D. Milne & J. Savage (ed) Reporting Economics: A NZ Guide to Covering the Economy, (JTO, 1999) p.103-110.

Keywords: Distributional Economics;

The morning this section was drafted, I listened to a lead story dealing with a statistical study which showed income inequality had been increasing over the last 15 years. The work was familiar because the results had been first reported three years earlier. The presentation made the New Zealand cricket team look world class. The journalists confused wealth and income, gross income and net income, income levels and income shares, and percent change and percentage point change. Deductions were drawn that simply are not in the data. Given it was one of our top news teams, the inference to any journalist faced with a story about the income distribution is “dont”.

Yet the material was newsworthy. The public cries out for this sort of information. “Is it fair” underpins many stories, in the economic, social, and criminal domains. Instructively, the interviewed spokesman for an Opposition party said the income distribution was as important as real income growth (and the government minister said it was not). How is a journalist to meet the strong public demand for discussion on equity issues, but avoid the minefield of the technical complexities? (How is this writer to pack into the space provided an adequate coverage, when writing for a technical audience is difficult enough?)

This section first explains the general issue of assessing equity, defines a few key notions in the following section, and finishes with some remarks on strategies of how one might handle economic equity stories.

Equity and Other Treacherous Concepts

As any philosopher will tell you, that most people are ready to talk about fairness, does not mean it is a rigorous concept. Economists appear to avoid the issue by referring to distributional “equity” (the noun also has sharemarket meanings). But of the nine economic dictionaries by my desk, five do not define the term, and the rest say it is “fairness”. In comparison many other well used notions – like GDP – are reasonably precisely defined.

Once economists thought they could avoid making judgments about equity, and still draw policy conclusions. To simplify, it seemed that “efficiency” could be increased without affecting economic equity. Unfortunately, as so often happens with technical jargon, the economist’s notion of efficiency is treacherous to the layperson. Economists sometimes talk about the “efficiency-equity tradeoff”, with the implication that in order to achieve greater efficiency there will has to be a loss of equity. That cannot be true, in the normal meaning of the words. Imagine a leftwing Treasurer announcing their government’s policy was “to pursue equity inefficiently”, which appears to be the commonsense meaning of the tradeoff.

Economists have some rigorous notions involving equity. Here are some common examples, although inevitably the simple exposition has suppressed caveats. Horizontal equity refers to two people in similar circumstances being treated exactly the same. Typically the similarity of circumstances includes the same income, and similar personal and family characteristics. For instance it would be equitable (i.e. fair) to give two people on the same income the same health or educational entitlements, even though one was, say, a beneficiary, the other a worker. Vertical equity is the principle that people in similar circumstances, except for different incomes, should be treated on the basis of ability to pay. This was a traditional justification for progressive income tax, but has not been a central policy principle in recent years. Intergenerational equity is a form of horizontal equity, in which the situation of different generations with similar circumstances are compared. It has been important in recent years in regard to government borrowing (which places a burden on future generations), and retirement policy. Very often we want to make comparisons involving those not “in similar circumstances”, but rigorous comparisons are not defined. It can be argued that two people are always in different circumstances.

In these various judgements income plays a major role. Economists can measure the income distribution, without necessarily making judgements about it. There are a myriads of ways of doing this, for there is no agreed universal measure. A little of the complexity is shown in the appendix table, where 17 estimates of “average” weekly income from 5 data bases – all quickly available from the 1998 New Zealand Official Year Book and the most recent (1996) census. In case the basic point be missed, another 17 could have been presented. It will be evident to the casual reader that here are treacherous waters, because anywhere between $883 a week and $244 a week can be shown to be as an average. There is not the room to go through all the lines, so a couple of examples will be sufficient.

First, I have seen enthusiastic journalists combining the average male wage and the average female wage (line 1) to get an “average” household weekly income of $1226, whereas even with benefits added in it is about $880 (lines 4,8).

Second, the ACT party proposed a private contributory superannuation scheme which appeared to be very favourable to most people compared to the state New Zealand Superannuation Scheme with the same contribution. But it was based upon the mean wage (line 1) being a fair representation of the average incomes. (Averages, means, and medians are explained in the next section.) Inspection of the table shows that it is not. Many households and adults are not in receipt of wages and salary (for a full year), and on average their income is less than average earnings. Moreover, because the distribution is skewed towards higher incomes, less than half of wage earners receive more than the mean wage. A better measure is the person in the middle of the distribution. Typically this median is 75 to 80 percent of the mean. Thus the ACT figures were favourable because they chose the wrong average, which meant a scheme which benefitted upper income people was presented as if it was of benefit to those on middle incomes.

This last point draws attention to the simple statistical point that an average says something about the middle of an income distribution, but little about the dispersion in the distribution. There are formal measures of these (including standard deviations, variances, coefficients of variation, gini coefficients) which one will find in standard statistical textbooks. That adds to the difficulties. Not only are there numerous relevant definitions of income, but there is no simple way of summarizing an income distribution to determine uncontroversially whether there is an increase or decrease in equality. Consider the following three cases of the income distribution shared between five people A, the poorest, to E, the richest, and in which the total income in each case is the same (150).

	CASE 1	CASE 2	CASE 3
A	10	15	13
B	20	25	18
C	30	30	28
D	40	35	38
E	50	45	53

In Case 2, 5 units of income has been taken away from the two richest and given the two poorest. Undoubtedly Case 2 involves a more equal distribution than Case 1, on any common sense or rigorous definition. But in Case 3 in comparison to Case 1, 2 units have been taken away from the middle income people (B,C,D), and shared between the poorest and the richest. Is Case 3 more or less equal than Case 1? There is no rigorous value free way of answering this question. We can add our values, but a journalist’s story may not be nearly so interesting if the “objective” expert has to admit the values they are using. On the other hand the reader/viewer is surely entitled to know that what is being reported is not some objective statistical fact, but a lot of opinion.

Defining Some Concepts

The previous section has been largely about problems which are so fundamental they puzzle economists. There is an underlying conceptual structure, which if misused compounds the confusion. Here are some frequently muddled standard dichotomies. (To get the main ideas across, some simplifications have been made.)

Wealth is the stock of assets held by a person: income is the flow of (mainly) money revenue to an individual which can be spent and save (so it excludes wealth transfer, such as receipts form selling a house, and employment costs). Income may be earned from labour activities (including wages and salary and the labour earnings of the self employed), it may be market consisting of earned income and income from wealth (interest, dividends, profits of the self employed, (possibly capital gains), and so on). Typically, but not always, total income includes market income plus social security benefits.

Gross income, without any tax deductions: net income is after income taxes are deducted (and usually) social security benefits added. Sometimes the expressions before tax, and after tax are used. After tax (and benefit) income is called disposable income.

An income level is some many dollars per period: an income share refers to the share of a group (often a decile or ten percent of the total individuals/households involved). A fall in income share need not necessarily mean a fall in the income level, since the shares of other income levels may have risen even further.

A percent increase measures the relative increase in the variable: a percentage point increase measures the absolute increase in a percent. For instance if an income share rises from 6 percent to 9 percent, that is a 50 percent increase, and a 3 percentage point increase. (Incidentally, try not to use a percentage when it exceeds 100 percent. Even if the journalist does not get it wrong – about four out of five times they do – the public will. For instance a 200 percent increase, means the new level is 3 times the old one.

The mean income is the total income divided by the total number of units; it is the most common form of “average”: the median income is the income of the middle person ranked from top to bottom. both are measures of “central tendency” of a distribution (as is the mode, or most common income). As a general rule means have superior statistical properties, but the median, may be a better indicator if the concern is a typical person because it is not so influenced by the extreme top incomes. Because the public is uneasy about “median”, use “middle income” or “the income of the person in the middle of the distribution.” Note there are also measures of the dispersion, or the width, of a distribution, the most common being the standard deviation.

To do justice to poverty, a state of economic deprivation, would take as much space as this again. Absolute poverty is when the person (or household) has insufficient material goods and services to subsist on: relative poverty is when the individual has insufficient to be able to participate and belong to their community. There is no official poverty line, but the most commonly used one is that set by the 1972 Royal Commission on Social Security as the level for someone on a social security benefit. Subsequent research suggests this benefit datum level (BDL) remains a plausible estimate of a relative poverty line. In March year 1998 prices the BDL was about $300 per week for a married couple. In the March 1993 year about 16.3 percent of the population had incomes less than the BDL. There is a vigorous debate, some of the contributions to it are nutty, few are value free.

Dealing with the Income Distribution

The basic advice remains “dont.” If the chief reporter has asked you to do an income distribution story and the usual avoidance strategies have failed (refusal, handing to a subordinate, taking a sickie, asking for a pay rise, resignation) the following seem to be the main strategies to prevent yourself preforming like a New Zealand middle order batsman.

1. Be humble. You dont know much about it, so admit that when you are pressing the expert.

2. Watch for flannel from the expert. If they cant explain it to you, they probably dont understand it themselves.

3. Repeat what you have been told in your own words. Show the expert any text you have prepared.

(The problem here is how do you identify an expert. Some economists are so keen for media coverage they will claim expertise on anything, especially where they can expound their personal values in the guise of objectivity. Most of the top journalists I know build up a relationship with good economists, who tell them who is the experts in the field.)

4. Ask the expert, if he or she has not indicated already, where they are making personal value judgements (as well as technical judgements). Ask them who is likely to give an alternative account of the story.

5. Beware of your own values. Why have I never heard a journalist mention that a hike in interest rates means that those lending to financial institutions will receive a hike in their incomes? Why do they always emphasize mortgages (according to the 1996 census only in 37 percent of private dwellings do the usual residents make mortgage payments, 32 percent own their own home without mortgage, 27 percent make rent payments, and the rest dont pay anything)? OK. so this is a rhetorical question. I reckon the answer is that most journalists are so badly paid they still have mortgages. In other words, journalists have personal values, like economists, and you need to be aware of them too.

Go to top

Bibliography

Here are works that should be in your library. I have omitted a few poor quality references, like one which does not even source its data.

The official data is in sources like
Statistics New Zealand/Department of Statistics (various years) New Zealand Census of Population and Dwellings. (various publications)
Statistics New Zealand/Department of Statistics (various years) New Zealand Official Year Book.
Statistics New Zealand/Department of Statistics (various years) Incomes, Wellington.
Department of Statistics (1990) The Fiscal Impact on Income Distribution 1987/88.

Easton, B.H. (1983) Income Distribution in New Zealand, NZIER Research Paper No 28, Wellington. (This has a lot of useful definitions.)
Easton, B.H. (1996) “Income Distribution”, in B. Silverstone, A. Bollard, & R. Lattimore (eds) A Study of Economic Reform: The Case of New Zealand, North Holland. (Instead of saying this is the most authoritative study of the 1980s and early 1990s, let me state that this is probably why I was asked to write this section.)
Mowbray, M. (1993) Incomes Monitoring Report: 1981-1991, Social Policy Agency. (Hopefully there will be an update to this report.)
New Zealand Planning Council (1988) For Richer or Poorer: Income and Wealth in New Zealand, Wellington.
New Zealand Planning Council (1990) Who Gets What? The Distribution of Income and Wealth in New Zealand, Wellington.

Also the Social Policy Journal of New Zealand, which has most of the contemporary debate, and New Zealand Economic Papers. An important paper is B.H. Easton,“Poverty in New Zealand: 1981-1993”, New Zealand Sociology, Vol 10, No 2, November 1995, p.182-213.

Go to top

Definitions are in the text above except as follows this table. (Adult: over 15 years old.) The website has limitations relative to a page, which means the table here is less attractive.

ESTIMATES OF AVERAGE INCOME ($ p.w.)

Name	Average	House- hold	All Adults	Male Adults	Female Adults	All	Notes Below
1. Gross Earnings	Mean		620	693	533		1
2. Gross Income	Mean		409	506	326		2
3. Gross Income	Median		300	430	239		3
4. Gross Income	Mean	883					4
5. Gross Income	Median	667					5
6. Gross Income	Mean					323	6
7. Gross Income	Median		244			244	7
8. Gross Income	Mean	879					8
9. Gross Income	Median	667					9
10. Gross Income	Mean					318	10
11. Gross Income	Median					242	11
12. Gross Income	Mean		426	541	317		12
13. Gross Income	Median	0	291	428	242		13
14. GDP	Mean					491	14
15. GDP	Mean		635				15
16. Net Income	Mean					305	16
17. Net Income	Mean		395				17

Notes

1. Gross Earnings: Total wage bill divided by Full time equivalent employees (i.e. part-time people treated as a fraction). (Feb, 1997) Source: Quarterly Employment Survey, Table 14.13 of NZOYB

2,3. Gross Income per adult. (June 1997) Source: New Zealand Income Survey, Table 14.14 of NZOYB

4,5. Gross Income per household. (1996-97) Source: Household Economic Survey, Table 6.13 of NZOYB

6,7. Gross Income per person. (1996-97) Source: Household Economic Survey, Table 6.13 of NZOYB

8.9: Gross Income per household. (1995-96) Source: 1996 Population Census, 1996 National Summary, Table 37.

10,11. Gross Income per adult. (1995-96) Source: 1996 Population Census, 1996 National Summary, Table 37.

12,13. Gross Income per person. (1995-96) Source: 1996 Population Census, 1996 National Summary, Table 37.

14. GDP per person. (1996-97) Source: SNA (System of National Accounts), Table 17.1 of NZOYB

15. GDP per adult. (1996-97) Source: SNA (System of National Accounts), Table 17.1 of NZOYB

16. Net Income per person. Source: SNA (System of National Accounts), Table 17.9 of NZOYB

15. Net Income per adult. (1996-97) Source: SNA (System of National Accounts), Table 17.9 of NZOYB

Go to top