becomes quite large.

There have been lots of epidiological studies of that sort done in several countries that have not seen any real evidence that vaccines cause autism. Some of them:

Pervasive Developmental Disorders in Montreal, Quebec, Canada: Prevalence and Links with Immunizations
Authors: Fombonne E, Zakarian R, Bennett A, Meng L, McLean-Heywood D
Source: Pediatrics, July 2006, Vol. 118(1):e139-e150

The Incidence of Autism in Olmsted County, Minnesota, 1976-1997
Authors: Barbaresi WJ, Katusic SK, Colligan RC, Weaver AL, Jacobsen SJ
Source: Archives of Pediatric and Adolescent Medicine, January 2005, Vol. 159(1):37-44

No Effect of MMR Withdrawal on the Incidence of Autism: A Total Population Study
Authors: Honda H, Shimizu Y, Rutter M
Source: Journal of Child Psychology & Psychiatry, Online Early Release

Age at First Measles-Mumps-Rubella Vaccination in Children with Autism and School-Matched Control Subjects: A Population-Based Study in Metropolitan Atlanta
Authors: DeStefano F, Bhasin TK, Thompson WW, Yeargin-Allsopp M, Boyle C
Source: Pediatrics, February 2004, Vol. 113(2):259-266

Association Between Thimerosal-Containing Vaccine and Autism
Authors: Hviid A, Stellfeld M, Wohlfahrt J, Melbye M
Source: Journal of the American Medical Association, October 1, 2003, Vol. 290(13):1763-6

Thimerosal and the Occurrence of Autism: Negative Ecological Evidence from Danish Population-Based Data
Authors: Madsen KM, Lauritsen MB, Pedersen CB, et al
Source: Pediatrics, Sept. 2003, Vol. 112(3 Pt 1):604-606

Prevalence of Autism and Parentally Reported Triggers in a North East London Population
Authors: Lingam R, Simmons A, Andrews N, et al
Source: Archives of Disease in Childhood, August 2003, Vol. 88(8):666-670

A Population-Based Study of Measles, Mumps, and Rubella Vaccination and Autism
Authors: Madsen KM, Hviid A, Vestergaard M, et al
Source: New England Journal of Medicine, November 7, 2002, Vol. 347(19):1477-1482

And there is a review of some of the literature here:
http://pediatrics.aappublications.org/cgi/content/full/114/3/793