1	Country bias in international dairy bull evaluations
2	MACE Evaluation Process Interbull combines yield evaluations from 27 national Holstein bull populations. Estimated genetic correlations between evaluations from different populations are applied. Results are reported on the scale of each participating country organization.
3	Investigation of Country Bias Full brothers are expected to have the same genetics on average. Full brothers with daughters in different countries should have the same MACE evaluations (on average). Do full brother evaluations reveal country bias?
4	Data May 2005 Interbull evaluations of Holsteins for milk, fat, and protein yields, and SCS Eligible bulls ≥1 eligible full-brother Daughters in only one eligible country Eligible countries Represented in ≥25 multi-country families 18 countries for yield 15 countries for SCS
5	Data
6	Multi-Country Families
7	Ties Between Countries
8	US Ties to Other Countries
9	Model Full-brother family (absorbed) Country of bull daughters Analysis of MACE evaluations on the US scale Difference between evaluations based on foreign daughters and those based on domestic (US) daughters Evaluations on several other country scales also analyzed
10	Results Overall significance for country differences on all country scales for yield traits Overall, no significance for SCS For SCS, South African bulls significantly disadvantaged on all scales
11	Results for US-Scale Evaluations
12	Results for Other Evaluation Scales
13	Results for Other Evaluation Scales
14	Results for Other Evaluation Scales
15	Possible Complications Genetic correlations Parent averages (completeness and phantom grouping) National animal model vs. sire-MGS MACE Preferential treatment Pre-selection (markers, etc.)
16	Conclusions Apparent biases for yields but not SCS Some countries appear disadvantaged while others are favored across all country scales Caused by shortcomings in national systems or international system? Source of biases unknown and may be different for different country pairs