Legistics
Rounding of Numbers

Numbers are often rounded in order to emphasize the important or relevant digits in a numerical expression and to produce a simpler number that is easier to read and use. The process of rounding consists in omitting one or more of the least significant digits in the number and adjusting the remainder in accordance with specific rules. These rules may vary slightly depending on the desired purpose (readability, precision, etc.) and the type of information (statistical, scientific, etc.).

The following are examples of rounding provisions using the most common rules that are appropriate for legislative texts:

Monetary Amounts

(X) If an amount to be used for a taxation year contains a fraction of a dollar, the amount is to be rounded to the nearest whole dollar or, if the amount is equidistant from two whole dollars, to the higher of them.

For example:

Decimals

(X) If the quotient referred to in paragraph … contains a fraction, the fraction is to be expressed in decimal form and rounded to three decimal places, the digit at the third decimal place being increased by one if the digit at the fourth decimal place is 5 or more.

OR

(X) The rate determined in accordance with the formula is to be expressed in decimal form and rounded to the nearest one-thousandth or, if the rate is equidistant from two consecutive one-thousandths, to the higher of them.

For example:

Truncation

Although it is more a matter of shortening the numerical expression than rounding it, truncation is often mentioned in the rules for rounding. It consists in simply dropping one or more digits after the decimal point. For example 1.677754 truncated to the fourth decimal place becomes 1.6777. It should be noted that this is different than if the number had been rounded to the fourth decimal place, which would be 1.6778.