Sometimes, it may appear that a country may have different scores depending on where the Fragile States Index is published. This does not indicate an error in the data. Rather, this is due to the rounding up- or down- of scores to one decimal place, depending on when that rounding takes place. If the individual indicators are rounded, this may affect the total score versus if only the total score is rounded. (See also this article)

Take this example from 2015:

## Austria |
|||

Indicator |
Raw Score |
Rounded atTotal Score |
Rounded atIndicator Level |

DP | 2.4 | 2.4 | 2.4 |

REF | 1.99437482787231 | 2.0 | 2.0 |

GG | 4.25530726256983 | 4.3 | 4.3 |

HF | 1.5 | 1.5 | 1.5 |

ECO | 3.4 | 3.4 | 3.4 |

UED | 2.16741573033708 | 2.2 | 2.2 |

SL | 1.35865921787709 | 1.4 | 1.4 |

PS | 1.57039106145251 | 1.6 | 1.6 |

HR | 1.7 | 1.7 | 1.7 |

SA | 1.1 | 1.1 | 1.1 |

FE | 2.7 | 2.7 | 2.7 |

EXT | 1.7 | 1.7 | 1.7 |

Total |
25.8461481001088 |
25.8 |
26.0 |

Where the scores are rounded at the indicator level, the total score reflects the sum of individual indicator scores that have already been rounded. Where the scores are rounded at the “final post”, i.e., the total score, it means that only the total score itself is rounded and ** may not actually appear to add up** to the sum of the indicators as they are commonly represented to one decimal place.

Which is correct? Both scores are technically correct, as they are the same base scores, but simply rounded in different ways. Fortunately, this discrepancy is rare and affects very few countries’ scores, and where it does it is minimal and within what is considered the normal “margin of error.”