Explain differences with Kaukin and Zhemkova (2023)

[memoryfull]

Our replication of a study estimating production function of Russian manufacturing firms in 2012-2018 by Kaukin and Zhemkova (2023) revealed unrealistic capital elasticities.

industry	labour	capital
01. Food, beverages, and tobacco	0.78	0.09
02. Textiles and textile products	0.83	0.03
03. Wood and products of wood and cork	0.74	0.05
04. Pulp, paper, printing and publishing	0.82	0.07
05. Coke, refined petroleum	0.73	-0.04
06. Chemicals and chemical products	0.72	0.02
07. Pharmaceutical products	0.82	0.20
08. Rubber and plastics	0.79	0.09
09. Other non-metallic mineral	0.69	0.05
10. Basic metals	0.63	0.06
11. Fabricated metals	0.75	0.05
12. Electrical and optical equipment	0.83	0.12
13. Electrical equipment	0.78	0.01
14. Machinery, not elsewhere classified	0.74	0.04
15. Transport equipment	0.69	0.16
16. Furniture	0.84	0.07
17. Manufacturing, not elsewhere classified	0.81	-0.01
18. Equipment repairs and installation	0.85	0.00

We provide 5 explanations for the apparent failure to replicate KZ exactly:

• Data differences between Orbis and the RFSD
• Differences in sample restrictions
• Differences in deflators
• Value added definition
• Proxy variable

Thank you for any contributions.