When the number of crorepati taxpayers in the country rises from 88,649 in FY14 to 140,139 in FY17\u2014for individual taxpayers, it rose from 48,416 to 81,344\u2014it is undoubtedly a big achievement. The number of individual taxpayers declaring an income of over Rs 10 lakh\u2014the highest tax rate becomes effective at this level\u2014has risen even faster, from 24.4 lakh to 45.2 lakh in the same period. What is disconcerting, however, is that, despite this sharp hike in the number of people declaring higher incomes, the direct tax growth has been somewhat muted except in FY18; detailed data for FY18, though, is not available so far. So, in FY14, the direct tax-to-GDP ratio was 5.62% and this fell to 5.57% in FY17; provisional data for FY18 shows that this rose to 5.98%; with this, direct tax buoyancy rose from 1.16 in FY14 to 1.81 in FY18; it was just 1.17 in FY17. How much this increase in buoyancy is related to better tax compliance due to the taxman\u2019s efforts\u2014also, demonetisation and GST\u2014and how much is due to higher economic growth is not clear though. In FY08, when nominal GDP grew 16.1%, the direct tax-to-GDP ratio was 6.3% versus 5.98% in FY18 when GDP grew at a lower 9.96%; as GDP growth rises, people move into higher tax brackets\u2014\u201cincome-creep\u201d in jargon\u2014and that is why tax buoyancy rises. At 140,139 in FY17 for all taxpayer groups\u2014individuals, HUFs, AoPs, companies\u2014and 81,344 for individuals, the number of crorepatis is quite small compared to the PRICE estimate of one million crorepati households in the country. PRICE\u2019s data is for households, but that translates to around 6.6 lakh individuals with an annual income of more than Rs 1 crore a year. This means that just around an eighth of eligible households are declaring their incomes. The tax data for FY17 shows that there were 1.4 million taxpayers declaring an annual income of more than Rs 20 lakh; 1.3 million in the case of individual taxpayers. Once again, this compares poorly with PRICE\u2019s estimate of 8 million households or 5.3 million individuals with an income of more than Rs 20 lakh per annum. In all probability, what this suggests is that the taxman has been quite successful in getting taxes from the salaried class\u2014according to PRICE, roughly a fifth of all households earn a salary and an equal number are self-employed in agriculture (they don\u2019t have to pay taxes) and self-employed in non-agriculture. More worrying is the fact that, despite the growth in the number of rich people declaring their incomes, the average tax rate is quite low. In FY17, the tax data shows a total of 4.7 crore individuals filed their incomes and declared their incomes to be Rs 28.2 lakh crore; the time-series data shows they paid a total tax of Rs 3.5 lakh crore at an average rate of 12.4%. In FY14, a total of 3.7 crore individuals paid a tax of Rs 2.4 lakh crore on an income of Rs 18.4 lakh crore, making the effective tax rate for that year 13.2%. This gets worrying when juxtaposed with another statistic, that of what proportion of taxes are collected due to the taxman scrutinising returns and telling people they need to pay more taxes. Since 2000-01, roughly 35% of direct taxes come from TDS, another 40% from advance tax and around 7-8% from self-assessment; only around 8-10% is paid based on the taxman\u2019s demand after the assessment of returns filed by individuals Ideally, this post-assessment ratio of taxes should be zero, implying that everyone is paying as much as they should. Given that a small fraction of higher income groups are paying their taxes, this low post-assessment ratio of taxes, however, means the taxman needs to work harder on catching non-filers or those paying less than is due. The good news here is that, with GST and even demonetisation, a lot more people will automatically come into the net and will need to file more realistic tax statements. That is why, in FY18, the government is looking at a sharp increase in tax buoyancy\u2014to 1.81 as compared to 1.18 in FY17\u2014despite GDP growing at a slower rate.