Deeds, not words. For all the talk, the Shannon remains undrained (unless perhaps to water Dublin). And the top rate of tax starts below average earnings.
Those were fine words from the Taoiseach, saying the limit of the standard 20pc tax band would be increased from the present €35,300 to €50,000 - just above average earnings of €46,000.
But fine words should be kept a safe distance away from deeds which do not match them. In this case, curiously, they came just weeks after deeds which went the other way, in the October Budget.
As with most budgets of the past 35 years, this one neglected to raise the standard rate band fully in line with earnings - a process known as indexation. The increase was just over 2pc, whereas earnings are forecast to rise by more than 3pc.
Full indexation would have reduced government revenues by €600m. Bringing the top rate entry closer to average earnings would have cost even more. It is an expensive business, which is why progress has been slow to non-existent so far.
During the crash there was reverse indexation, where the entry point to the top rate fell significantly, to a low of around €33,000. That makes the average earnings target even harder to attain.
Mr Varadkar blamed "reckless economic mismanagement of the past" but there are reasons why politicians do not like the process, even in better times. Budget 2019 shows them quite clearly.
First of all, indexation seems to bring limited electoral benefit. Government ministers can hardly feel they got €160m worth of headlines from the measures they did take.
Nowadays, things are further complicated by the existence of the unpopular USC. The €500 increase in its bottom rate band, along with a quarter point cut in the middle rate, took another €120m out of government coffers.
The total cost of both measures is a not inconsiderable €280m, yet they received few cheers in the generally critical reaction to the Budget.
Before tax-cutting got a bad name in the crash, past governments often preferred rate cuts to indexation. Reducing the 40pc rate to 39pc would cost about €350m a year but, back then, such cuts got more political approval per million. So indexation came a poor second and the low entry point problem remained unsolved.
Cutting rates like that would now elicit howls of disapproval - although not on USC for some reason. That seems to be regarded as almost a government obligation rather than a policy choice, so there is not a lot of kudos to be had from it.
Quite the opposite. Even the most progressive budgets of recent years were widely decried as unfair. Something else is needed politically, and the Taoiseach has delivered it.
One can get remarkable numbers by rolling a few years together, as his Ard-Fheis speech demonstrated. Three billion in tax cuts, worth up to €250 a month for the squeezed middle. Suddenly, indexation is hot.
For that is all the Mr Varadkar has promised - and perhaps not even that in full. Average earnings will probably be above €50,000 in five years. It is a moving target. There is also the little matter that, in any common sense definition, indexation is not a tax cut, even if it suits politicians in general to call it that.
Non-indexation, on the other hand, is a tax rise for those whose earnings move from below the standard rate cut-off point to above it. This was the original "stealth tax" before the phrase came to mean just about any unpopular tax rather than one which is largely invisible.
It is very handy for finance ministries and some, as in Britain, make it automatic before they go on to real tax cuts. There won't be much room for any of those, under Mr Varadkar's plans, once the annual cost of full indexation is accounted for.
Mr Varadkar is probably right in hoping that it will be seen as a tax cut rather than a tax rise foregone, but he will face the same dilemmas as all those predecessors in doing without the stealthy revenue from non-indexation.
Spending is the bugbear, as it nearly always is. Last week's analysis of the budget from the Fiscal Advisory Council (IFAC) painted an unflattering portrait of a government which is missing its EU commitments because spending is following a pattern not dissimilar to the early years of the bubble.
Corporation tax has taken the place of property based taxes in funding this spending. At least we knew where property revenues originated and what might stop them in their tracks (even if there were conflicting views on whether they would. With corporation tax, we do not know even that much.
This opens a distinctly black hole in that grand Five-Year Plan. The budget figures were used to show it is all possible, with their picture of tax revenues increasing by €12bn, current spending by €7bn and capital by €3bn over the period. But all is not what it seems.
The revenue estimates are simple enough, even if based on rather cheery forecasts about economic performance over the next five years. However, as IFAC pointed out, the spending figures are based on technical assumptions. They include some costs of an ageing population, as well as increased public investment, but no other new spending.
They purport to be the kind of medium-term framework which other countries produce, but are nothing of the kind. By IFAC's calculation, they imply that spending will fall from the present 38.7pc of national income (GNI*) to 36.2pc by 2023.
Fancy a walk across the Shannon? No words could be further from recent deeds than these ones.
Far from flattening, never mind falling, public spending has been rising at bubble-era rates in the past four years.
Should it continue to do so for the next four or five, the fiscal targets will be blown away once tax revenue growth slows, as eventually for whatever reason, it will. Some calculations suggest there is already a potential deterioration of five percentage points in the deficit after even a normal global shock.
It seems clear that Finance Minister Paschal Donohoe lost the late summer budget battle. The now almost routine end-of-year increase means spending will rise by 5.7pc this year, as compared with the planned 4.1pc.
To meet this, the budget had a €350m increase in taxation, to which non-indexation added a further €600m.
If this was a pre-election budget, we can see where the election is headed: the tried and tested formula of spending promises which are more or less honoured if possible, and tax reductions which usually are not, because the two things are rarely possible together.
The exception in taxation promises is the mantra that those on the minimum wage must be kept out of the tax net. The unthinking application of this formula has produced the tax system of which the Taoiseach and many others complain, with its early application of the top rate and steep progression thereafter.
Mr Varadkar's answer is to tack another storey on this shaky edifice. The two-thirds of workers who are exempt or taxed at 20pc would increase to more than three-quarters. Such an income tax system could not fund a modern welfare economy, certainly not one with Ireland's insatiable desire for public spending.
In the meantime, the widespread idea will be reinforced that the solution to the issue of early entry into the top tier is to keep more people in what used to be the bottom tier, but is now the standard 20pc bracket.
Therein may lie the only possible solution: raising the top rate entry point, but not the 20pc band. Instead, earnings above the present cut-off would be taxed at a new rate of, for the sake of argument, 33pc.
It would still be slow and expensive, but it might just get to a €50,000 top rate by 2023, if accompanied by spending growth in line with the economy's potential and no nasty shocks.
Those are two enormous ifs, but anything else, however grand it may sound, is just more of the same old, same old.