Recently we heard that government agencies are fudging the numbers to make themselves look better. Or as The Salvation Army's latest State of the Nation report argues, government statistics can be tweaked and manipulated to illustrate favourable results.

This has caused what The Salvation Army call a "moveable feast" mentality: searching for any reason to celebrate successes.

The army brings up a valuable point. While government and ministries collect and report an abundance of statistics on a range of key social indicators, they cannot be taken at face value. Numbers may be numbers. But different measures capture different things about this nation, and represent different priorities.

And the interpretation of these measures is generally politicised. Following The Salvation Army's report, Question Time in parliament certainly brought that point home.

But is the Government hell-bent on promoting a "moveable feast", or have others simply been setting a "moveable funeral" for too long?

Let's review, using poverty as an example. The most comprehensive and most reported-on poverty statistics come from the Ministry of Social Development (MSD). MSD produces hundreds of pages of household income measures and hardship measures. Inevitably, these hundreds of pages get boiled down to one headline-grabbing figure.

How is that one figure decided? Politics, of course.

While the government may be accused of cherry-picking the numbers that make it look good, perhaps there are others: advocacy groups, opposing political parties, and commentators, that cherry-pick numbers that make the situation look worse than it is.

This is one of the key arguments in The New Zealand Initiative's latest report on poverty: Poorly Understood: The state of poverty in New Zealand. As the name suggests, we believe poverty in New Zealand is often misunderstood, misrepresented, and misreported.

For example, inequality looks worse before taxes and transfers. A 2015 Treasury report (Inequality in New Zealand 1983/84 to 2013/14) found that measures of consumption inequality yielded much lower rates than traditional income measures. While pre-tax measures give more shocking statistics, post-tax measures show the effectiveness of the redistribution system.

Meanwhile, the MSD reports there is not a significant overlap between relative material hardship and measures of relative income inequality. Typically, the overlap in these two different ways of measuring poverty is around 35-45 per cent for the population as a whole.

Housing costs matter, too. In 2014, one measure (using the 1998 median income as the threshold) saw households in poverty before housing costs as 8 per cent of the population, while after housing costs that proportion rose to 13 per cent. From a long-term perspective, while poverty had decreased from 1982 to 2014 before housing costs, it had increased after housing costs were considered.

Finally, fixed line and moving line thresholds make a difference. As MSD describe it, "declining fixed line rates mean that the incomes of low-income households are on average rising in real terms, whereas declining moving line rates means that the incomes of low-income households are getting closer to incomes of middle-income households (whether the low incomes are rising or falling in real terms)." While the former seems like a good measure, the latter more reflects inequality.

The range of measures and social indicators collected in New Zealand should be a celebrated strength for those interested in evidence-based policies and solutions. But it is impossible to report on these measures objectively. In fact it is hard to believe objective ways to measure poverty even exist in developed countries in New Zealand.

Definitions of poverty in New Zealand, and ways of measuring it, are necessarily subjective. And that's OK.

Some organisations may be incentivised to promote a "moveable feast", while others, such as those whose primary role is to advocate for those in poverty, may be incentivised to promote "moveable funerals" to keep poverty at the forefront of the public's mind.

Overcoming subjectivity may be impossible, but transparency around definitions and ways of measuring social indicators is important.

After all, to make long-lasting and effective change, surely the first step is identifying who welfare policies are intended to help in the first place.