Rosstat revised up its construction growth estimate for the end of 2018 from zero to 5.3% y/y, in a move that raised eyebrows amongst Moscow’s analysts.

The state statistics agency was under fire last year for the poor quality of its work and several large revisions that dramatically altered the picture of Russia’s economy. Minister of Economy Maxim Oreshkin publicly lambasted the agency last year and its boss was replaced.

Specifically, the statistics agency drew attention last year when it revised its figure for 2016 gross domestic product growth to a 0.2% contraction from a much deeper fall reported previously. It also revised 2015’s contraction in GDP to just 2.8% from an initial 3.7%.

The latest revision to the construction number — one of the three big drivers of economic growth — will again alter the picture of Russia’s growth. Alexander Isakov, chief economist at VTB Capital said in a note that the revision to such a key sector would take the estimate for 2018 growth up from the disappointing 1.5% — well below the expectation of 2.1% growth at the start of 2018 — to around 2%.

Rosstat is expected to release fuller figures and revise its preliminary estimates with a big data release next week, although the revised construction numbers may not make it into next week’s full report.

Isakov tried to check the construction number by looking at physical proxies, like the production and transport of cement, for building activity rather than looking at the revenue-based indicators.

“What this new revision highlights is the need for robust alternative gauges for the level of activity in construction, which would foreshadow such revisions,” Isakov said in a note. “The approach which we have advocated to supplement industrial production analysis — deliberately limiting our analysis to publicly available items which are reported in physical quantities, not deflated from the revenue data — might not have been as useful in this case. The data on physical output and about the railway turnover of construction material (cement, in particular) did not, we think, foreshadow the reported revision of construction growth.”

To be fair, Isakov points out the discrepancy may be due to reporting lags, but the big jump in the number for construction itself doesn't inspire confidence in Rosstat’s methodology.