Summary: Twin studies that estimate the genetic influence on human traits wrongly assume that parents treat all their children similarly. Instead, parents are well known to treat different children differently. I believe that parents treat identical twins more similarly than they treat non-identical twins, and that this difference is large enough to account for almost all of the differences between the twins. In other words, environment alone, with little genetic influence, can explain almost all of the differences between people.

Recommended background for this section: How Personality Develops

Here are the details:

Twin Studies

Twin studies estimate genetic influence by assuming that identical twins (monozygotic, MZ) and fraternal twins (dizygotic, DZ) share equally similar environments. This is known as the equal environment assumption (EEA).

Here’s how this works: If identical twins share an environment that is, say, 90% similar, the EEA assumes that fraternal twins also share an environment that is about 90% similar. Since their environments are mostly the same, if the DZ twins are more different than the MZ twins, that must be because the DZ twins have different genes.

To measure environmental similarity, parents and the twins complete surveys that ask, for example: 

* How similarly the twins were dressed

* How much time they spent together

* Whether they had the same teachers

* Whether they slept in the same room

* How many friends they had in common

I believe these measured factors have minimal impact on personality development. Instead, as I discuss in the section on personality formation, the emotional environment while interacting with their parents is the most critical factor.

Because of this, I argue that MZ twins actually experience a significantly more similar environment than DZ twins.

Here’s how this works. With MZ twins:

* When MZ twins express the same emotion, it appears virtually identical because of their identical facial features and voices.

* The caregiver then responds to both twins in an almost identical way.

* This reinforces nearly identical emotional responses in the twins, shaping similar personalities over time.

With DZ twins, however:

* Even if they express identical emotions, their differing facial features and voices cause those emotions to be expressed differently.

* The parent perceives slightly different emotions and responds to each twin slightly differently. 

* As a result, the twins incorporate slightly different emotional reactions to that parent.

* In subsequent interactions, the twins' emotional responses to the same stimulus will be different.

* Those different emotional responses will appear even more different because of their differing facial features and voices.

* The parent will then respond even more differently to each twin.

* Over time, this feedback loop leads to divergent personalities.

As a result of the above, MZ twins experience a much more similar emotional environment than DZ twins, making the assumption of equal environmental influence flawed.

Population Studies

Studies that attempt to link genetics to human traits are riddled with methodological problems—so much so that Jay Joseph devoted an entire book, The Gene Illusion (2004, Algora Publishing), to exposing their flaws. Here are just two major issues he highlights:

1. Problems with “twins reared apart” studies:

    * Many twins were not separated at birth, meaning they shared similar environments during crucial early developmental years.

    * Even when separated, twins often maintained significant contact with each other.

    * Many were raised by relatives, who likely had similar parenting styles to the biological parents.

2. Inconsistencies in schizophrenia concordance rates:

    * Before 1962, schizophrenia diagnoses were much broader, leading to a reported concordance rate of about 60% in identical twins.

    * Between 1962 and 2004, as diagnostic criteria became stricter, this rate dropped to around 21%.

    * Some researchers include pre-1962 studies in their analyses, inflating the perceived genetic link, while others exclude them, leading to much lower estimates.