Birth order, an effect that melted away

The idea that rank among siblings influences personality dates from the early twentieth century and was popularised in the 1990s, notably by Frank Sulloway's work on the propensity to challenge the established order. Its origin is generally attributed to Alfred Adler, who gave position among siblings an important place. Sulloway's 1996 book, Born to Rebel, drew on historical data about scientists and actors in revolutionary periods.

The proposed mechanism is differentiation within the family: each child would occupy a niche left free by the previous ones. The hypothesis is appealing because it offers a simple mechanism and echoes family experiences shared by many.

Comparing eldest and youngest children across different families is misleading, because rank is tied to sibship size, itself tied to social background and other characteristics. Large families by definition include more younger children, and if they differ from other families, those differences are wrongly attributed to rank.

The valid comparison is made within one family, between siblings, which few older studies allowed for want of data. It requires surveys that question several children from the same family, or registers that make it possible to link them to one another.

Julia Rohrer, Boris Egloff and Stefan Schmukle published in 2015 an analysis of large German, American and British samples, comparing children within the same sibship.

They find no notable effect on the major personality dimensions. They do find a weak but consistent effect on cognitive scores and educational attainment, declining with rank.

That twofold result matters: the absence of an effect on personality cannot be explained by a lack of power, since the same data detect a weak effect on other measures.

The measured gap in cognitive scores is very small at the individual level, of the order of a few points, which is invisible within a given family but detectable across tens of thousands of observations.

Proposed explanations bear on resource dilution and on the teaching role eldest children may occupy, none of them established. Other work, conducted on Norwegian conscript registers, finds a gap in the same direction in cognitive scores.

Character descriptions by rank are general enough to find confirmation in almost any sibship, a mechanism known as the Barnum effect and documented since the 1940s.

In addition, parents compare their children with each other, not with the general population, which amplifies real but small differences into a strong narrative contrast. A small average difference, taken up in family stories, thus ends up being perceived as a character trait.